Multi-channel Incomplete Measurements Estimation Algorithm Under Study

In target tracking area, there widely exists incomplete measurement. In practical applications, measurements are often multi-channel. Therefore, the study on estimation algorithm in the case of multi-channel incomplete measurement is of great significance. For the nonlinear in target tracking system and multi-channel incomplete measurement, this paper focuses on the estimation algorithm in the case of multi-channel incomplete measurement, which contains two kinds of situations. One situation is that the multi-channel incomplete is coupled, and the other situation is that the multi-channel incomplete is not related. The critical detection probability is also studied in this paper. The main research results are as follows:First, the multi-channel incomplete measurement tracking system is modeled. Based on the previous research and according to the practical applications, we designed two kinds of models. One kind is a multi-channel incomplete measurement with coupled detection (MIMCD); the other is a multi-channel incomplete measurement with decoupled detection (MIMDD).Next, we study the estimation algorithm based on MIMCD. Using the projection theorem we derived UCMKF based on MIMCD. By solving modified Riccati difference equation we get the robust filter gain which makes the filter results uniform stability. By simulation method these two methods are compared and analyzed, the simulation results show that the robust filter get higher precision and UCMKF get better real-time.Then, we also study the estimation algorithm based on MIMDD. With the help of the projection theorem we derived the EKF algorithm based on MIMDD. In this basis, According to the standard Kalman filter we designed the robust filter based on MIMDD. We designed the robust filtering algorithm based on MIMDD. Simulation results show that the robust filter has better robustness and higher precision. EKF may be unstable in the low probability of detection, and the error variance is large. But EKF run much faster than robust filter.Finally, critical detection probability is studied. This paper proved the existence of the critical detection probability, and it also gives algorithms to calculate the critical detection probability.