Font Size: a A A

Research On Multi-model Filtering Algorithm For Tracking Multiple Maneuvering Targets

Posted on:2020-09-17Degree:MasterType:Thesis
Country:ChinaCandidate:X L HuFull Text:PDF
GTID:2518306047497614Subject:Master of Engineering
Abstract/Summary:PDF Full Text Request
Multi-target tracking provides a basis for subsequent information processing by accurately estimating the state of multiple targets.With the development of society and technology,multitarget tracking has been more widely used in the military and civilian fields.Scholars have conducted detailed research on them and achieved many results.Based on the existing results,this paper makes a further study on multi-maneuvering target tracking in clutter background.The main research contents are as follows:The basic principle of multi-target tracking is expounded.The coordinate system commonly used in multi-target tracking is given.The mathematical model of coordinated turn of the target in the ground coordinate system is deduced in detail.The mathematical models of target for constant velocity,constant acceleration,and snake-like maneuver are studied.The CS model that models the target acceleration as a stochastic process is studied.In addition,the observation models under different observations and the tracking gates of multi-target tracking are also studied.Single target tracking in a non-clutter background is studied.The KF,EKF,FMKF,and IMM algorithms are studied.When the observation is the target position,the classical IMMKF algorithm is studied and verified by simulation.When the observation is relative distance and angle,the IMMEKF algorithm is studied based on the EKF which is linearized and then discretized.Aiming at the shortcomings of IMMKF algorithm,an IMMFMKF algorithm based on FMKF is proposed and verified by simulation.In view of the situation that the observation comes from the missile seeker,the missile-target relative motion model is established,and EKF is used to track the single target in polar coordinate system and the simulation is carried out.The multi-target tracking problem is usually decomposed into multiple single-target tracking problems.The study of single-target tracking lays a foundation for the subsequent study of multi-maneuvering target tracking.The multi-maneuvering target tracking in the clutter background is studied.The JPDA algorithm for tracking non-maneuvering multi-targets and the IMMJPDA algorithm for tracking maneuver multi-targets in clutter background are studied.The IMMJPDA algorithm is illustrated by simulation.Aiming at the optimization and improvement of the IMMJPDA algorithm,an improved IMMJPDA algorithm based on FMKF is proposed and verified by simulation.Two kinds of multi-maneuvering target tracking algorithms for engineering application in clutter background are proposed.Aiming at the shortcomings of JPDA algorithm,two simplified JPDA algorithms,Cheap JPDA algorithm and Suboptimal JPDA algorithm,are studied.The JPDA algorithm in the improved IMMJPDA algorithm is replaced by the Cheap JPDA algorithm,and an IMMCheap JPDA algorithm for engineering application is proposed and verified by simulation.The JPDA algorithm in the improved IMMJPDA algorithm is replaced by Suboptimal JPDA algorithm,and an IMMSuboptimal JPDA algorithm for engineering application is proposed and verified by simulation.
Keywords/Search Tags:multiple maneuvering targets tacking, interacting multiple model filtering, joint probabilistic data association, simplifying joint probabilistic data association, fading memory Kalman filtering
PDF Full Text Request
Related items