Abstract: The performance of tracking association algorithm using target topology information is influenced by the estimation accuracy of the reference target state because of sensor bias. To solve such problem, a grey track anti-bias association algorithm based on centroid topology of reference is proposed. First, the algorithm takes the position and course of the sensor commonly observed targets' fusion centroid as the reference, and constructs the target topological vector. Then, the subsequential modified grey association analysis algorithm is used to calculate the grey association degree of the sensors. Finally, the global optimal track anti-bias association is carried out using the grey association degree as test statistical vector. Simulation results show that the performance and robustness of the proposed algorithm are better than the traditional algorithm apparently in random cross targets and dense parallel formation environment.

Key words: communication, track association, system bias, topology of reference, grey association degree, anti-bias association