Research On Formation Control Of Multi-Robot System

This paper focuses on the leader-follower formation maneuvers of a group of nonholonomic wheeled robots. On the basis of achievement of formation control by sliding mode algorithm, the core research of the paper concentrates on the uncertainties rooted in the formation system.Firstly, consider the influences of load variations, parameter perturbations, measurement noise, friction, nonlinearity of drive system, external disturbance, etc., dynamic model with uncertainties of a single robot is introduced in accordance with kinematic model. Subsequently, dynamic model of leader-follower formation system is formulated and the uncertainties are classified as matched and unmatched. It’s worthy mentioning that, for the purpose of minimizing the communication burden between leader and follower, the velocity information of leader robot is not needed any more to the follower robot and considered as unmatched uncertainties.Secondly, in the matched condition, conventional sliding mode control (SMC) strategy and terminal sliding mode control (TSMC) strategy are addressed. Both of the design processes and the selection principle of parameters of two controllers are illustrated, besides, the system stability is also touched. Take a platform composed of three robots as example, simulation demonstrates the effectiveness of these two proposed strategy and the invariance of SMC.Thirdly, the effects of unmatched uncertainties are particularly significant in practice, hence, under the assumption that the unmatched uncertainties are bounded with a known boundary, as a modified strategy, integral sliding mode control (ISMC) is adopted. According to Lyapunov stability theory, the formation system is proven to be locally asymptotically stable in the slidingmode stage, meanwhile, the reaching condition of sliding mode is guaranteed by the developed ISMC law. Finally, simulations are performed in two cases individually.Fourthly, the boundary of unmatched uncertainties is usually a subjective estimate and hard to get precise value. Thus, in case of the unknown boundary, an adaptive sliding mode control (ASMC) approach with fuzzy compensator is proposed. The fuzzy compensator is applied to approximate the unmatched uncertainties in real time in the formation system. The simulation results verify that the proposed method achieve to improve the formation precision efficiently.The last part concludes the research results and prospects the directions of future study.