Abstract:

A generalized probabilistic data association algorithm fusing measurements of multi-sensor to estimate the position of target is proposed. Firstly, the algorithm implements pre-correlation statistical test on all possible multi-tuple of measurement, and estimates the targets position that the valid multi-tuple of measurement which has passed through the pre-correlation statistical test corresponds to. Then the association probability between the valid measurement and target track is calculated according to generalized probabilistic data association algorithm, and is further used to calculate the update state of target. In comparison with sequential processing of multi-sensor generalized probabilistic data association algorithm and joint probabilistic data association algorithm fusing measurements of multi-sensor to estimate the position of target, the new algorithm has the advantages of both the optimal estimate and effective reuse of information. Both theory analyses and simulation results have verified the effectiveness of the proposed algorithm.

Key words: multi-sensor multi-target, data association, generalized probabilistic data association, multiple tuple of measurement