Proc. SPIE. 9682, 8th International Symposium on Advanced Optical Manufacturing and Testing Technologies: Large Mirrors and Telescopes
KEYWORDS: Mathematical modeling, Telescopes, Digital signal processing, Detection and tracking algorithms, Image segmentation, Field programmable gate arrays, Control systems, Antennas, Charge-coupled devices, Control systems design
The angle between slant and azimuth axes is 45 degrees in slant-mount telescopes, which use the slant rotation to replace altitude movement in traditional mounts. The circular motion can achieve more smooth driving in altitude direction, and can effectively overcome problems in traditional mounts such as vibrations around the zenith. It is a hot topic in researches of current astronomical instruments. While the vertical displacement of the structure is composed by circle rotation of the slant axis, which causes displacements of the two direction coupled. Moreover, the slant rotation can also leads to the revolving movement of CCD images of the tracking system. These two problems bring great challenges to the slant mount driving. In this paper, based on a slant mount, we establish accurate mathematical model of the drive system to effectively compensate for the revolving movement of the image field, decompose in real time the deviation of vertical and horizontal directions by the decoupling algorithm, and design the control system using FPGA to carry out testing experiments. We strive to solve the displacement coupling and the image revolving movement problem, and thus provide a feasible technical support for driving control of slant mounts.
Large astronomical optical telescopes are badly needed in order to learn more remote universe. There exist some key
problems of the control systems of large astronomical optical telescopes. Since they have voluminous bodies that would
encounter heavy external disturbance, one of the key problems is focused on how to accurately control them.
Additionally, in order to get nicer ultra-low velocity performance and a steady field of view, friction drive is widely
applied in contemporary large optical telescopes. One serious disadvantage of friction drive is that it will cause some
nonlinear uncertainties to influence telescope controls because of the mechanical characteristics between the principal
and subordinate friction wheels. These two aspects of external and internal disturbances will make a telescope very
difficult to be controlled. In this paper, we introduce a method of higher order sliding modes (HOSM) to control
telescopes, which overcome these two disadvantages of traditional Proportional-Integral-Derivative approach and can
achieve excellent control performance. Conventional sliding mode approach has been applied in many other mechanical
control systems owing to its high accuracy in anti-jamming. By discontinuous switching, it is invariable to disturbances
based on keeping some constraints with a sufficiently energetic effort. However, such conventional sliding mode
approach may cause dangerous high-frequency vibrations in the corresponding control system, which may influence
systemic control performance or even lead the system unstable. In this work, we use the newly developed HOSM
approach in the control systems of the large astronomical optical telescopes. The HOSM approach inherits the dominant
merits of conventional sliding mode. Moreover, it acts on the higher order time derivatives of the system deviation from
the constraint. And the discontinuous dynamics are restricted to the highest state while the counterpart in standard sliding
mode is in first derivative. Thus the HOSM approach can mostly removes high-frequency vibration effects on telescope
control. This control approach needs all states of the system to be observable. We use robust exact differentiator to
estimate the immeasurable state. Simulations have been done in the environment of MATLAB language. The results
show that this approach can realize the tracking performance of accurate ultra-low velocity for telescope control.
The workshop test of mount drive for Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) was completed in June of 2005. Now the giant mount has just been erected on Xinglong station, and is due to test in the summer of 2006. LAMOST mount mechanism features friction drive on both axes, and oil pad is employed specifically for the azimuth. For further improving the tracking accuracy in worse surroundings some nonlinear phenomena in the drive chain have to be addressed. Moreover, external uncertainties on Xinglong site, wind buffeting in particular, could affect load variation on the drive. The control system parameters would change with time, thus eventually degrade the tracking performance. All these reasonable assumptions call for a more robust controller than conventional PID approach to cope with. This is where H-Infinity controller comes in. This paper focuses on the mount drive of LAMOST by using H-Infinity technique and comparison with the PID servo. The load disturbance rejection is discussed, as well as transmission rigidity improvement is analyzed. Study and simulation are done in Matlab. The model test in our friction drive lab is presented.
The concept for Chinese Future Giant Telescope (CFGT) with 30-m aperture has been around for several years, although
the requirements for control system are still far from completed and conclusive at this stage. Since the project was
proposed more study on a number of key issues relevant to the control system has been conducted. In particular the
mount control system for the giant telescope has been put forward under exploration. With our ongoing 4-m LAMOST
telescope just underwent a successful mount drive test the LAMOST control group has become more knowledgeable
with hands on experience that would be quite useful for mount drive design of even large telescope. This paper focuses
on the mount control system design for CFGT telescope in general. Particular aspects such as the effect of large moment
of inertia with ultra low-speed and multi-disturbance are included. Friction drive is opted for both historical and
economical reasons. Drive stiffness and servo control parameters optimization are discussed based on the workshop test
with LAMOST mount that could possibly be mapped to CFGT.