Several Control Algorithms Of Optical Testing Simulator

Three-axis optical testing simulator is a key testing device in HIL experiment of CCD imaging guidance system, position accuracy, velocity accuracy and performance in low velocity are important indexes of the simulator. The factors influencing accuracy (position and velocity) and performance in low velocity of the simulator are friction torque, coupling among axis, power frequency, noise, mechanical resonance, etc. For the purpose of improving accuracy and performance in low velocity of the simulator without hardware costs, the dissertation study servo control, friction compensation, decoupling and filtering in the control of the simulator, and corresponding solutions are given.Control of the simulator is essentially control of DC servo motors, chapter 2 focuses on control of DC servo motors. Three cascaded loops (current loop, velocity loop and position loop) are adopted to control DC motors, for improving response time, fuzzy control is introduced. PI controller is used in current loop, FuzzyPID switching controller is used in velocity loop, in which PID control is used in range of small error while fuzzy control is used in range of large error. Feedforward+FuzzyPID switching control scheme is used in position loop. Simulation shows that the combination of traditional PID control and modern fuzzy control improves response time of the simulator.Friction is a main factor influencing performance in low velocity of the simulator, in chapter 3, a composite adaptive friction compensation algorithm is designed. A adaptive friction compensation algorithm is calculated based on Lyapunov theorem, the other adaptive friction compensation algorithm is calculated from the view point of pattern recognition using least square method, then the two algorithms is organized linearly to construct a composite adaptive friction compensation algorithm. Both theoretical analysis and simulation show the composite adaptive friction algorithm have characteristics of both asymptotically convergence of adaptive control and online update of least square method. The composite adaptive friction compensation algorithm proposed in the dissertation is effective and practical.The coupling among the three axis of simulator deteriorates performance of the simulator, so decoupling is indispensable. In chapter 4, Singh decoupling algorithm based on invert system is studied. Firstly the inverse of the coupling physical system is calculated by using Singh algorithm, and then the invert system is connected with the physical system anteriorly, the coupling physical system and its invert system construct a pseudo liner system, and thus a dynamic decoupling design is implemented. The Singh decoupling algorithm based invert system avoids singularity of Falb-Wolovich matrix. An example is given. Simulation shows nonlinear coupling can be compensated completely if the mathematic coupling model is accuracy enough.Power frequency interference, mechanical resonance, measure noise, Gaussian noise, high frequency noise influence on accuracy of the simulator directly or indirectly. Chapter 5 discusses filtering techniques to eliminate the noise or disturbance. 50 Hz adaptive notch filter and mechanical resonant adaptive notch filter are designed to remove power frequency interference and mechanical resonance respectively. Kalman filter is utilized to remove measure noise and process noise, furthermore, acceleration feedback control is discussed based on Kalman filtering. Besides, the application of other filtering technologies such as low pass filtering, etc, are simply discussed.The dissertation study several key algorithms on control of simulator. Theory analysis and simulation result are both available in the dissertation.