Maneuvering target tracking based on information fusion

In target tracking. target maneuvering and nonlinear estimation are often confronted Extended Kalman Filter (EKF) is the most general method for nonlinear problem. and Interactive multiple model (IMM) filter is now regarded as the most effective algorithm in tracking maneuvering target But when the assumption of its linearity does not agree with the reality. linearization will make the filter highly instable We present a method of modified Extended Kalman Filter (MEKF) based IMM tracking in this paper. On the one hand we adopt unscented transformation to alleviate the influence of model error. on the other hand we use IMM to raise the accuracy of model. thus solve the problem of maneuvering target tracking under nonlinear condition. The simulation result shows that this method is practical.