Research On AGV Motion Control Method For Industrial Applications

With the continuous advancement of the modernization process,autonomous control theory and sensor technology have been improved and developed.As a kind of automatic logistics transmission equipment,AGV is widely used and its autonomous navigation technology also has important research significance.High-precision positioning,flexible path planning and accurate track tracking are important indicators to measure the performance of AGV.This dissertation carries out on the motion control method of AGV,the main research contents are as follows:(1)In view of the problem that the cumulative error of a single inertial measurement unit increases over time,a data fusion algorithm base on multiple inertial navigation combination is designed to improve the positioning accuracy of AGV.The update equations of position and attitude are derived,and the error model of inertial navigation is established.The magnetometer is used to assist inertial MEMS to complete the transfer alignment process.A federated data fusion method is designed and the error model is extended to federal Kalman filter to improve the positioning accuracy.The state value is corrected by real-time feedback and the best navigation parameters are obtained by optimal fusion.Simulation and experiment are designed to verify the feasibility of the proposed method.(2)Given the shortcoming of ant colony algorithm for path planning,such as slow convergence speed,excessively long path length and easy to fall into local optimum,traditional ant colony algorithm is improved by introducing discussion mechanism.Through group discussion,pheromone factor and heuristic function factor in state transition probability can be adjusted adaptively with iteration,the algorithm is avoided from falling into local optimum.Simulation and experimental results show that the algorithm can effectively reduce the number of iterations and the length of path planning,and can plan an effective path in multi-obstacle environment,which proves the effectiveness of the algorithm.(3)In order to overcome the problem of low accuracy and stability when AGV performs trajectory tracking,an improved model predictive control algorithm base on vehicle kinematics model is proposed to accomplish the assignment of trajectory tracking.The state value of vehicle is estimated by particle filter to improve the positioning accuracy of AGV.The gradient projection method is introduced in the design of controller.The new method reduces the number of iterative steps and expedites the convergence,improving the computational efficiency of AGV tracking controller.The algorithm is verified by simulation experiments of straight,curve,circular and compound paths,the results confirmed that this method has strong tracking precision.