Persistent Tracking of Maneuvering Target Using IMM Filter and DMPC by Initialization-Guided Game Approach

Abstract

In this paper, we focus on planning the paths of multiple unmanned aerial vehicles (UAVs) attached with camera sensors for persistently tracking a maneuvering ground target in urban environment. Considering the case of changing among various models, the interacting multiple model filter along with the state-vector fusion is utilized to estimate and predict the target motion, observed by multisensor networks. Then, based on the estimation results, the distributed model predictive control (DMPC) is taken as the framework of optimizing multi-UAVs paths with applications to persistent target tracking. When it comes to the DMPC solver, the game approach based on Nash optimization is adopted iteratively, and the guided initialization is introduced especially to improve the quality of initial solutions. Finally, we give numerical simulations to verify the high efficiency and feasibility of our proposed method.