Research On Perfect Tracking Control Of Dual-axis Opto-electronic Tracking Turntable

The dual-axis opto-electronic tracking turntable is a complex system combining optics, machinery, electronics and computer application. It is widely applied in military and civil fields. This paper is focused on the problems of its low tracking precision and poor dynamic performance when it moves fast in the wild. According to that, the research of perfect tracking control as well as object detection and tracking algorithm is carried out to improve the tracking precision and guarantee the dynamic performance of the system, which is of great practical significance.Firstly, considering the problems of low tracking precision and lack of dynamics performance, the research on perfect tracking control is carried out. And based on controlled quantity division, a perfect tracking control algorithm which is not limited by the system order is proposed. The proposed algorithm can overcome the shortcoming of the perfect tracking controller and has a good performance on tracking precision and dynamics.Secondly, the research on object detection and tracking is carried out to extract the position of the object and provide an accurate command signal for the tracking turntable. When tracking turntable moves fast in the wild, it is difficult to track the moving object because of the illumination change and the object of the similar shape and other interference factors. A hierarchical particle filtering algorithm based on multi-feature fusion is proposed. This algorithm utilizes the corner, color and edge features of the object hierarchically to obtain the weights of particles, improving the tracking precision and having good robustness.Finally, a physical experiment environment is built. And a semi-physical simulation system based on Real-Time Workshop and xPC Target is established. These two experiment environment offer the experimental platform for this research and lay the foundation of the future theoretical research and engineering application. The good tracking performance and robustness of the proposed algorithms are proved through the experiments.