Multi-Model Cardinalized Probability Hypothesis Density Filter For Multiple Maneuvering Target Tracking Algorithm Research

Tracking multiple maneuvering targets,including jointly estimating the number of targets and their states,is a challenging problem in theory and practice in the presence of clutter,noise,and uncertainty from target maneuvering,data association and detection uncertainty.The CPHD filter is suitable for tracking targets of multiple fixed motion models,but when the motion model of the target constantly changes,it will no longer be suitable.Using a combination of multiple motion models to deal with a situation where the target motion models switching between several models results in better performance.This paper combines the JMS model with the CPHD filter to solve the problem of multiple maneuvering targets tracking after reading the existing literature on multiple maneuvering targets tracking.The main research content of this paper includes the fo llowing two aspects:First,the multi-model CPHD filtering algorithm under the linear gaussian maneuveri ng targets tracking models.The multiple maneuvering targets tracking problem mainly use BFG approximation method to approximate the multiple models density funcition under linear condition,so that the multi-targets intensity function is independent of t he target's motion model,the multi-model maneuvering target tracking problem is tran sformed into a single-model maneuvering target tracking problem,and then the singlemodel CPHD filtering algorithm can be used to jointly estimate the state and number of multiple maneuvering targets.The second is the study of the multi-model CPHD filtering algorithm under the no nlinear non-Gaussian maneuvering target tracking models.The nonlinear multiple mane uvering targets tracking problem mainly use the Gaussian sum approximation method t o linearize the nonlinear non-Gaussian system,and then combine the multi-model tech nique with the nonlinear Gaussian sum CPHD filter to track multiple nonlinear maneu vering targets.The Gaussian mixture implementation of the multi-model CPHD filter under linear and nonlinear conditions is described respectively,and through the MATLAB simulation experiment,the performance of the algorithm is analyzed and evaluated.