Multisensor Multitarget Tracking and Collision Avoidance Based on the Labeled RFS Void Probability and KL-PaC Cost Function

Abstract

Existing sensor control methods for multi-sensor and multi-target tracking seldom consider comprehensive control strategy and collision avoidance between sensors and targets or between sensors together. In this paper, we first propose a cost function for multi-sensor control strategy, which includes Kullback-Leibler (K-L) divergence, detection probability and sensor usage cost (KL-PaC cost function). Second, in order to solve collision avoidance between sensors and targets, we propose a collision avoidance strategy based on the labeled random finite set (RFS) void probability. The strategy is essentially to set up a safe region around the sensor or target, and then combine the void probability and cost function in the safe region as the evaluation function. The final simulation and experimental results verify the effectiveness of the proposed approach.