关键词: 目标跟踪, 无味卡尔曼滤波, 交互式多模型, 非线性估计

Abstract:

Aiming at the uncertainty of target motion-mode and the nonlinearity of target motion model,this paper proposes an Interacting Multiple Model (IMM) algorithm based on Unscented Kalman Filter (UKF).The constant velocity model,the constant acceleration model,the coordinated turn model for fixed angular velocity and the “current” statistical model are chosen as the model set,and the Unscented Kalman Filter is proposed to realize the nonlinear state estimation.The simulation results show that this algorithm has higher tracking performance than the traditional IMM algorithm.

Key words: target tracking;UKF;IMM;nonlinear estimation