The Square Root Cubature Kalman Filter-Markov Ahead Estimation Based Tracking Maneuvering Photoelectric Target Systems

A new method of ahead estimated tracking maneuvering targets is proposed to overcome the inherent measurement delay and control hysteresis. Square Root Cubature Kalman Filter-Markov (SRCKF-Markov) ahead estimation is used to predict the maneuvering target states at the next sampling time in advance. The “current” statistical models of maneuvering targets are set up, and the maneuvering states at the next sampling time are one-step ahead estimated by SRCKF. Then, the Markov matrixes of transition probability are set up with the errors between the prior estimates by SRCKF and the measured values. The next error between the prior estimate by SRCKF and the measured value is one-step ahead predicted by the matrixes of transition probability. The estimates of maneuvering target states with SRCKF are revised by the estimates of next errors. Meanwhile, a self-correcting predictive controller is used in the control system to overcome the control hysteresis. Simulations and experiment results show that the proposed estimation and control method can improve the accuracy of ahead estimated tracking greatly.