| An Electro-Optical Search and Track system (EOST) is the device to search and track the target which is detected by the infrared or CCD sensors. The EOST system can be used in many applications such as real time surveillance, homing missile guidance, laser designation systems and weapon fire control systems etc. As the targets are developed toward more maneuverable, the tracking servo system is desired for performance improvements of fast response and precise positioning. Based on the author's practice experience on developing several EOST systems, this thesis carry on the research work on several aspects such as kinematics and dynamics modeling of tracking servo system, simulation of tracking control in whole time of tracking a target, servo system bandwidth optimization, multi-mode switching control of tracking control loop, matching between image tracker and servo control, and the evaluation of tracking performance.1. Firstly, this thesis analyzes the primary principles of the EOST system, derives kinematics equations and Newtonian mechanics equations of a two axes stabilized platform, and establishes the system components'mathematical models. To analyze the effects of error factors on the tracking system's performance in systemic view, the working process of the system is modeled and simulated, which provides references for the confirmation of its performance indexes and the research of control problems.2. To fulfill the demand of quick response performance, the thesis investigates on the optimization design of servo system bandwidth. After analysis on the effect of sensor noise, system damp and resonance on the system bandwidth, this thesis proposed the Bode-Step method to expand the bandwidth of the tracking servo system, and established a systemic method for bandwidth design and optimization. It is an efficient means for improvement of servo capability.3. In EOST systems, there are special requirements for better dynamic and steady-state performance, which cannot be satisfied by using only single controller. In order to respond to the changing requirements for better control performance at different stages, several different and independent controllers can be used by switching. This control method is normally called multimode control. The direct switching of the parallel channels might result in an unstable system, or in a system with small stability margins and, therefore, producing large transient error. Several methods were proposed for mode switching from one controller to another. After comparing several methods such as initial value compensation (IVC), composed nonlinear feedback (CNF), an adaptive regulator with a smoothly variable high-order linear response controlled by a single real parameter is employed in an acquisition-tracking system. Using normalized design simplifies the design procedure for practice application. The experiment result shows that the transition between the modes is smooth and rapid, and the transient responses are fast and without substantial overshoots.4. A method for servo and image tracker matching problem is proposed after study of the imaging tracker's lag and low frame frequency, and their influence on performance of tracking control loop. To compensate the tracker's delay effect, a two-stage Kalman estimator for tracking maneuvering and non-maneuvering targets are provided. The method utilize the first stage of which is a bias-free filter providing target position and velocity estimates, and the second stage of which is a bias filter providing estimates of target acceleration. These two filters act together to provide parallel processing calculations thereby achieving high speed target state determination. During target maneuvers, the output of the second stage is used to correct the output of the first stage. In the absence of maneuver, the second stage is turned off and the first stage provides the target position and velocity estimates. To eliminate the influence of low frame frequency, a direct multi-rate feedback (DMF) control algorithm is brought forth by combining the low rate sub-controller with the fast rate one. The desired rise time of system response is depended on the low rate sub-controller. While the stability margin of system is guaranteed by the fast rate one. It is equivalent to increasing damping for increasing the number of effective controlled input signals in unit time though the fast rate controller. Combing the two-stage Kalman predict filter and the DMF control algorithm, we got the unified multirate predict control scheme (UMPC), which is effective for matching the servo control and tracker output. It is proved that the UMPC algorithm is effectiveness in improving system stability, dynamic performance and tracking accuracy.5. Tracking performance evaluation indictors of EOST systems have been investigated. The evaluation of tracking performance is provided by introducing the notion of random sensitivity function. This function, obtained using the method of stochastic linearization, characterizes the stationary tracking error for band-limited random references. Several of its functions, referred to as tracking quality indicators, are introduced to characterize complex tracking behaviors that arise in EOST systems. Various causes of poor tracking (e.g., static unresponsiveness, lagging, oscillatory responses, controller windup, and limitations due to finite trackable domains) can be identified by these indicators. The random sensitivity function and tracking quality indicators are useful for predicting the quality of tracking in EOST systems.
|
PDF Full Text Request