Study on Kalman dynamic prediction and feedback parameter optimization of laser tracking system

Under the environment of MATLAB, a closed-loop feedback laser tracking system was established based on the dynamic prediction with Kalman filter and some other filtering processes. Different motion states of the tracked target are simulated to test the tracking performance. The following conclusions are obtained through simulations. After adding the dynamic prediction with Kalman filter, the tracking hysteresis can effectively be avoided when tracking the dynamic object. Taking advantage of the high frequency of PSD, the coordinate value can be read for several times in a feedback loop and then filtered to obtain a more accurate spot coordinate. Under static condition, the instability due to noise can be reduced through segmenting the feedback coefficient by distinguishing between the dynamic and static state of the object. According to the above designs, the results of the laser tracking system simulations show that the dynamic tracking performance is better than 100°/s, and static stability has an order of magnitude improvement than before.