A Hybrid Motion Planning Algorithm for Multi-robot Formation in a Dynamic Environment

Abstract

Motion planning with the multi-robot system is complex and challenging in a dynamic environment with unknown obstacles. This paper proposes a two-stage motion planning strategy to guarantee the formation of multi-robot when they pass through some obstacle areas. First, the multi-robot formation is constructed by a leader-follower strategy. Then a two-level algorithmic framework is introduced to optimize both global and local path planning. A Self-adaptive Dynamic Window Algorithm (SDWA) is proposed to dynamically search both global and local optimal path, which aims to balance the speed and safety of the multi-robot system. In SDWA, the weight parameter of the objective function is adaptively controlled based on the distance between the multi-robot system and obstacles. Finally, results of experiments demonstrate convincingly that the SDWA can avoid dynamic obstacles by switching between global path planning and local path planning modes, which makes the trajectory more reasonable in a complex environment, and ensure the smooth path and safety of multirobot formation.