Formation Control of Multiple Nonholonomic Mobile Robots with Limited Information of a Desired Trajectory

In the study of task coordination for multiagent systems, formation control has received considerable attention due to its potential applications in civil and/or military practices. Fundamentally, formation control problem for multiagent systems can be formulated as making a group of agents follow the desired trajectory while maintaining certain prescribed geometric distances among agents. In this paper, we consider the formation control problem for mobile robots with nonlinear dynamics and moving in a 2D environment. To address the inherent challenges due to nonlinear system dynamics and agents' limited sensing/communication capabilities, we instill an idea of integrating the recently developed distributed consensus theory into the standard feedback control, and propose a new time-varying cooperative control strategy to solve the formation control problem for multiagent systems. In particular, the proposed design only requires the local and intermittent information exchange among agents to achieve the formation control objective. More importantly, we remove the restriction on the need of the desired trajectory for every agent, and instead design a distributed observer for obtaining the desired trajectory in order to establish the formation in the design. The overall distributed formation control system stability is rigorously proved by using a contraction mapping method under the condition that the sensing/communication network among robots is sequentially complete. Simulation is provided to validate the effectiveness of the proposed design.