Formation Optimization and Control for Maneuvering Target Tracking by Mobile Sensing Agents

Abstract

This paper presents a geometric formation control strategy of multiple sensing agents for maneuvering target tracking, which ensures agents to track the target with optimal target state estimation. Three sub-problems are solved for achieving the target tracking. First, an IMMCEKF algorithm is proposed to estimate the maneuvering target state by fusing multiple bearings-only measurements. This algorithm combines the interacting multiple-model estimator and the extended Kalman filter-based augmented measurement fusion algorithm, which is a centralized filter. Second, the algorithm of constructing optimal configuration for target tracking is proved and verified by maximizing the determinant of the Fisher information matrix. Third, we transform the problem of target tracking with optimal configuration into the problem of formation control. A geometric formation control approach based on Jacobi vectors is proposed. The formation shape controller and the formation tracking controller are decoupled because we use the formation center to describe the formation motion which is not relative to the formation shape described by the Jacobi vectors. The simulation results show that multiple sensing agents can track the moving target with optimal configuration such that the estimation error is obviously reduced.