Radar data association method based on simplified multi-hypothesis algorithm

Abstract

The invention provides a radar data association method based on a simplified multi-hypothesis algorithm. In view of problems of combination explosion and calculation amount exponential rising caused by a delay decision making mechanism of the traditional multi-hypothesis algorithm in a multi-target tracking system, a likelihood ratio scoring function and a linear distribution (LAP) pruning method are introduced, a track association logarithm likelihood ratio is used for simplifying hypothesis association probability calculation, M optimal hypothesis at the current moment is obtained through LAP algorithm track level pruning, through further global level hypothesis pruning, an optimal hypothesis matching sequence is obtained quickly as an effective association measurement point, and the purpose of reducing the calculation amount finally is realized. The method particularly comprises three main steps of track scoring calculation based on the likelihood ratio function, LAP-based track level M-optimal hypothesis pruning and LAP-based global level optimal hypothesis generation. By using the method provided by the invention, on the basis of ensuring the association accuracy, the calculation amount during a multiple hypothesis algorithm data association process can be reduced to a large degree.

Description

technical field [0001] The invention belongs to the technical field of radar multi-target tracking data association methods, in particular to a radar data association method based on a simplified multi-hypothesis algorithm. Background technique [0002] In the environment of dense radar echoes, multiple targets interfere with each other, and it is easy to cause false correlation between measurement and real targets or interruption of target track, which seriously affects the accuracy and efficiency of target tracking. The role of the data association algorithm is to correlate and match the measurement data with the real target track, so as to realize accurate tracking of multiple targets and correct update and maintenance of the target track. In radar multi-target tracking, the data association algorithm mainly solves two types of problems: first, how to select the correct target trajectory for association when the measurement of a target falls into two target correlation ga...

AI Technical Summary

Problems solved by technology

The tracking effect of this method relies on prior knowledge, has high requirements for the initialization of relevant parameters, and has high computational complexity

Although the optimal global hypothesis can also be obtained by using the ordinary m-optimal hypothesis pruning method, it needs to calculate the associated probability of all hypothesis branches in the pruning process, so the calculation load is large and the real-time performance is limited.

Images