This PDF file contains the front matter associated with SPIE Proceedings Volume 6794, including the Title Page, Copyright information, Table of Contents, Symposium Welcome, and Symposium Committee listing.

Brake system of railway vehicles has a crucial role for the safety as well as riding quality of passengers. Its core technology for successful development of the brake system is to design of ECU (Electric Control Unit) containing antiskid control, brake blending control, load compensating control, and so on. Each development procedure of ECU involves a verifying test for each step and is completed by an evaluation test of the comprehensive performance verification for the overall systems. The development tool which is applied to this objective is the HILS (Hardware-In-the- Loop-Simulation). In order to design a controller, a good representative model of the system is needed. This paper describes the dynamic modeling of brake system of railway vehicle for HILS and analyzes some dynamic behavior under emergency braking force.

In urban transit systems, railway vehicles are often required to negotiate tight curves. During curve negotiation, the wheelsets of conventional vehicles generally misalign radically with the track increasing wheel/rail contact forces and resulting in increased wheel and rail wear, outbreak of squeal noise, fuel consumption, and risk of derailment. This paper presents active steering controller design in the railway systems using Linear Quadratic Gaussian. Simulation results have been shown that the proposed LQG method methodology robustly yields uniform performance with adequate response over the axle weight and wheel conicity variation range.

This paper presents a passivity based control combined with velocity estimation by the extended Kalman filter for a magnetically levitated flexible beam with both ends free. Passivity analyses result in that the system can be decomposed into two passive subsystems: a mechanical subsystem that consists of the flexible beam with both ends free and an electrical subsystem that is comprised of electromagnets. An output feedback controller combined with velocity estimation by the extended Kalman filter for the mechanical subsystem computes desired force required to achieve position convergence and vibration suppression of the flexible beam. In a practical point of view, economic considerations restrict accurate measurements of both position and velocity of the supporting object while the control strategy requires accurate measurements of them. Thus, we employ the extended Kalman filtering technique in order to reject noise in the velocity signals at high sampling frequencies and/or micro-displacement vibrations. A feed forward controller that generates the desired force is designed for the electrical subsystem. Effectiveness of the proposed controller and velocity estimator is demonstrated by a numerical simulation.

The nuclear steam generator feedwater control system is designed by the robust control methods. The design is divided into two steps. First, the feedwater controller in the feedwater station is designed by H_∞ and MWS methods. Then the controller located on the feedback loop is designed both by classical PID and by robust technique. It is found that the feedback controller of simple PID whose coefficients vary with the power is proper for the system performance. The simulations show that the hybrid system of H_∞ and PID has a good performance with proper stability margins.

A novel approach to back-EMF detection of sensorless brushless DC motor drive systems was proposed. Theoretical analysis of back-EMF sensing is presented, and indicates the back-EMF signal can be extracted for each floating phase whether the PWM is on or off time. As a result, the proposed method rejects the high frequency switching noise and no filtering. Moreover, it provides a highly accurate zero crossing point. This proposed sensing scheme is implemented into a microcontroller for a sensorless BLDC motor. The microcontroller has been successfully applied to the aeromodelling motor. The applied results verify the theoretical analysis and demonstrate that the microcontroller provide a good motor performance.

As an important part of diesel engine, the speed regulator plays an important role in stabilizing speed and improving engine's performance. Because there are so many model parameters of diesel-engine considered in traditional PID control and these parameters present non-linear characteristic.The method to adjust engine speed using traditional PID is not considered as a best way. Especially for the diesel-engine generator set. In this paper, the Fuzzy PID control strategy is proposed. Some problems about its utilization in electronic speed regulator are discussed. A mathematical model of electric control system for diesel-engine generator set is established and the way of the PID parameters in the model to affect the function of system is analyzed. And then it is proposed the differential coefficient must be applied in control design for reducing dynamic deviation of system and adjusting time. Based on the control theory, a study combined control with PID calculation together for turning fuzzy PID parameter is implemented. And also a simulation experiment about electronic speed regulator system was conducted using Matlab/Simulink and the Fuzzy-Toolbox. Compared with the traditional PID Algorithm, the simulated results presented obvious improvements in the instantaneous speed governing rate and steady state speed governing rate of diesel-engine generator set when the fuzzy logic control strategy used.

This paper presents a passive control scheme for a force reflecting bilateral teleoperation system via the Internet. To improve the stability and performance of the system, the host and client must be coupled dynamically via the network and Internet technology provides a convenient way to develop an integrated teleoperation system. However, as use of Internet increases, congestion situation of network increased and transmission time and packet loss increased accordingly. This can make system unstable at remote control. In this paper, we present an optimal passive control scheme for a force reflecting bilateral teleoperation system via the Internet and we investigated how a varying time delay affects the stability of a teleoperation system. A new approach based on an optimal passive control scheme was designed for the system. The simulation results and the tracking performance of the implemented system are presented in this paper.

In this paper, the problem of an adaptive tracking robust control for robot manipulator is studied. The proposed controller always guarantees closed-loop stability, and the stability of the proposed controller is proven by Lyapunov method. The proposed controller can easily be implemented, and the tracking error and adaptation parameter error are uniformly ultimately bounded. Finally, a simulation illustrates the performance of the scheme.

This paper considers robust critical control of uncertain systems which require that tracking errors and control inputs are strictly maintained within pre-specified constant bounds. To achieve such critical control in the presence of unknown external disturbances, measurement noises and parameter variations of plants, the disturbance observer-based control is applied. A controller is designed as a 2-degree-of-freedom controller which consists of a servo compensator and a disturbance observer. Through a numerical example of servo control system design of the actively-controlled bed for ambulances, it is shown that, even if a measurement noise is fairly small relative to an output, it gives much influence on achievement of the critical control and so it is important to consider an effect of the noise in design explicitly.

UV lamp systems have been used for cleaning of display panels of TFT LCD or Plasma Display Panel (PDP). However, the needs for high efficient cleaning and low cost made high voltage plasma cleaning techniques to be developed and to be improved. Dielectric-barrier discharges (DBDs), also referred to as barrier discharges or silent discharges have for a long time been exclusively related to ozone generation. In this paper, a 6kW high voltage plasma power supply system was developed for LCD cleaning. The -phase input voltage is rectified and then inverter system is used to make a high frequency pulse train, which is rectified after passing through a high-power transformer. Finally, bi-directional high voltage pulse switching circuits are used to generate the high voltage plasma. Some experimental results showed the usefulness of atmospheric plasma for LCD panel cleaning.

In this paper, we examine a design method for unknown input observers for non-minimum phase systems. The unknown input observer has been used to estimate the state variable of the plant in the presence of unknown input. In addition the unknown input observer is applied to the systems with disturbance inputs or unknown varying parameters. Initially, the unknown input observer is examined by Kudva, Viswanadham and Ramakrishna. According to past studies, the unknown input observer for the plant (A,B,C,0) can be designed if and only if following 2 expressions hold true: (1) rankCB= rankB and (2) the plant (A,B,C,0) has no invariant zero in the closed right half plane. Many papers have been considered to design unknown input observers, even if above mentioned 2 expressions are not satisfied. In this paper, we propose a new design method of unknown input observers for non-minimum phase plants.

The modified repetitive control system is a type of servomechanism for the periodic reference input. That is, the modified repetitive control system follows the periodic reference input with small steady state error, even if a periodic disturbance or uncertainty exists in the plant. Using previously proposed modified repetitive controllers, even if the plants does not includes time-delay, the transfer function from the periodic reference input to the output and that from the disturbance to the output have an infinite number of poles. When the transfer function from the periodic reference input to the output and that from the disturbance to the output have an infinite number of poles, it is difficult to specify the input-output characteristic and the disturbance attenuation characteristic. From the practical point of view, it is desirable that the input-output characteristic and the disturbance attenuation characteristic are easily specified. In order to specify the input-output characteristic and the disturbance attenuation characteristic easily, the transfer function from the periodic reference input to the output and that from the disturbance to the output are desirable to have a finite number of poles. Yamada et al. proposed the concept of simple repetitive control systems such that the controller works as a modified repetitive controller and the transfer function from the periodic reference input to the output and that from the disturbance to the output have a finite number of poles. In addition, Yamada et al. clarified the parametrization of all stabilizing simple repetitive controllers. However the method by Yamada et al. cannot be applied for the plant with uncertainty. The purpose of this paper is to propose the parametrization of all robust stabilizing simple repetitive controllers for the plant with uncertainty.

In this paper, a novel sliding mode controller is proposed by using neural network for IPM machine. The current control for interior permanent magnet machines is more complicate than surface permanent magnet machine because of its torque characteristic depending on the reluctance. For high performance torque control, it requires state decoupling between the dcurrent and q-current dynamics. However the variation of the inductances, which couples the state dynamics of the currents, makes the state decoupling difficult. This paper presents a novel approach for fully decoupling the states cross-coupling using sliding mode control with neural network. The sliding mode control method is based on the error between reference currents and the currents with state decoupling which have to follow the references. In the conventional sliding mode control, the dynamic of sliding surface is not as same as nominal dynamic of original system. To overcome this problem, this paper proposes a new design method of a sliding surface without defining any additional dynamic state by using neural network. Finally, the proposed sliding surface can have the dynamics of nominal system controlled by PI controller.

