A Joint Probability Data Association Filter Algorithm for Multiple Robot Tracking Problems

摘要

This paper dealt with the problem of multi-robot tracking taken into account as one of the most important topics in robotics. The JPDAF algorithm was presented for tracking multiple moving objects in a real environment. Then, extending the aforementioned algorithm to robotics application was discussed. To enhance the quality of tracking, different motion models were introduced along with a simple near constant velocity model. Proposing a new approach for observer trajectory planning was the key part of this paper where it was shown that because of some problems such as increasing the variance of the additive noise by increasing the distance between targets and the reference robot, the tracking performance may be corrupted. Therefore, a fuzzy controller was proposed to find an optimal trajectory for the reference robot so that the effect of the additive noise is minimized. Simulation results presented in the paper confirmed the efficiency of the proposed fuzzy control approach in enhancing the quality of tracking.