Tracking Control of Multiple Mobile Robot Trajectory by Genetic AlgorithmsFull text in English

摘要

This paper proposes a robust trajectory tracking approach for formation control of multiple nonholonomic wheeledmobile robots. Using Lyapunov theory, A ? φ controller is proposed to derive the linear and angular velocity inputsof each follower robot to keep a desired separation distance and deviation angle with respect to the leader robotwhile the overall system is asymptotically stable. In addition, genetic algorithms (GA) is applied for optimizing thegains of the A ? φ controller. Optimizing the controller makes the tracking errors smaller between the referenceand the actual trajectory, and augments the precision of the formation tracking. To promote the stability androbustness of the formation in case of communication failure among the mobile robots, which may occur very oftenin real time applications. Therefore, least square (LS) approach is adopted for online estimation of the leader’strajectory whenever communication is failed between the leader and the follower robots. The optimal values ofthe controller’s gains and the least squares (LS) coefficients are estimated on offline mode with genetic algorithms(GA) using Matlab Toolbox. Finally, some Matlab simulations are carried out to demonstrate the effectiveness ofthe proposed approach.