The SVM is one of the methods which can introduce the statistical learning theory for solving the pattern recognition problem with small samples and learning problems such as function estimation. The relationships between the states with desirable responses can be expressed by some functions and these functions are estimated by using SVM. There are classification problems and regression problems in support vector machines. Only the support vector regression problem is used in this paper. This paper proposes a very novel method which makes it possible that state feedback controller can be designed for unknown dynamic system with measurable states. The SVR algorithm is used for the identification of inputoutput relationship. A virtual state space representation is derived from the relationship and the SVM makes the relationship between actual states and virtual states. For unknown dynamic systems, a state feedback controller can be designed based on the virtual system and the SVM makes the controller being with actual states. The results of this paper can give many opportunities that the state feedback control can be applied for unknown dynamic systems. The first step of this design method is to identify the input-output relationship of the unknown system as a transfer function by using SVR. Next step is to design a virtual system based on the transfer function. Final step is to derive the relationship between the actual states and virtual states by using SVM. The linear kernel function is used in SVR. A state feedback controller is designed based on the virtual system and the virtual system can be replaced by actual states by using the above relationship and it gives the actual states feedback controller. And simulation results are provided to show the performance of the proposed control method. Finally, the results of this paper make it is possible to the state feedback control theory to be used for unknown dynamic systems. This result can be expected to be applied to unknown nonlinear systems.

We describe an integrated local path planning method whereby a particle filter is used for computation of the steering direction based on the Vector Field Histogram. In the original VFH, a typical polar histogram contains peaks (i.e., sectors) with a high polar obstacle density (POD) and valleys (i.e., sectors) that contain low polar obstacle density. In order to determine a desired path allowable for the width of the vehicle, the valley below some threshold is selected for the traversable local path. In the proposed approach, the particle filter is used to select the widest valley without the predefined threshold over the polar histogram and the measurement model of the particle filter is used for the uncertainty of each cell in the VFH as well. Experimental results from a mobile robot traversing the obstacle courses demonstrate the proposed approach successfully selecting and driving the most traversable path.

The use of the interior permanent magnet synchronous motor (IPMSM) has increased in the industrial field because of its excellent characteristics such as high efficiency, wide range of speed operation, and flexibility of the rotor structure. Therefore, IPMSM, especially one using the rare earth permanent magnet, is being actively studied. In IPMSM using the rare earth permanent magnet, typically the permanent magnet is deeply embedded in the rotor. The washing machine has large load variation and needs high torque. So, IPMSM is better than any other motor types for washing machines. This paper suggests a control algorithm of IPMSM for washing machines. Some experimental results are given in order to show the feasibility of the proposed control schemes for washing machines.

Hybrid electric vehicle (HEV) has been widely studied recently due to its high potential in reduction of fuel consumption, exhaust emission, and lower noise. Because of comprised of two power sources, the HEV requires an energy management system (EMS) to distribute optimally the power sources for various driving conditions. The ITRI in Taiwan has developed a HEV consisted of a 2.2L internal combustion engine (ICE), a 18KW motor/generator (M/G), a 288V battery pack, and a continuous variable transmission (CVT). The task of the present study is to design an energy management strategy of the EMS for the HEV. Due to the nonlinear nature and the fact of unknown system model of the system, a kind of simplex method based energy management strategy is proposed for the HEV system. The simplex method is a kind of optimization strategy which is generally used to find out the optimal parameters for un-modeled systems. The way to apply the simplex method for the design of the EMS is presented. The feasibility of the proposed method was verified by perform numerical simulation on the FTP75 drive cycles.

BLDC motors have many advantages over brushed DC motors and induction motors. So, BLDC motors extend their application to many industrial fields. In this paper, the digital simulation and modeling of a 7-phase brushless DC motor have been presented. The 14-switch inverter and a 7-phase brushless DC motor drive system are simulated using hysteresis current controller and logic of switching pattern with the Boolean¡s function. Through some simulations, we found that our modeling and analysis of a 7-phase BLDCM with PWM inverter would be helpful for the further studies of the multi-phase BLDCM drive systems.

20 years has passed since the prototype algorithm of Human Simulated Intelligent Control (HSIC) was brought out in 1979. A basic HSIC theory system and the systematical design method have formed, and successfully applied in many fields, especially in the nonlinear control of a Double Pendulum System. A Multi-Pendulum System (MPS) is a typical nonlinear, multi-variable, strong-coupled and under-actuated, complex system. The study on a MPS has been a hotspot in the field of automation control. The successful application of HSIC in swing-up and handstand control on a DPS adequately testified the validity of HSIC theory. This paper will focus on two topics: (1) structural design of controller for swinging up a circular-rail double pendulum (CDP) form its down-down equilibrium position to down-up equilibrium position (DD2DU); (2) real-time adjustment of controller parameters.

An electromagnetic levitation controller design considering vehicle-guideway interaction is presented. Even though to design a Maglev controller to make a vehicle steady suspension under guideway deflection is a key technology in Maglev vehicle system, the vehicle-guideway interaction effects haven't well explored and yet settled. The vehicle is approximated to a single moving reaction force and the primary and secondary suspension dynamics are applied to the model. The Bernoulli-Euler beam equation is used to model the dynamic deflection of guideway, which acts as a disturbance to the Maglev control system. Based on the model, a PID controller and a state feedback controller with a state observer are proposed. The feasibility of the controller is verified through the Simulink simulation.

Sliding mode control with disturbance estimator scheme is proposed for the control of a maglev vehicle. The conventional sliding mode control requires a priori knowledge of the upper bounds of parameter variations and disturbances to assure robustness, but it is hard to know the upper bounds of guideway disturbances like guideway irregularity and guideway deflection in a maglev vehicle. A sliding mode control with disturbance estimator, which offers a robust control performance without a priori knowledge about the disturbances, is derived to maintain the desired robustness and minimize the effect of guideway disturbances. Simulation results of the proposed controller are given to illustrate the effectiveness of the proposed scheme.

Heat-pump system has a special feature that provides heating operation in winter season and cooling operation in summer season with a single system. It also has a merit that absorbs and makes use of wastewater heat, terrestrial heat, and heat energy from the air. Because heat-pump system uses midnight electric power, it decreases power peak load and is very economical as a result. By using the property that energy source is converted to low temperature when losing the heat, high temperature energy source is used to provide heating water and low temperature energy source is used to provide cooling water simultaneously in summer season. This study made up a heat-pump system with 4 air heat sources and a water heat source and implemented the optimal operation algorithm that works with numbers of heat pumps to operate them efficiently. With the heat-pump system, we applied it to cooling and heating operation in summer season and in winter season operation mode in a real building.

The problem in improving the high positioning precision of permanent magnet linear synchronous motor (PMLSM) for direct drive conveyer is the large detent force, which contains two components: one is the cogging force caused by the interaction between the permanent magnet (PM) and the iron core, the other is the end effect caused by the finite mover length. In this paper a force compensator in the field oriented control (FOC) algorithm is employed to reduce this detent force, so the thrust and the normal force characteristics analysis of PMLSM are of the most importance. First the characteristics of the thrust and the normal force are simulated by finite element method (FEM), and the relation of the thrust and the normal force to the q-axis current is analyzed. Then the components of detent force are defined by Fourier series using the curve fitting method, and compensated by injecting the instantaneous currents using the FOC method. Finally the dynamic characteristics of this PMLSM for direct drive conveyer are investigated, and the numerical results are reported to validate the effectiveness of this proposed method.

In recent years, wireless sensor networks (WSN) have attracted a great deal of attention in the research community. This is due to the applications that will be enabled once wireless sensor networks are in place. Wireless sensor networks are emerging as a promising solution for various types of futuristic applications both for military and public. Security in wireless sensor networks is very crucial as these networks are expected mainly to operate in the hazardous and hostile environments. Efficient key management could guarantee authenticity and confidentiality of the data exchanged among the nodes in the network. Depending on the application requirements, the WSN may operate in star topology, in which the communication is established between sensors (i.e. nodes or devices) and a single central controller, called the WSN coordinator. In order to make the power consumed as efficiently as possible, a typical WSN architecture is first described and then energy strategies and model are specified on account of the hardware constraints of the sensors in the paper. Furthermore, energy estimating method is formulated and deduced. At last, computer simulation based on the above architecture and model gives the analysis illustrations and the results. Experiments and simulation show how energy savings of 50% can be achieved and can delay the deadline of the WSN system. This energy model can also be used for power management in any system characterized by different levels of power consumption in various stages of shutdown.

The authors proposed the singular LTR method for the integral controller design based on the step disturbance cancellation. In this paper, the singular LTR method is extended to more general class of disturbances. It is assumed that the disturbance is modeled as an output of a linear function generator. By the separation principle, the optimal disturbance cancellation controller is constructed using an observer estimating the plant state and the disturbance state. It is shown that the partial LTR method can be used for the efficient design of the disturbance cancellation controller.

