The bearing only target tracking of UUV based on cubature Kalman Filter with noise estimator

A Cubature Kalman Filter with noise estimator is proposed to solve the problem that the selection of the statistic property parameter is not reasonable, which leads to filtering algorithm declining in accuracy and even diverging, when the noise statistic property is unknown in the bearing only target tracking of Unmanned Underwater Vehicle. This algorithm can estimate the noise statistic property by the noise estimator when the Cubature Kalman Filter estimates the position and the speed of the target. The simulation results show that the Cubature Kalman Filter with noise estimator has better filtering accuracy and has better fault tolerance to the unknown noise statistic property compared with traditional Cubature Kalman Filter when the noise statistic property is unknown.