An Asynchronous track-to-track Association Algorithm without Time Alignment

摘要

Local sensors in distributed multi-target tracking systems are different in types for different missions. So local sensors usually have different sampling rates and transfer asynchronous track data to fusion centre. The current track association algorithms are mostly synchronous track association algorithms based on time alignment. Tracks need to be synchronized before the algorithms applied. But it brings extra estimation error when using the time alignment method. And the estimation error spreads at the same time, which affects the performance of the track association algorithm. To solve this problem, this paper presents an algorithm for asynchronous track to track association without time alignment. This paper provides a method of Real to Interval Transformation (RTIT) to describe the real number track sequences as interval number track sequences. Then a new distance measurement for the interval sequence is defined to measure the differences of each track sequence. So the correlation degree can be calculated, which describes the similarity degree between each track. Also the track association conclusion can be made. Simulation results show that the presented method can effectively solve the asynchronous track-to-track association problem, and the correct association rate maintains on high level.