Research And Design Of AGV Path Tracking System Based On Improved Sliding Mode Control

It is showed that automated guided vehicle plays an important role in the realization of high-speed,informatization and intelligence in the field of modern industrial logistics.As the core technology of AGV system,path tracking control is widely concerned in recent years.Sliding mode control algorithm is an effective control method for nonlinear systems with strong robustness,but the biggest disadvantage of the algorithm is that it is easy to cause chattering of the control system.In this paper,the chattering suppression of sliding mode controller and the optimization of AGV path tracking effect are studied with the help of intelligent algorithms such as RBF neural network and PSO,which breaks through the application limitation of sliding mode control algorithm in the field of AGV motion control,forms a reproducible improvement and optimization scheme,and also has great practical significance for other nonlinear systems.Firstly,the overall framework of AGV path tracking system is designed,the running environment and communication mechanism are set up for Turtlebot robot,Raspberry Pi vehicle controller and magnetic navigation sensor system.In order to design the sliding mode control law of AGV velocity and angular velocity,two procedures are carried on.On the one hand,the general AGV kinematics model and pose error model are established.On the other hand,the application of backstepping design method and the constant velocity reaching law of sliding mode switching function are studied.Simulation shows that the sliding mode control algorithm is better than the traditional PID algorithm in AGV path tracking effect.Subsequently,an improved sliding mode controller based on RBF neural network algorithm is proposed,in which the chattering of AGV running speed is used as the network learning signal to realize the online intelligent adjustment of the gain of sliding mode switching term to suppress the chattering of the system.After this improvement,the PSO algorithm which takes the integral of time multiplied by the absolute value of error criterion as the evaluation function is further studied to optimize the fixed parameters and initial values of the dynamic parameters in the controller.According to the improved scheme of sliding mode control algorithm,the software is designed to form a visual operation interface to obtain the optimal value of controller parameters.Simulation experiment shows that RBF neural network algorithm can effectively suppress the chattering of sliding mode controller,and PSO algorithm can also improve the convergence effect of AGV pose error.Finally,aiming at verifying the performance of AGV path tracking system designed in this paper,steady-state response experiments and interference experiments are carried out in the straight line,curves with different curvatures and combined path.It shows that the sliding mode control algorithm has faster speed,higher accuracy and better stability than PID algorithm.The improved sliding mode controller can fully suppress the chattering and accelerate the convergence of the system.The AGV path tracking system designed in this paper has the stability,reliability and robustness of tracking different paths.