Two-Wheel Differential Drive AGV Path Planning And Trajectory Tracking Research

Automated Guided Vehicles(AGV)is a kind of intelligent mobile robot with a high degree of automation.With the intelligent development of logistics systems,AGVs are used in a large number of smart warehouses,smart factories and logistics transportation,etc.The use of AGVs can reduce labor costs and increase the efficiency of logistics transportation.However,in the operation environment with narrow space and complex environment,the performance of AGV path planning algorithm affects its operation efficiency and even the feasibility of the path,while the trajectory tracking algorithm affects the movement accuracy of AGV.To this end,this thesis takes two-wheel differential drive type AGV as the research object and carries out research on path planning and trajectory tracking technology,the main research contents are as follows:(1)Kinematic analysis and modeling of AGVs.Based on the incomplete constraint characteristics of two-wheel differential drive AGVs,the posture coordinate system is established and the coordinate transformation equations are derived,and the differential AGV kinematic model and the posture error model are constructed.(2)Study the AGV global path planning algorithm.First,to address the blind search and local optimum problems of the traditional ant colony algorithm,the improvement of the ant colony algorithm is realized by adjusting the initial pheromone distribution,designing a new heuristic function,and optimizing the pheromone update rule and volatility factor.Then,two typical environments are designed and the global path planning simulation of AGV based on the improved algorithm is carried out to prove the theoretical feasibility and effectiveness of the algorithm.(3)Study of AGV local path planning algorithm.To solve the obstacle avoidance problem,the optimized artificial potential field algorithm is used for local path planning.By designing a new repulsive force function,the relative distance between AGV and the target point is accounted for in the original repulsive force function to solve the target reachability problem;by introducing virtual traction,the AGV is able to jump out of the local minima trap.Simulation experiments show that the algorithm can make the AGV avoid obstacles effectively and achieve collision-free smooth path planning.(4)Study the AGV trajectory tracking control problem.On the basis of path planning,the AGV trajectory accuracy problem is explored based on the kinematic model and the positional error model.Four common convergence law characteristics are analyzed,and the backstepping method is used to design the sliding mode control switching function to obtain the improved convergence law.Simulation experiments on linear and circular trajectory tracking show that the improved sliding mode controller can effectively reduce jitters and verify the stability of the method in improving convergence speed and accuracy.(5)Build the experimental platform and validate it by example.The traditional algorithm and the algorithm of this paper are imported into the ROS system,the AGV operation environment is built,the motion start and stop points are set,and the path planning of the AGV in a typical environment is realized.From the analysis of the experimental data,it can be seen that under the same environment,this method has a more obvious advantage in the search time and path length.