PRT( Personal Rapid Transit ) system is a automated operation, so that it is important exactly finding position of vehicle. Many of PRT system has accepted the GPS system for a position, speed, and direction. in this paper, we propose a combination of Kalman Filter and H∞ Filter known as Hybrid Kalman/ H∞ Filter for applying to GPS navigation algorithm. For disturbance cancellation, Kalman Filter is optimal but it requires the statistical information about process and measurement noises while H∞ Filter only minimizes the "worst-case" error and requires that the noises are bounded. The new Hybrid Filter is expected to reduce the worst-case error and exploit the incomplete knowledge about noises to provide a better estimation. The experiment shows us the ability of Hybrid Filter in GPS navigation algorithm.

The forced commutation makes that the trigger angle changes between π and 3π/2 so that the inverter absorbs the leading current of the power network. This improves the power factor of the whole system. This paper builds the eightpulse generator model, the simulating model of the forced commutation inversion bridge, the model of the cascade control system with forced commutation inverter eight-pulse commutation inversion circuit model, and proofs the feasibility of the system. This paper also analyzes the voltage wave and the current wave everywhere of the main circuit at different control angle, puts forward that the forced commutation brings some problems that must be noticed and the setting problem of the relative system parameter.

This paper deals with the global optimization algorithm of the Bilinear Matrix Inequalities (BMIs) based on the Unimodal Normal Distribution Crossover (UNDX) GA. First, analyzing the structure of the BMIs, the existence of the typical difficult structures is confirmed. Then, in order to improve the performance of algorithm, based on results of the problem structures analysis and consideration of BMIs characteristic properties, we proposed the algorithm using primary search direction with relaxed Linear Matrix Inequality (LMI) convex estimation. Moreover, in these algorithms, we propose two types of evaluation methods for GA individuals based on LMI calculation considering BMI characteristic properties more. In addition, in order to reduce computational time, we proposed parallelization of RCGA algorithm, Master-Worker paradigm with cluster computing technique.

This paper presents a new technique of designing a control system for PWM converter based on the LabVIEW and an advanced FPGA reconfigurable embedded control and acquisition equipment, Compact RIO (cRIO). Assuming that a PWM controller is composed in the double loop structure and that the inner loop controller is given, a constructive design system that is able to make data acquisition, modeling, controller design, simulation with PC, and hardware in the loop simulation (HILS) is represented. In this approach, a key point is to obtain an equivalent linear model of the inner loop system including nonlinear components by using the closed loop identification and to use it when the outer loop controller is designed. Experimental results applied to a real PWM converter will be given as a demonstration.

In the paper, a novel trajectory shaping scheme to improve the manipulation skill of the articulated manipulator for the liquid container transportation is proposed. The paper investigated the transient free surface behavior of liquid inside a cup during the horizontal transportation. Based on the observation of the free surface vibration of the liquid, the adaptive command shaping filter is proposed for the generation of the transportation trajectory to reduce the fluctuation magnitude without increasing the time required to travel the given transportation distance. Simulation results verify the effectiveness of the proposed scheme. An experimental testbed of an articulated manipulator with a simplified model of a liquid container is established. The proposed manipulation approach is implemented in the experimental setup and the experimental results also verify the effectiveness of the proposed scheme.

In the high-precision servo system, not only request quickness but also response characteristics without overshoot, usually request strong robot for position loop. The traditional control system design usually is finished under the linear condition which neglects some factors. It makes the more differences between designed model and actual system, so the general PID adjuster and the single basic fuzzy control are different to meet the need. The paper designs a complex fuzzy controller based on double structure and variable coefficient, at the same time adopts Cooperative Processing module with fuzzy reasoning function, which make the system fuzzy reasoning speed improved and realize the online reasoning. So the system capability is improved, the experiment proved the system has the good capability.

The intention of this paper is to present the control system of the physical simulation for the Three Gorges HVDC transmission system. It covers all essential control functions which are involved in the continuously working power flow control and power modulation including converter firing and pole current control. This system include two convert stations, They design the fault arbitrator for the dual-controller system based on fault-tolerant technology and figure-safety-comparing technology. The arbitrator can endure long working time with high reliability. Controller A and B not only works independent, but also can controlled by the master controller from CAN bus. At the same time, because the master controller is a industrial PC, we can easy adjust and enhance the functions. This characteristic is especially fit the physical simulation system.

This paper presents a networked simulation for an integrated chassis control system of AFS (Active Front Steering) and ITD (Intelligent Torque Distribution). Integration of each chassis control system is used to overcome the limit of performance when each chassis controller is used individually. We show an integration method AFS and ITD under the supervisory controller which determines the operation modes based on vehicle variables. The experimental set-up for the networked simulation of the integrated chassis control consists of four microcontroller boards, a steering wheel sensor and a data acquisition board where all the microcontrollers and a sensor are communicated using the CAN protocol. It is shown by experimental results that the integrated control system can achieve better performance than simply combined individual controllers in the sense of energy consumption.

Sensitivity of fiber optic sensor is critical concern, in case of concealed application such as buried fiber-optic sensor, because ground soil behaves as buffer to isolate from outside disturbance by intruder. Thus coherent intrusion sensor is getting popularity where the sensitivity is primary concern, since the superior sensitivity comes from optical phase relationship of two received light signal. The performance of coherent sensor depends on the coherent degree of light source. Conventional laser light source is not sufficient for this application, thus enhancing technique to increase coherence length is required to have stable sensor response. Fiber laser is the suitable candidate because emission spectra can be adjusted to the wavelength where transmission loss through glass fiber medium is at its minimum. The performance of erbium doped fiber laser with single optical feedback loops are compared with the one using selfinjection locked DFB laser experimentally in Mach-Zehnder interferometer configuration. The response of coherent fiber-optic sensor, in Mach-Zehnder interferometer configuration, using erbium doped fiber laser with single optical feedback loops as a light source is attempted to the various stimulus, including pressure, open flame and vehicle passing, for the purpose of evaluating sensor application. The possible application of this fiber laser with enhanced coherence is in zone perimeter security within ~20km where buried fiber-optic sensor installation is necessary.

This paper presents a decentralized method for people tracking with multiple 2D laser range sensors (LRSs) allocated in environments. Moving people which are detected from laser images with the LRSs are tracked based on Kalman filter, the assignment algorithm and Covariance Intersection method. A rule based on track management system is embedded into the tracking system in order to improve the tracking performance. The experimental results of people tracking validate this proposed method.

Supported Bilayer Lipid Membrane (s-BLM) affords a great direction to study the lipid membrane. It becomes a key to study the stability of s-BLM. The study is adopting a method with optics to detect data not electronics, because it can avoid the disturbance which is brought by the electric charge. Based on the basic principles of SPR, we assembled a mini-SPR equipment, and it has the accuracy fitting for our works. To adjust the angle of the polarizer and move the fiber, we can make the spectrogram that membrane is coupled on the surface of the gold slice. Comparing to the reference spectrogram, we can make a certain relationship between the light absorbed maximum and the quality of the phospholipid molecules coupling on the surface. It can token the resident time with the time which the detected spectrogram resumes to the reference one. And the resonance angle has a definite shift. We can assess the stability of phospholipid BLM with the three references. The mini-SPR equipment can easily change the parts and the parameter with the up-and-down equipment, the movement of the fiber and replacing the triangular prism plated the thin metal. So we can explore and find the conditions of the stability of BLM effectively with it^[1].

A blood pressure sensor unit for ubiquitous healthcare monitoring was newly developed. The digital wrist band-type blood pressure devices for home are popular already in the market. It is useful for checking blood pressure level at home and control of hypertension. Especially, it is very essential home device to check the health condition of blood circulation disease. Nowadays many product types are available. But the measurement of blood pressure is not accurate enough compared with the mechanical type. It needs to be upgraded to assure the precise health data enough to use in the hospital. The structure, feature and output signal of capacitor type pressure sensors are analyzed. An improved design of capacitor sensor is suggested. It shows more precise health data after use on a wrist band type health unit. They can be applied for remote u-health medical service.

The capacity of a high speed camera in taking high speed images has been evaluated using CMOS image sensors. There are 2 types of image sensors, namely, CCD and CMOS sensors. CMOS sensor consumes less power than CCD sensor and can take images more rapidly. High speed camera with built-in CMOS sensor is widely used in vehicle crash tests and airbag controls, golf training aids, and in bullet direction measurement in the military. The High Speed Camera System made in this study has the following components: CMOS image sensor that can take about 500 frames per second at a resolution of 1280*1024; FPGA and DDR2 memory that control the image sensor and save images; Camera Link Module that transmits saved data to PC; and RS-422 communication function that enables control of the camera from a PC.

The high accuracy of the vehicle digital dashboard makes it difficult to check its error in real time. On taking the advantage of is production condition, the digital image processing method can be used to check the dashboard's precision automatically. The image edge detection method is the key of our dashboard check method. The snake model has been extensively used today. The GVF snake model overcomes the traditional snake model's shortcoming, it has a large capture range and is able to move into boundary concavities. But it still needs large amount of computation and is easily to be disturbed by noise. The wavelet-based multi-scale GVF snake took the advantage of the wavelet transform and GVF model. In the lower resolution, there were less wavelet coefficients and the GVF snake was easy to deform to the contour without much computation and was less interfered by noise. In higher resolution, with taking advantage of the initial position of the foregoing resolution, much more computation would be saved. Experiments show this method can detect the position of the pointer automatically and exactly.

