Research On Path Following Control Method Of Single Rudder Wheel AGV

AGV is an automatic guided vehicle that is guided through different navigation methods and then through differential or steering wheel movement.Due to the rise of the logistics industry,it has gradually become the main force of Industry 4.0,so it has been highly valued by scholars in the automation industry,especially the AGV path Tracking the problems in this area is currently the most difficult and most in need of solution in the AGV industry.In the final analysis,the problem of AGV path tracking is to eliminate the problem of the deviation between the AGV and the set trajectory.This article first combines the LQR control and MPC control,which are widely used in unmanned driving technology,to carry out the path tracking control method of the single steering wheel AGV with laser navigation After the design and research,the positioning optimization based on sensor fusion is designed for the input error of the AGV of the laser navigation mode.Finally,the dynamic obstacle avoidance method based on the Bezier curve is proposed for the special situation with obstacles.The main contents include:First,analyze the kinematics model of AGV with single steering wheel structure,and introduce the method of trajectory generation.Then,the deviation model for identifying the tracking effect is obtained through the reference point of the trajectory and the reference point of the AGV vehicle body.Then,two control methods are proposed based on the LQR control,and an iterative method is proposed for the computer system to solve the Riccati equation.The analysis shows that based on The LQR control of the kinematic model has better robustness and rapidity because it obtains the steering wheel compensation angle by solving the optimal solution,but it also has the defect of not being predictive.Furthermore,the path following control based on MPC control is designed,and a method of dynamically changing the control time domain is proposed to solve the problem of insufficient computing power of the embedded control system.Aiming at the positioning problem of the laser navigator,according to the sensors commonly used in AGV car bodies,a sensor fusion optimization method based on extended Kalman filtering is proposed to filter and optimize the measured values of the laser navigator.Finally,according to the actual situation,path tracking may encounter obstacles,a dynamic obstacle avoidance method based on Bezier curve is proposed,and the effect of path tracking algorithm is verified again.