The posture and size of workpiece in pixel level can be determined through optimization of the internal parameters of the camera, utilizing the enclosing rectangle method based on rotation axis to position the workpiece. After selecting ROI of the object, the sub-pixel edge of workpiece is extracted using the bilinear interpolation algorithm and Hessian paradigm line fitting is utilized to find the object edge accurately. The industrial camera with 5 million pixels is used, and the sizes from 80mm to 150mm of the workpiece are measured under the condition of panorama shooting, with measurement repeatability reaching 0.015mm.
Local mean decomposition (LMD) is a time-frequency analysis approach to deal with complex multi-frequency signal. However, as the decomposition process is sensitive to noise, there is a distinct limit when it is applied to analysis of the vibration signals of machinery with serious background noise. An improved LMD algorithm based on extracting the extrema of envelope curve is put forward to reduce the influence of high-frequency noise effectively. To verify its effect, three different de-noising methods, i.e., band-pass filter method, wavelet method and lift wavelet method are used, respectively. And the comparison result of the 4 methods shows that the proposed method has satisfactory reproducibility. Then the new algorithm is applied to real bearing signal, and experimental results show that it is effective and reliable. The method also has certain significance for the subsequent eigenvector research in intelligent fault diagnosis.
To solve the problem of feature extraction of weak gear fault under strong noise background, an early feature extraction method based on cascaded monostable stochastic resonance (CMSR) system and empirical mode decomposition (EMD) with teager energy operator demodulation was proposed. The model of monostable stochastic resonance expanded the processing range of characteristic frequency of the measured signal, and had a good effect on denoising performance by cascading. Firstly CMSR was employed as the preprocessor to remove noise, then the denoised signal was decomposed into a series of intrinsic mode functions (IMFs) of different scales by EMD, and finally teager energy operator demodulation was applied to obtain amplitudes and frequencies of each effective IMF to extract the weak gear fault feature. Simulation and application results showed that the proposed method could effectively detect the characteristic frequency of gear fault of local damage after the noise reduction by CMSR.
This paper concerns the mathematical model and simulation of forces on Mie micro-particles in laser trap. To make the
mathematical models suitable for particles with arbitrary shape, the surface of Mie particle is divided into many microplanes,
then the radiation pressure resulting from the interaction between a single incident light and a micro-plane and a
individual ray tracing is obtained using the geometrical optical theory. A computer simulation program based on this
mathematical model is designed to calculate the trapping force on Mie micro-particles with arbitrary shape (sphere, cone,
polyhedron and particle with smooth surface).The simulation result for the trapping forces of spherical particles is
consistent with that in reference , and comparisons are made between spherical particles and spherical particles
reconstructed from NURBS, and between conical particles and conical particles reconstructed from polyhedron. These
results verify the simulation program, so it can be an effective design and analysis tool for the application of laser
trapping in nanometer testing and manufacturing.
A 2D high-precision motion unit plays an important role in an automatic optical inspection (AOI) system, and its
positioning accuracy and speed have an decisive effect on the inspection accuracy, efficiency and stability of the AOI
system, but the present AOI systems cannot ensure the high speed and high-precision at the same time. To solve this
problem, an automatic optical platform is developed, using the whole-closed-loop, fast, high-precision positioning
technology to construct the mathematical models of its motion control unit, including models of current loop, velocity
loop and position loop from inside to outside. Experiments were made on the motion platform using a laser
interferometer. Experimental results show that the 2D motion platform can be used to reach a scanning speed of 20m/s
and a positioning accuracy of 9ìm in a work area of 300mm * 3000mm with a load of 5kg.
This paper analyzes the effects of birefringences of acousto-optical crystals on the diffraction angle of the o and e beam,
deduces the exact tuning relationship between the two diffracted polarized beams in large-aperture and non-collinear
AOTFs, and explores the influence of birefringences on the wavelengths difference of the two diffracted beams, Δλ,
acoustical angle θa, and relation between the external angle of incidence θi and the ultrasonic frequency f. Theoretical
calculations show, that the difference of the internal angles of incidence of the two diffracted beams, θie-θio, varies in
the range of 0°~1.5° in accordance with the change of the external angle of incidence, and that the maximum
wavelengths difference of the diffracted beams of order±1 can be as much as 80 nm. The direction θa of the acoustical
vector is approximately 108.8°, which satisfies the conditions of momentum match and wavelengths difference to be
zero for diffracted beams of order±1, but the internal angle of incidence varies about θie=55.6°, and the range can reach
1.4°. The obtained theoretical results are verified by experiments to measure the diffracted beams of order ±1 of