H.264/AVC is the newest video coding standard jointly developed by the ITU-T Video Coding Experts Group and the ISO/ISE Moving Picture Expert Group. Compared to previous standards, aggressive compression techniques such as intra prediction, adaptive block size motion compensation, and 4x4 block based prediction are employed in this standard. As a result, this standard greatly outperforms other standards with significant bit-rate reduction. In this paper, we present a new intra-prediction algorithm for the H.264/AVC standard. The proposed algorithm generates additional prediction blocks by using previously reconstructed blocks in order to reduce the amount of bits for video representation. Experimental results show that the proposed algorithm is superior to conventional methods in terms of compression ratio.

In this paper, the aortic and mitral insufficiency murmurs analysis method using the wavelet packet technique is proposed for classifying the valvular heart defects. Considering the different frequency distributions between the normal sound and insufficiency murmurs in frequency domain, we used two properties such as the relative wavelet energy and the Shannon wavelet entropy which described the energy information and the entropy information at the selected frequency band, respectively. Then, the signal to murmur ratio (SMR) measures which could mean the ratio between the frequency bands for normal heart sounds and for aortic and mitral insufficiency murmurs allocated to 15.62-187.50 Hz and 187.50-703.12 Hz respectively, were employed as a classification manner to identify insufficiency murmurs. The proposed measures were validated by some case studies. The 194 heart sound signals with 48 normal and 146 abnormal sound cases acquired from 6 healthy volunteers and 30 patients were tested. The normal sound signals recorded by applying a self-produced wireless electric stethoscope system to subjects with no history of other heart complications were used. Insufficiency murmurs were grouped into two valvular heart defects such as aortic insufficiency and mitral insufficiency. These murmur subjects included no other coexistent valvular defects. As a result, the proposed insufficiency murmurs detection method showed relatively very high classification efficiency. Therefore, the proposed heart sound classification method based on the wavelet packet was validated for the classification of valvular heart defects, especially insufficiency murmurs.

In this paper, we propose a modified watershed algorithm considering zero-point of gradient to obtain a correct edge of a region. The conventional watershed algorithm decide priority considering gradient of a region to be extended. However if the grey levels of two neighboring seeds are quite different, the ridge line obtained by the conventional watershed algorithm is not coincided with the correct edge. It is probable that the point where gray level extreme value occurs will be estimated as an edge of two neighboring regions. Therefore more correct edge of a region can be obtained if the expansion of watershed algorithm is delayed where the extreme value occurs. The position where the gray level extreme value occurs corresponds to the position where the zero-crossing of gradient occurs. So we adjust the priority to postpone expansion at the zero-crossing point of gradient. We compare the proposed method with the conventional method and validate that this proposed method can extract more correct edge of a region.

In this paper we propose a method of optimal camera exposure estimation in the image for a structured light system. In structured light system, it is important to discriminate the patterns form the captured images which are illuminated by the projector. But every object in real environments has a different reflection due to object's material and surface's color property, so the precise discrimination of the projected pattern from the image in indoor environments is a hard problem where many objects are located together especially. And the camera exposure setting is not good. For better 3D range data, we estimate the optimal exposure at pixels so that makes the illuminated light by the projector can be captured properly by the camera. In order to obtain the optimal exposures about specific environments in structured light system, we should know how much intensity and/or brightness varies while the exposure is changed. So we introduce the novel method to estimate intensity by characteristic curves. This proposed method overcomes the conventional method's exposure problem and presents what is the optimal exposure at pixels is and how to obtain this setting. This has been proved to be feasible in many applications which need to measure the objects with different surface properties.

This paper presents a correlation model based error concealment method for Polyphase Downsampleing (PD) applied multiple description coding (MDC) using H.264. With MDC, when one description is lost due to errors, the lost description can be reconstructed from other descriptions correctly received. The PD based MDC provides coded stream capability of reconstructing pixels in the lost description from neighboring pixels which are coded in other descriptions. The proposed error concealment method explores correlation between pixel to be concealed and its neighboring pixels, then applies correlation model based concealment method to reconstruct the lost pixels. Simulation results show that the proposed method improves coded stream error resiliency performance by an average 0.62 dB PSNR enhancement in comparison with an existing PD based MDC method.

In the reconstruction of high resolution (HR) images from low resolution (LR) image frames, the accuracy in the inter-frame motion estimation is a critical factor. With a motivation that motion estimated over the reconstructed HR frames would have higher accuracy, several attempts have been made to perform motion estimation and HR reconstruction simultaneously. In this paper, we propose a HR reconstruction method, in which a regularized block matching for motion estimation and POCS-based reconstruction are performed in turns. In experimental results, regularized block matching produces reliable motion estimation since the motion of an image pixel is highly correlated with the motion in neighboring region. The improved accuracy of the motion estimation results in higher PSNR of the reconstructed HR images.

One of the most important problems faced when building Augmented Reality (AR) systems is registration error. In mobile and outdoor AR systems, dynamic errors are the largest source of registration errors. Hybrid tracking is a key technique in combating the effects of dynamic errors. In this paper we present a multi-stage close-loop correction algorithm for reducing dynamic registration errors in AR system. The control law in this algorithm is based on 2D visual servoing or image-based camera control that allows control of a virtual camera to reduce registration error. The experimental results illustrate the effectiveness of this algorithm.

As required of multi-level and network based fault diagnosis system for large-scale rotating machinery, a new kind of embedded intelligent local set is introduced in the paper for machine condition monitoring and fundamental diagnosis. Its functions include high-speed acquisition of vibration signals, time-frequency analyses, ANN-based fault diagnosis and remote communications. The experimental results show that it performs with higher efficiency with real time, reliability and accuracy in data acquisition, vibration monitoring and typical fault diagnosis.

Based on the Multi-analysis wavelet threshold denosing method which is put forward by D. L. Dohono and I. M. Johnstone, a new threshold function is posed. This new threshold function has many advantages over the classical soft and hard threshold function. It is clear in physics meaning and simple in expression. The experimental results show that this improved method is effective in removing white noise, and gives better RMSE performance and SNR gains than soft and hard threshold methods, and makes it possible to embody the advantages of the wavelet threshold denoising method more fully.

With regard to the over-segmentation of traditional watershed algorithm and the problems of sensitivity to noise, a new algorithm that can effectively restrain noise, eliminate image edges and detect the image edges exactly was presented. Firstly, morphological gradient image edge was enhanced by high frequency emphasize filter effectively. Then, the enhancement image is fit by B spline function many times. Finally, the watershed segmentation is used for the curved face. Experimental results demonstrate that watershed segmentation with the proposed algorithm resolves the phenomena of over-segmentation well. The scheme can detect the image edges exactly and improve segmentation precision.

In this work, a prototype Polyvinylidene Fluoride (PVDF) tactile sensor for endoscopic application has been developed. The sensor aims to measure hardness, which is one of the information of tactile perceptions, of biomedical tissue. This sensor is composed of two PVDF films, a silicone cylindrical column, and an aluminum cylinder. And the classification of hardness is concerned with the ratio of these PVDF outputs. In this paper, two sensors are fabricated using two silicone cylindrical columns with different Young's modulus. The performance evaluation of each sensor is conducted using 6 silicone rubbers as measuring object. The experimental results correspond with the simplified theoretical analysis and the proposed sensor can distinguish a difference of elastic property.

This paper is concerned with the development of a tactile sensor using PVDF (Polyvinylidene Fluoride) film as a receptor of the sensor to evaluate a detergent. Tactile sense is the most important sense in the sensation receptor of the human body along with eyesight. When the dish which washed cleanly is rubbed with a finger of human, good tactile sense and sound (vibration signal) like "Kyu-kyu" are obtained. From this tactile sense and sound, we judge that a dish becomes squeaky-clean. This tactile sense and sound are evaluation parameters when consumer selects a detergent. In this study, a tactile sensor using PVDF film as the receptor is fabricated. Sensory test of detergents was conducted. Measurement experiment by the sensor is carried out. Experiment results show that sensor output have good correlation with the result of human sensory test of detergent.

This paper presents the 2-DOF haptic device in which DC motors and MR brakes are equipped. To apply MR brakes to the 2-DOF haptic device, its kinematic requirement is described by Jacobian analysis. Through the analysis, MR brakes can be used to display the damping force in a multi-DOF haptic device when it has the orthogonal Jacobian such as Cartesian manipulator. A five bar linkage type haptic device, which has the nonorthogonal Jacobian, is developed to verify the theoretical effort. Experimental results show that the five bar linkage type can't display the 2 DOF virtual environments due to its kinematic constraints, that is, nonorthogonal Jacobian.

