Adaptive Nonlinear Filter Algorithm Based On Current Statistical Model

According to current statistical model algorithm leading to poor tracking accuracy and divergent, it is presented a new adaptive nonlinear filter in this paper. It is not only to compensate the defect of the current statistical model algorithm, but also can be effective to adjust the system gain and covariance in real-time to enhance maneuverability of the tracking target. Meanwhile it can overcome the trap of residual error´s asymmetric information. The simulation and experiment show that it has excellent tracking characteristic. The error of a new adaptive nonlinear filter is less than current statistical model algorithm.