Radar multi-target tracking optimization method based on data correlation method

Abstract

The invention discloses a radar multi-target tracking optimization method based on a data correlation method, comprising: separately determining a total number T' of radar tracking targets; determining a measurement number nk corresponding to k time; successively calculating the candidate measurement set Zt' (k) of a t-th object at k time after optimization, a vector C (k) formed by appearing times of nk measurements in respective correlation window corresponding to T' targets at k time, and a vector C t' (k) formed by appearing times of candidate measurements in an nk*T' dimensional measurement-target incidence matrix Omega, the candidate measurements falling in a t-th object correlation window at k time; furthermore successively calculating the appearing times cit' (k) of an i-th candidate measurement in the nk*T' dimensional measurement-target incidence matrix Omega, the i-th candidate measurement falling in the t-th object correlation window at k time, a probability beta it (k) that an i-th candidate measurement zit' (k) in the candidate measurement set Zt' (k) of a t-th object at k time after optimization is from the t-th object, and a probability beta 0t (k) that no candidate measurement in the candidate measurement set Zt' (k) of the t-th object at k time after optimization is from the t-th object; and furthermore successively calculating a state equation X <^> t (k|k) of the t-th object at k time, and an error covariance matrix Pt(k|k) of the t-th object at k time.

Description

technical field [0001] The invention belongs to the technical field of radar, and in particular relates to a radar multi-target tracking optimization method based on a data association algorithm, which is suitable for a radar to track a single target or multiple targets in a clutter environment. Background technique [0002] In recent years, with the complex and changeable application environment, radars are required to have multi-target tracking capabilities and can simultaneously realize multi-target tracking; the basic concept of multi-target tracking was proposed by Wax in an article in the Journal of Applied Physics in 1955. After that, in 1964, Steer published a paper entitled "The Optimal Data Association Problem in Surveillance Theory" on IEEE, which became the forerunner of multi-target tracking, but at that time Kalman filtering was not yet widely used, Steer A track bifurcation method is used to solve the problem of data association; in the early 1970s, in the pre...

AI Technical Summary

Problems solved by technology

[0002] In recent years, with the complex and changeable application environment, radars are required to have multi-target tracking capabilities and can simultaneously realize multi-target tracking; the basic concept of multi-target tracking was proposed by Wax in an article in the Journal of Applied Physics in 1955. After that, in 1964, Steer published a paper entitled "The Optimal Data Association Problem in Surveillance Theory" on IEEE, which became the forerunner of multi-target tracking, but at that time Kalman filtering was not yet widely used, Steer A track bifurcation method is used to solve the problem of data association; in the early 1970s, in the presence of false alarms, the Kalman filter method (Kalman) was used to systematically track and process multiple targets; in 1971 The nearest neighbor method proposed by Singer is the simplest method to solve data association, but the correct association rate of the nearest neighbor method in the clutter environment is low; during this period, Y.Bar-Shalom played a pivotal role, he in 1975 Proposed a probabilistic data association algorithm (PDA) especially suitable for tracking a single target in a clutter environment, which effectively solved the problem of multi-target tracking in a clutter environment; T.E.Formann and Y.Bar-Shalom proposed a joint probability data Association Algorithm (JPDA), JPDA arranges and combines all targets and measurements, and selects a reasonable joint event to calculate the joint probability. JPDA considers the possibility that multiple measurements from other targets are in the same target interconnection domain, It can well solve the problem of multi-target measurement in an interconnected domain in a clutter environment; but at the same time, JPDA is more complicated, with a large amount of calculation, and with the increase of the number of targets, the split of the confirmation matrix will cause a combinatorial explosion. Happening

Therefore, JPDA is more difficult to implement in engineering

Images