This research developed a position control system of the pneumatic cylinder to minimize an error of displacement of the pneumatic cylinder system and to increase the repeatability of displacement. The maximum air pressure of the position control test was 5 kg/cm² and the displacement accuracy of a stroke was measured by using a dial gauge. The supply- and discharge-side of air pressure and the length of the stroke of the pneumatic cylinder were adjusted and set before experiment. The test of the position control of the pneumatic cylinder was carried out 50 times at each supply- and discharge-side air pressure of 2.5/3.5, 3.0/4.0, 3.5/4.5, and 4.0/5.0 kg/cm² and replicated three times. The accuracy of the displacement of the pneumatic cylinder stroke increased as the supply- and discharge-side of air pressure increased with the fixed stroke length of 133 mm. Also it increased as the stroke length increased with the fixed supply- and dischargeside of air pressure of the pneumatic cylinder as 3.5/4.5 kg/cm². The most accurate displacement of the pneumatic cylinder (i.e., standard deviation of 0.01 mm) was obtained with the supply-and discharge-side of air pressure of 4.0/5.0 kg/cm²and the stroke length of 170 and 190 mm.

This research designed and developed a groove type die set which improved life span of the die set by eliminating pointcontacts of steel balls with the guide post. The guide post consisted of a steel-ball retainer, a steel-ball retainer stopper, a guide bush, a guide pin, a snap ring, and a spring. The steel-ball retainer has 72 holes with 8 columns of 9 holes in each column. The inner surface of the guide bush was grinded (surface roughness: Ra = 0.2μm, accuracy: 0 ~ -0.002 mm) after NC turning and it was treated with heat. Also, a line of small intermediate pocket was processed inside of the guide bush for lubrication and elimination of foreign materials. Guide grooves for steel balls were processed using a wire EDM (Electrical Discharge Machining) after the heat treatment. With such a design of the guide post stated above, loads against steel balls could be dispersed greatly by the line contacts through the guide groove between the guide pin and the guide bush, and the life span of the guide post could be expanded semi-permanently.

Humanoid by ROBO-BLOCK (robot block system) and the rational formula of robots were proposed. ROBO-BLOCK is composed of servo motors, the parts for servomotor rotor, the brackets for servomotor fixation, the board parts and the controllers. A robot can be assembled easily by ROBO-BLOCK. Meanwhile, it is convenient when the structure of the robot can be described easily as a character. The whole structure of the robot is expressed as rational formula of the robot to show molecule structure in chemistry. ROBO-BLOCK can be useful for not only the research but also the education. Creative student experiment was done in the college of industrial technology.

This paper discusses a structural analysis of a gripper with a simple gripping mechanism and clarifies deformation and gripping efficiency of the gripper. This gripper is analyzed as a rigid frame which is built-in at both ends. This structure is a statically indeterminate structure. To analyze such a structure, the virtual work method is used. Moreover, to obtain an unknown gripping force, deformation of the gripped object replaced with load cells is considered. The effectiveness of the constructed mathematical model is verified in a comparison with respect to gripping efficiency between a simulation with the model and an experimental result.

Hysteresis is one of major properties of air spring during its compression and rebounding sequences. This study shows hysteresis property of air spring theoretically and experimentally. At first, the theoretical model of the air spring was developed, considering the factor of heat exchange. And then experiments of force responses with displacement excitation signal of various frequencies were implemented. The simulations by using obtained model were performed to compare with experimental results. The results show that simulation results are close to experiment in low excitation frequency from 1 Hz to 0.1 Hz, but there exits some difference between them in very low excitation frequency (quasistatic process).

In order to solve the energy supply problem for ubiquitous equipments, a gyroscopic power generator is proposed which utilizes kinetic energy of human movement. In the generator, a rotor increases spinning velocity by the precession and friction caused by input vibration. In this paper, first, an analytical method of the rotor movement is presented. Then the output energy of the prototype generator is estimated by using the analysis.

This paper presents a new design of precision stage for the reticle in lithography process and a low hunting control method for the stage. The stage has three axes for X,Y, θ_Z, those actuated by three voice coil motors individually. The proposed precision stage system has three gap sensors and voice coil motors, and supported by four air bearings, so it do not have any mechanical contact and nonlinear effect such as hysterisis which usually degrade performance in nano level movement. The reticle stage has cross coupled dynamics between X,Y,θ_Z, axes, so the forward and inverse kinematics were solved to get an accurate reference position. When the stage is in regulating control mode, there always exist small fluctuations (stage hunting) in the stage movement. Because the low stage hunting characteristic is very important in recent lithography and nano-level applications, the proposed stage has a special regulating controller composed of digital filter, adjustor and switching algorithm. Another importance factor that generates hunting noise is the system noise inside the lithography machine such as EMI from another motor and solenoids. For reducing such system noises, the proposed controller has a two-port transmission system that transfers torque command signal from the DSP board to the amplifier. The low hunting control algorithm and two-port transmission system reduced hunting noise as 35nm(rms) when a conventional PID generates 77nm(rms) in the same mechanical system. The experimental results showed that the reticle system has 100nm linear accuracy and 1μ rad rotation accuracy at the control frequency of 8 kHz.

A new kind of motion base is proposed for the purpose of setting up a user-friendly interaction system on the desktop. It aims at motion in various situations with a large range of movement using a simple mechanism, and generating motions adequate for an interaction with a man and for computer application. In this paper, a prototype of the motion base is developed. Its mechanism consists of a combination of a wheel drive locomotion mechanism and a 3-degrees of freedom (DOF) parallel link mechanism in parallel. It is aimed at generating 3 DOF translational motion on the desktop with a simple mechanism as a whole and with a large range of movement, especially in the horizontal direction. In consideration of application for man-computer interaction, the prototype is controlled by a microprocessor, a microcontroller and FPGA. This paper presents a description of the concept of the motion base for a desktop interaction system, the prototype mechanism, the control system and the preliminary experiment which demonstrates typical translational motions.

Self-excited vibration of a film caused by air flow is analyzed to realize an electric power generator driven by fluid energy which exists in various artifacts. First, generation of the self-excited vibration is verified through experiments. Second, an approximate calculation method on the basis of the experimental results is contrived. Third, with the appropriate analysis, effect of the inlet air pressure, film tension, boundary pressure on the vibration frequency and easiness of vibration generation are clarified. Also the calculated results are compared with experimental ones to verify the validity of the analysis.

For thin steel plates, which are used in many industrial products including those of the automobile industry, we have proposed a magnetic levitation control system and confirmed its realization by means of a digital control experiment. However, the use of a limited number of electromagnets cannot suppress static deflection and high-order-mode elastic vibration, which are characteristics of a flexible magnetic material. To solve this problem, we have proposed a hybrid levitation control system for the thin steel plate using the magnetic force generated by permanent magnets, which have no operational costs, in the areas where the attractive force of the electromagnets is negligible. In this study, we attempt to determine the optimal placement of permanent magnets to reduce the deflection and elastic vibration of a thin steel plate under the generated magnetic field. To verify the usefulness of optimal placement of the permanent magnets, experiments concerning elastic vibration were performed on a magnetically levitated thin steel plate. As a result, it was confirmed that the control performance was improved by the optimal placement of the permanent magnets.

We have developed a noncontact guide system for a high-speed traveling elastic steel plate in which electromagnetic forces are applied by actuators at the edges of the plate to control its position. Recently, we have been examining a noncontact guide system, in which one of the loops of a belt-shaped thin steel plate is supported by a pulley, and another loop is guided without contact using electromagnets. It was confirmed experimentally that the loop shape of the thin steel plate changes with increasing traveling speed when no control is carried out. In this study, a basic examination of the technique for forming a guideway using electromagnets, in which the change in the uncontrolled loop shape is considered, was carried out. By comparing the results for guideways with several different shapes, the effect of the guideway shape on the vibration suppression performance of the steel plate was discussed experimentally. As a result, it was confirmed that the vibration suppression performance of the steel plate can be improved by adjusting the shape of the noncontact guide system that uses electromagnets to conform with that of the loop of the uncontrolled steel plate.

We have examined the improvement of ride quality and the reduction of riding fatigue brought about by the active control of the seat suspension of small vehicles such as one-seater electric automobiles. A small active seat suspension, which is easy to install, was designed and manufactured for one-seater electric automobiles. For the actuator, a maintenance-free voice coil motor used as a direct drive was adopted. For fundamental considerations, we designed a one-degree-of-freedom model for the active seat suspension system. Then, we designed a disturbance cancellation control system that includes the observer for a two-degree-of-freedom model. In an actual driving test, a test road, in which the concavity and convexity of an actual road surface were simulated using hard rubber, was prepared and the control performance of vertical vibrations of the seat surface during driving was examined. As a result, in comparison with the one-degree-of-freedom control system, it was confirmed that the control performance was improved by the two-degree-of-freedom control system that includes the observer.

This paper exposes the complete procedure of suppressing the chatter and how to realize it on a direct-drive CNC hobber. In a series of experiments, we discovered the phenomenon that vibration energy was centralized at a natural frequency of the machine gradually during the chatter occurred. Based on the phenomenon and the theory of regenerative chatter, a multi-degree of freedom model about the chatter of machine is built. According to this model, the way to identify chatter frequency and weak component is proposed. Taking the bobber spindle system as an example, several key steps are introduced: building optimization model of spindle, reducing the modified parameters based on sensitivity of anti-chatter and the optimization algorithm. Optimal result of hob spindle system is provided finally.

