Research On Motion Control Of Multi-Agent Persistent Formation

In this paper, we mainly focus on the path planning and control of 3-dimensional multi-agent formation, which will be a challenging problem. The main works are as follows:Firstly, as some practical applications, 3-dimensional multi-agent persitent formation is required to maintain a specified shape during continuous motion, such as surveillance, reconnaissance. Thus, we focus on decentralized control laws for the cohesive motion of 3-dimensional multi-agent formation in this paper, where each of the agents makes decisions based only on its own observations and states. Considering a point-agent system model, we categorize the agents according to the numbers of degrees of freedom (DOFs) of them, i.e., we label the agents as 0-DOF agents, 1-DOF agents, 2-DOF agents, 3-DOF agents, respectively. Then we design various control laws to different classes, distinctively. Furthermore, let the target of the formation move continu-ously along the prescribed trajectory, and then the 3-dimensional persistent formation can move cohesively along the trajectory. The effectiveness of the proposed control laws is demonstrated via the simulations.Sencondly, this paper provides a practicable framework for planning and control of 3-dimensional multi-agent formation. In this work, in order to define various regions of interest, 3-dimensional environment is divided into cubic grid cells. Then the A* search coupled with a optimizing algorithm are performed to find a collision-free and deadlock-free feasible path for the whole formation.Finally, many numerical simulations are given to verify the effectiveness and feasibility of the corresponding theoretical results.