Maneuvering Target Tracking Using IMM Kalman Filter Aided by Elman Neural Network

Abstract

In order to improve the performance of Interactive-Multiple-Model (IMM) Kalman filter in maneuvering target tracking, Elman neural network is applied to learn and predict the estimation errors of IMM Kalman filter, and correct the output results of IMM Kalman filter. The simulation experiment suggested that the proposed approach was able to further improve the tracking results from the aspects of the tracking trajectory and the root mean square error. Compared with the maneuvering target tracking using IMM Kalman filter aided by RBF neural network, maneuvering target tracking using IMM Kalman filter aided by Elman neural network is more valid.