A new space and time resolved focusing elliptical curved crystal spectrometer has been developed and applied to diagnose X-ray of laser-produced plasma in 0. 2~2 nm region. According to the theory of Bragg diffraction, four kinds of crystal including LiF, PET, KAP, and MiCa were choosed as dispersive elements.The distance of crystal lattice varies from 0.4 to 2.6 nm. Bragg angle is in the range of 30°~67.5°, the spectral detection angle is in 55.4°~134°. The dispersive crystal sizes are 120×8×0.2mm.The characteristic of optical system is an elliptical geometry.The X-ray source is located at the front focal point.The X-rays diffracted by the elliptically curved crystal are focused at the rear focal point where a width-adjustable exit slit is positioned.The Curved crystal spectrometer mainly consists of dispersive elements, vacuum configuration, aligning device, spectral detectors and three dimensional (3D) micro-adjustment devices. The spectrographic experiment was carried out on the XG-2 laser facility.The PET and KAP crystals are adopted as the dispersive elements,which measure X-ray in the 0.44~0.81 and 1.33~2.46nm region. Emission spectrum of Al plasmas and Ti plasmas have been successfully recorded by using X-ray CCD camera. It is demonstrated experimentally that the measured wavelength is accorded with the theoretical value. At the same time, experimental result shows that spectral resolution of PET and KAP crystals is 956 and 1123.

The problem of task allocation for multiple robots is to allocate more relative-tasks to less relative-robots so as to minimize the processing time of these tasks. In order to get optimal multi-robot task allocation scheme, a twodimensional artificial swarm algorithm based approach is proposed in this paper. In this approach, the normal artificial fish is extended to be two dimension artificial fish. In the two dimension artificial fish, each vector of primary artificial fish is extended to be an m-dimensional vector. Thus, each vector can express a group of tasks. By redefining the distance between artificial fish and the center of artificial fish, the behavior of two dimension fish is designed and the task allocation algorithm based on two dimension artificial swarm algorithm is put forward. At last, the proposed algorithm is applied to the problem of multi-robot task allocation and comparer with GA and SA based algorithm is done. Simulation and compare result shows the proposed algorithm is effective.

Using double loop DC motor drive system (DLM) of the RoboCup middle size robots as research subject, the model of DLM has been reduced to a simple state space one by the "quasi-equivalent" modeling method, based on the characteristic analysis of the system. Further, the parameters of the model can be exactly identified applying the improved genetic algorithm. The results of compared experiment proved that using this model and identification ways can get a reasonably structural and high parameters precision model. The model can describe the DLM to design the control system for robot soccer.

In many cases exploring robots work under harsh conditions such as presence of various kinds of obstacles, high temperature, and limited visibility, etc. So they are required to have tough mechanical structure, good sensing capability, and reliable communication capability for receiving operator's command and for sending information on the state of the robot and on environment to operator. In this study the methodology of implementing these capabilities on exploring robot is explained, and actually an exploring robot that can investigate hazardous or disaster sites is designed and implemented. Finally, the performance of the proposed exploring robot is tested under simulated disaster environment.

Line tracer robots are widely accepted in various industry for the purpose of moving objects. In this study feasibility of using line tracer robots under disaster situation is studied as the line tracing capability may be useful in case of disaster. A small scale line tracer robot is designed, implemented, and tested under disaster situation. Specifically, the MCU based robot hardware including sensor, encoder, motor, and power supply is designed and implemented. Also, the software for receiving sensor signal, for driving motor and for implementing line searching algorithm is designed. The devised robot detects the guide line using 6 infrared optical sensors and can adjust the direction of motion with the front steering servo motor according to the sensor signal. In driving test, the robot showed reliable performance both on the straight and curved paths having obstacles under limited visibility condition as in disaster situation.

HAJIME ROBOT 18 is a fully autonomous biped robot. It has been developed for RoboCup which is a worldwide soccer competition of robots. It is necessary for a robot to have high mobility to play soccer. High speed walking and all directional walking are important to approach and to locate in front of a ball. HAJIME ROBOT achieved these walking. This paper describes walking control of a small size humanoid robot 'HAJIME ROBOT 18' and shows the measurement result of ZMP (Zero Moment Point). HAJIME ROBOT won the Robotics Society of Japan Award in RoboCup 2005 and in RoboCup 2006 Japan Open.

We made a self controlled moving robot (SCMR) which tracks the object using the images obtained from two cameras which are installed on the front of the system. In order to reduce the effect of a color and a light we converted the true color images to the gray color images. We can estimate the object movements by utilizing four images which are obtained from two cameras of different times. The SCMR controls the motor to move by itself considering the result of the image processing. In the image processing we used the image zone partition method. The image zone partition method decides a direction that the object is moving. For fast calculation of the image processing and the motor control we installed the personal computer (PC) on the SCMR.

Robot vision technology is needed for the stable walking, object recognition and the movement to the target spot. By some sensors which use infrared rays and ultrasonic, robot can overcome the urgent state or dangerous time. But stereo vision of three dimensional space would make robot have powerful artificial intelligence. In this paper we consider about the stereo vision for stable and correct movement of a biped robot. When a robot confront with an inclination plane or steps, particular algorithms are needed to go on without failure. This study developed the recognition algorithm of distance and gradient of environment by stereo matching process.

In this paper, we propose a design of Service-Adaptive Engine for robot middleware. This middleware called the KOMoR (Korea Object-oriented Middleware of Robot) is a middleware for robot that composed of three layers (Service Layer, Network Adaptation Layer, Network Interface Layer). In particular, Service-Adaptive Engine in Service Layer is responsible for communication between distributed applications and provides a set of features that support development of realistic distributed applications for a robot. Also, it avoids unnecessary complexity, making the middleware easy to learn and to use. For writing application, both client and server consist of a mixture of application code, library code, and code generated from IDL definition called MIDL (Module Interface Definition Language). The Service-Adaptive Engine in SL contains the client-and server-side run-time support for remote communication. The generic part of the Service-Adaptive Engine (that is, the part that is independent of the specific types you have defined in MIDL) is accessed through the SL API. The proxy code is generated from MIDL definitions and, therefore specific to the types of objects and data you have defined in MIDL.

This study shows how a mobile service robot can avoid obstacles, and presents a VPH method using feature information for Local Path Planning. It is not easy to make a mobile service robot automatically move towards the goal. Path Planning lays out the path through which a robot follows to reach the goal. It can be divided into two folds: Global Path Planning (GPP) and Local Path Planning (LPP). Local path planning sets a path in a changing environment with moving obstacles such as in a museum and exhibition hall so that the robot reaches the goal without any collision. This study evaluates the Fusion Map-VPH (FM-VPH) Local path planning method with improved VPH by making use of the combined data drawn up through the ultrasonic sensor and laser sensor and by means of feature information. The results of the simulations and experiments have verified the validity of the methods described.

Human emotions could be expressed by many bio-symbols. Speech and facial expression are two of them. They are both regarded as emotional information which is playing an important role in human-computer interaction. Based on our previous studies on emotion recognition, an audiovisual emotion recognition system is developed and represented in this paper. The system is designed for real-time practice, and is guaranteed by some integrated modules. These modules include speech enhancement for eliminating noises, rapid face detection for locating face from background image, example based shape learning for facial feature alignment, and optical flow based tracking algorithm for facial feature tracking. It is known that irrelevant features and high dimensionality of the data can hurt the performance of classifier. Rough set-based feature selection is a good method for dimension reduction. So 13 speech features out of 37 ones and 10 facial features out of 33 ones are selected to represent emotional information, and 52 audiovisual features are selected due to the synchronization when speech and video fused together. The experiment results have demonstrated that this system performs well in real-time practice and has high recognition rate. Our results also show that the work in multimodules fused recognition will become the trend of emotion recognition in the future.

This paper provides a new approach of manipulability for robotic manipulators. While conventional manipulability is analysis in velocity domain and can not include force effect such as gravity of link and payload, the proposing approach can include the force effect. For the purpose, an operation range is introduced. The operation range is for actuator attached with every joint of robot and provides generable joint torque and velocity and their relation (between generating torque/velocity and addable velocity/torque). Using the operation range, we develop a new manipulability in velocity domain, including force effect.

In this paper, the concept of a walking guide system with tactile display is introduced, and experiments of 3-D obstacle detection and tactile perception are carried out and analyzed. The algorithm of 3-D obstacle detection and the method of mapping the generated obstacle map and the tactile display device for the walking guide system are proposed. The experiment of the 3-D detection for the obstacle position using ultrasonic sensors is performed and estimated. Some design guidelines for a tactile display device that can display obstacle distribution is discussed.

This paper proposes a human body posture estimation method based on back projection of human silhouette images extracted from multi-camera images. To achieve real-time 3D human body posture estimation, a server-client system is introduced into the multi-camera system, improvements of the background subtraction and back projection are investigated. To evaluate the feasibility of the proposed method, 3D estimation experiments of human body posture are carried out. The experimental system with six CCD cameras is composed and the experimental results confirm both the feasibility and effectiveness of the proposed system in the 3D human body posture estimation in real-time. By using the 3D reconstruction of human body posture, the simple walk-through application of virtual reality system is demonstrated.

