Constrained minimum fuzzy error entropy filtering for target tracking

Abstract

To improve the accuracy of target tracking results in nonlinear systems under complex noise, the constrained minimum fuzzy error entropy Kalman filter (CMFEE-KF) is proposed. In this proposed filter, the double indices-induced fuzzy membership is introduced to represent the different effects of different error samples on the estimation results, solving the problem of the same weight in ordinary error entropy. And then, the minimum fuzzy error entropy criterion (MFEEC) is constructed and used to optimize the Kalman filter. In this proposed algorithm, error information is obtained by model reconstruction. Then the objective function is constructed based on MFEEC, and finally, the posterior state estimation is achieved by the fixed-point iteration method. In addition, soft constraints can be implemented by adding a regularization term into the loss function, deriving the CMFEE-KF. Simulations show that the proposed filter has strong stability and more accuracy in target tracking.