Trajectory Tracking Control Method For An Autonomous Guided Vehicle

Automated Guided Vehicles, also known as the AGV or wheeled robot, is an unmanned intelligent truck traveling along a pre-set path, and trajectory tracking control system of AGV is a typical nonholonomic and nonlinear system. In the past few decades, since their potential applications in the study of the AGV, they have been drawing more and more attention by the researchers. This paper mainly studies the trajectory tracking control problem of two wheeled differential-drive AGV.Firstly, according to the methods of building the kinematics and mechanics model of the two wheeled differential-drive AGV, we analysis its dynamic characteristics and give the process to deal with the nonholonomic constraints, in order to conduct the further simulation of the AGV trajectory tracking control algorithm in Matlab.Secondly, Lyapunov direct method, sliding mode variable structure method and the backstepping method are used to design the trajectory tracking control law for the Kinematic model of the AGV to prove the global stability of the system under the control laws by designed Lyapunov function. And under the Matlab/simulink environment, the results of tracking a circular trajectory prove the effectiveness of the control laws. At the same time, after comparing the three designed methods above, we found that backstepping has better tracking performance.Thirdly, in order to give the full consideration to the impact of trajectory tracking control of the AGV’s dynamic characteristics, the proposed double closed-loop control structure to achieve the synergistic effect of the position/force. The outer loop uses backstepping method to design the position and attitude controller for the kinematic model of AGV; while the inner loop uses the integral sliding mode variable structure control method for mechanic model of AGV, in the controller, adaptive control law is used to estimate the uncertainty of external interference. To obtain better control performances to prevent severe vibration, we use the hyperbolic tangent function to replace the sign function to ensure the control input and states of system are continuous and smooth by tracking eight-shape trajectory in the Matlab/simulink environment. The simulation results show the effectiveness of the control structure and control algorithm.Finally, in order to further study of the application of intelligent control algorithm in the model in this paper, we propose an idea of decoupling fuzzy control to achieve trajectory tracking control of the AGV. Under the reasonable assumptions, we firstly decouple the mechanical model of the AGV, and the design fuzzy controller for the decoupled subsystems respectively. The simulation results show that good control effect and robustness of the control algorithm.