A new structure of dual arm robot manipulator which consists of two industrial 6-DOF arms and one 2-DOF Torso is introduced. Each industrial 6-DOF arm is able to be used as a stand-alone industrial 6-DOF robot manipulator and as a part of dual arm manipulator at the same time. These structures help the robot maker which is willing to succeed in the emerging dual arm robot market in order to have high competition for the current industrial robot market at same time. Self-collision detection algorithm for multi-arm robot and kinematics algorithms for the developed dual arm robot manipulator which are implemented in our controller are introduced.

This paper describes the development of a hand rehabilitation system for stroke patients. Our aim is to provide fine motion exercise for a hand and fingers. Thus, a hand rehabilitation device that assists patients' finger movements was developed. Because this device has 18 degrees of freedom of motion, it is diffcult for disabled patients to use it by themselves. Therefore, an appropriate control strategy and control system are required to allow its safe and effective use. In light of this requirement, a control system was constructed which is comprised of four separated controllers. This paper presents the structure of the control system and introduces the control protocols used in our hand rehabilitation system.

The traditional artificial potential field (APF) method is widely used for motion planning of traditional mobile robot, but there is little research about the application to the omnidirectional mobile robot (OMR). To propose a more suitable motion planning for OMR, an evolutional APF is presented in this paper, by introducing the revolving factor into the APF. The revolving factor synthesizes the anisotropy of OMR and the affect of dynamic environment. Finally simulation is carried out to demonstrate that, the evolutional APF is a high-speed and high-efficiency motion planning by comparing with the traditional APF, and the advantages of OMR is exerted.

In this paper, a modified version of dynamic encoding algorithm for searches (DEAS) is proposed and applied to generate walking patterns of a biped humanoid robot. For the controller of each joint motor to generate optimal trajectories, mDEAS is developed from the previous versions of exhaustive DEAS (eDEAS) and univariate DEAS (uDEAS). Modular DEAS (mDEAS) searches optimal coefficients of polynomials whose trajectories are assigned to joint motors. Since the number of the coefficients amounts up to 16, sharing search space and optimizing independently is expected to search efficiently. For validation of mDEAS, a simulation result about initial stepping is provided.

This paper addresses an analog input shaper introduced in haptic control to improve the stability when interacting with virtual environments. High frequency inputs to a haptic device, which can occur in collision with a virtual wall with high stiffness, can bring limit cycle oscillations and instabilities. In order to reduce the high frequency input to an haptic device an analog input shaper is added to the control system. Since the input shaper acts as a low-pass filter, when a haptic pointer leaves the virtual wall with high stiffness, a user may feel slow decrease of impedance, moreover there may be negative impedance as if the wall is pulling. In order to prevent this, we add half-wave rectifiers which allow fast decrease of impedance and no negative input to a haptic device. The input shaper reduces the total energy supplied to a haptic device by preventing inputs with high frequency from flowing into a haptic device. Therefore it can be regarded as an artificial damping element. In order to apply the Energy-Bounding Algorithm (EBA),¹ which can guarantee the stable haptic interaction, to a scaled teleoperation in a virtual nano-environment two scaling factors (velocity and force) are incorporated into EBA. By applying the analog input shaper to EBA in scaled teleoperation, the range of virtual wall stiffness which can be stably rendered by a haptic device is significantly extended.

The coordinated motion of group of autonomous mobile robots for the achievement of goal has been of high interest since the last decade. Previous research works have revealed that one of the essential problems in the area is to plan, navigate and coordinate the motion of robots, avoiding obstacles as well as each other while still achieving the goal. In this paper, Behavior Based approach for the control of distributed networked robotic system, concentrated towards the navigation, planning and coordination between them in unknown complex environment is addressed. A layered behavior based control architecture, with the basic behaviors of Message passing, Obstacle avoidance, Safe wandering and Pit sensing have been designed and assigned to the individual robotic systems to form a navigation algorithm. Validation of this guidance algorithm is carried out through simulations using SIMULINK/State flow.

This paper presents the position accuracy and sensitivity for the robot localization system using the ultrasonic wave. This localization system consists of many beacons and a receiver. Many beacons are located in the absolute coordinate defined by user and a receiver is loaded in the robot. The robot position can be computed from the known beacon positions and more than three distance information between the receiver and beacons measured by ultrasonic wave. But the distance measured by ultrasonic wave includes errors that are limited by ultrasonic sensor efficiency. We found that ±2cm tolerance for measured distance in Threshold method is independent with the distance range. Since more than 3 beacons are used, Least-Square-Method are used to compute the receiver position. This paper derives the method of position accuracy and sensitivity due to distance error by using Taylor Expansion and SVD(Singular Value Decomposition) method.

This paper presents a disaster prevention system using aerial images obtained by a small unmanned aerial vehicle (UAV). The immediate assessment of damage information and the continuous collection of information for constructing a revival plan are important for damage mitigation when natural disasters such as large earthquakes occur. In this study, we develop a small prototype UAV and onboard software using a GPS navigation system and propose a unique disaster prevention system that employs the small UAV and a geographical information system (GIS). We divide the period that follows a natural disaster on the basis of three processes, and we propose the use of the small UAV for gathering and integrating information into the GIS. We conducted field experiments to evaluate the proposed system employing the small UAV. From the results of our experiments, we concluded that the system is effective in disaster prevention.

Recently various types of robots are being studied and developed, which can be classified into two groups: humanoid type and animal types. Since each group has its own merits and demerits, a new type of robot is expected to emerge with greater strengths and fewer weaknesses. In this paper we propose a new type of robot called the "Centaur Robot" by merging the concepts of these two types of robots. This robot has a human-like upper body and a four-legged animal-like lower body. Due to this basic architecture, the robot has several merits, including human-like behaviors. It can also walk smoothly and stably even on non-smooth ground. We describe the basic concept of our centaur robot and then explain its hardware and software architectures as well as its merits.

Dance is one form of entertainment where physical movement is the key factor. The main reason why robots are experiencing a kind of "boom" is that they have a physical body. We propose a robot dance system that combines these two elements. First, various factors concerning entertainment and dance are studied. Then we propose a robot dance system in which a humanoid robot achieves various dance performances by concatenating different short dance motions. Also we describe the details of the system by focusing on its software functions and show evaluation results of robot dance performances.

To design human cooperative robot it is necessary to take into consideration the factor to make the robot move as smooth as possible during the cooperative task. This is to ensure that human can work with robot with high degree of smoothness that would ensure the task is completed without stress and fatigue to human. Since robot does not know the feeling of human, we need to replicate the human motion characteristic into the robot. In human-human cooperative task normally to achieve good cooperative task, human will use auditory, visual and touch senses. We want to understand what kind of sense at which moment it is uses to get good cooperative task. We arranged the experiment subjects so that they utilized their senses individually during the cooperative task. Experiment devices are equipped with 3D position sensors and force sensors to measure the position, angle and force value. This research is concentrating the force and torque characteristic that occurs to the human participants during human-to-human cooperative work where the human audio, visual and touch senses are applied.

Side-wall is the static obstacle in soccer robot game, reasonably making use of the Side-wall can improve soccer robot competitive ability. As a kind of artificial life, the Side-wall processing strategy of soccer robot is influenced by many factors, such as game state, field region, attacking and defending situation and so on, each factor also has different influence degree, so, the Side-wall behavior selection is an intelligent selecting process. From the view point of human simulated, based on the idea of Side-wall processing priority[1], this paper builds the priority function for Side-wall processing, constructs the action predicative model for Side-wall obstacle, puts forward the Side-wall processing strategy, and forms the Side-wall behavior selection mechanism. Through the contrasting experiment between the strategy applied and none, proves that this strategy can improve the soccer robot capacity, it is feasible and effective, and has positive meaning for soccer robot stepped study.

Human characteristics are supposed to be applied to the control systems of the human-friendly robots. Therefore, it is important to know the human characteristics in human-robot cooperative tasks. In this study, we considered a single rotational degree of freedom experimental system as well as a mass-spring-damper-friction dynamic model for the human arm and analyzed the master-slave and the master-semi master cooperative motions. Through the master-slave cooperative motion, we verified the application of the mass-spring-friction dynamic model of the slave's arm, and regarding its parameter we found the damping factor as zero and the stiffness as well as the coulomb friction torque almost individually invariable. Through the master-semi master cooperative motion, we got the torque characteristics and drew a comparison of them with those in the master-slave cooperative motion, and then we found improved torque characteristics, the constant torque characteristics.

The kinematics of two typical mechanisms is analyzed following the concept design of a 4MNm forging manipulator for a huge hydraulic press, i.e. a parallel links mechanism and a swing links mechanism. For the first kind of mechanism, when the jaw swings up and down, the swing angle is in proportion to the extending length of its driving hydraulic cylinder. For the second kind of mechanism, it is mostly difficult for the fluid driving control system to make sure three hydraulic pressure cylinders to keep moving synchronously. Based on these simulations and considering practical technological demands, the first kind of mechanism with parallel links is chosen as a better one.

This paper is mainly about a new robot control system, which has open structures, and whose architecture is based on CAN bus for a joint modular robot. The hardware system of the joint control module is designed for signal sampling and processing, data storage and communication. ATmega64 SCM is selected as the core processor. A CAN card is used for communication. FM256 RAM is used for data storage and condensation. The monitoring software and lower level control processing software are developed with ZLGVCI interface functions of CAN bus, and in the ICC AVR IDE and AVR Studio online debug tool with C language. Some experiments are carried out to validate the basic functions of the control system.

In determining the quality of a car, switch feeling is one of the important factors. Designing switches is time consuming because the relationships between characteristics of a desired switch and design parameters such as spring stiffness or stroke are not very clear, and furthermore, physical evaluation requires prototypes. A haptic Virtual Prototyping Environment (VPE), in which CAE (Computer-Aided Engineering) and haptic interface work together to display virtual feelings of mechanisms designed by CAD (Computer-Aided Design) will enable designers to feel the motions of the designed switches without prototypes. This paper presents design and quantitative evaluation of a haptic VPE for power window switches. We compared relationship between displacement and static reaction force, and acceleration during clicking motion. The similarity in the force vs stroke curve (FS curve) suggests the overall similarity in the feeling of the whole switch stroke. The acceleration showed some differences, which was felt as somewhat softer clicking effect. The lower frequency of the residual vibration of the haptic device indicates that the equivalent inertia of the haptic device is larger than that of the real switch. Although the overall motion of the switch stroke felt similar, quality of reality needs further refinement. Introduction of acceleration feedback for reducing the equivalent inertia is discussed.

Recently, the scientific investigations of marine environment, marine products and algae fields often require the use of underwater vehicle to perform survey and monitoring tasks. This paper describes a compact unthethered remotely operated vehicle (UROV). The vehicle is compact and light in weight by employing a VARIVEC propeller. The UROV operator controls by the UROV monitor including the data of GPS, electronic compass, sonar, depth and range sensor. The sensor data are extracted smartly by the Programmable System on Chip (PSoC) microprocessor. The performance of UROV is examined about the operating characteristics, image transmission, and remotely monitoring system. The temperature, conductivity, and turbidity of the sea surface are collected with accuracy.

The purpose of this paper is to show merit and feasibility of the component based approach in robot system integration. Many methodologies such as 'component based approach, 'middle ware based approach' are suggested to integrate various complex functions on robot system efficiently. However, these methodologies are not used to robot function development broadly, because these 'Top-down' methodologies are modeled and researched in software engineering field, which are different from robot function researches, so that cannot be trusted by function developers. Developers' the main concern of these methodologies is the performance decreasing, which origins from overhead of a framework. This paper overcomes this misunderstanding by showing time performance increasing, when an experiment uses 'Self Healing, Adaptive and Growing softwarE (SHAGE)' framework, one of the component based framework. As an example of real robot function, visual object recognition is chosen to experiment.

This paper proposes an improved octree representation method which is necessary for robot manipulation. To manipulate an object in a certain complex environment, robot needs simple, precise, and task oriented environment representations. Octree is a effective and broadly used three dimensional volumetric representation method to do this. But the problem of the conventional octree representation methods is that it generate many useless and incorrect cells because it generates the octree based on the number of 3D points. To solve these problems, we use the new 'Fill-Factor' concept. This paper introduces the concept and shows the experimental results to show the feasibility of the concept.

Some related works about the control architecture of robot system are briefly summarized. According to the discussions above, this paper proposes control architecture of service robot based on artificial psychology. In this control architecture, the robot can obtain the cognition of environment through sensors, and then be handled with intelligent model, affective model and learning model, and finally express the reaction to the outside stimulation through its behavior. For better understanding the architecture, hierarchical structure is also discussed. The control system of robot can be divided into five layers, namely physical layer, drives layer, information-processing and behavior-programming layer, application layer and system inspection and control layer. This paper shows how to achieve system integration from hardware modules, software interface and fault diagnosis. Embedded system GENE-8310 is selected as the PC platform of robot APROS-I, and its primary memory media is CF card. The arms and body of the robot are constituted by 13 motors and some connecting fittings. Besides, the robot has a robot head with emotional facial expression, and the head has 13 DOFs. The emotional and intelligent model is one of the most important parts in human-machine interaction. In order to better simulate human emotion, an emotional interaction model for robot is proposed according to the theory of need levels of Maslom and mood information of Siminov. This architecture has already been used in our intelligent service robot.

It's of great importance for assisting robot in path planning, position navigating and task performing by perceiving and recognising environment characteristic. To solve the problem of monocular-vision-oriented landmark recognition for mobile intelligent robot marching with complex background, a kind of nested region growing algorithm which fused with transcendental color information and based on current maximum convergence center is proposed, allowing invariance localization to changes in position, scale, rotation, jitters and weather conditions. Firstly, a novel experiment threshold based on RGB vision model is used for the first image segmentation, which allowing some objects and partial scenes with similar color to landmarks also are detected with landmarks together. Secondly, with current maximum convergence center on segmented image as each growing seed point, the above region growing algorithm accordingly starts to establish several Regions of Interest (ROI) orderly. According to shape characteristics, a quick and effectual contour analysis based on primitive element is applied in deciding whether current ROI could be reserved or deleted after each region growing, then each ROI is judged initially and positioned. When the position information as feedback is conveyed to the gray image, the whole landmarks are extracted accurately with the second segmentation on the local image that exclusive to landmark area. Finally, landmarks are recognised by Hopfield neural network. Results issued from experiments on a great number of images with both campus and urban district as background show the effectiveness of the proposed algorithm.

In this paper, we consider control of grasp and manipulation of an object in 3-dimensional space by a 3-fingered hand robot with soft finger-tips. We firstly propose a 3-dimensional deformation model of a semispherical soft fingertip and confirm its relevance by experiments. Second, we derive the rolling contact kinematics and the dynamical equations of the fingers and the object where the 3-dimensional deformation is considered. For the system, we thirdly propose a method to regulate the object and the internal force with the information of the hand, the object and the deformation. Simulations are shown to prove the effectiveness of the control method.

We propose a voice-controlled electric wheelchair with a system for detecting its position and direction. Thus far, we have studied systems that incorporated voice instructions with our indoor navigation equipment using RF tags. The operability of a wheelchair was measured in the same environment for different operating methods. We found that one very effective method is to store the possible stopping positions and angles of rotation for the desired directions in RF cards, which could then adequately assist the movement of a wheelchair.

In parallel mechanisms with redundant actuators, actuator forces are underspecified because of redundancy. In this paper, the method to minimize a maximum actuator force was proposed. At first, parallel mechanism with two redundant actuators and Stewart platform model using the linear shaft motor were developed. This redundant parallel mechanism was consists of Stewart platform and additional 2 redundant actuators. Inverse dynamics of each mechanisms were formulated. Secondly, effectiveness of dynamic control of Stewart platform was experimentally confirmed. Finally, the method to minimize a maximum actuator force using Linear Matrix Inequality (LMI) was proposed. And the effectiveness of the proposed method was confirmed by simulation.

In this paper, control of a two-link flexible manipulator using neural networks is presented. The neural networks are trained so as to make the error between the root strain and the desired root strain and the error between the joint angles and the desired joint angles almost zero. In the process, the neural networks learn the inverse dynamics of the system. Also we compared the learning algorithms between conventional back propagation method and adaptive learning method. The numerical results and experimental results show that this NN control system can suppress the vibrations of the flexible manipulator within a short time in comparison with no feedback. Numerical and experimental results for the vibration control of a two-link flexible manipulator are presented and verify that the proposed NN control system is effective at controlling flexible dynamical systems.

This paper describes quick and stable standing and sitting motion of an inverted pendulum type robot by wholebody motion including force control. The whole-body motion is achieved by controlling composite center of gravity of the robot, and damping force control of a wrist-roller is employed to assure the stability of inverted pendulum control on a contact situation with the ground. Motion planning of standing and sitting based on the proposed control method is also proposed. Standing and sitting motion using the proposed motion control and motion planning is simulated with a dynamic simulator, ODE(Open Dynamics Engine). From the simulation results, it was confirmed that the robot successively realized the standing and sitting motion quickly and stably.

Based on Landau-Devonshire Shape Memory Alloy SMA model and Nitzsche-Breitbach-Elmar SMA model, the bifurcation characteristic of rotor-shape memory alloy bearings(SMAB) system was investigated in this paper. Heteronomous system was transformed into autonomous system in averaging method and Van der Pol transformation, and the existence of Hopf bifurcation was proved in theory. The concept of broadened set of equilibrium point was introduced to improve centre manifold method to be adapted to heteronomous system. The equation of the flow on the centre manifold of rotor-SMAB system was obtained, and the existence of transcritical bifurcation and supercritical pitchfork bifurcation was proved in theory. Finally the result in centre manifold method was compared with the one in averaging method. The comparison shows that the results of the two methods were both the parts of global dynamic characteristic of rotor-SMAB system, while centre manifold method can be applied to research bifurcation behavior in the case of more codimensions. It means that the two methods both have limitation, and global dynamic characteristic must be obtained in kinds of method.