Research On Key Technologies Of AGV Precise Path Navigation System In Intelligent Workshop

With the "made in China 2025" plan approved by the State Council in March 2015,China’s planned 10-year industrial plan was officially launched.Using and manufacturing automation and intelligent equipment is the only way in the innovation driven road of digitalization,networking and intellectualization of manufacturing industry.In engineering practice,automated guided vehicle(AGV)is often used in the material transportation in the processing workshop.AGV is used instead of manual transportation to improve the work efficiency,but sometimes the transportation has deviation.If these errors are ignored,there will be material delivery failure or even collision.Therefore,this paper establishes a path planning navigation system,through the construction of grid map,path planning and path deviation correction,so that AGV can accurately deliver materials to the target point.The main work of this paper is as follows(1)According to the requirements of AGV transportation system,the AGV precise navigation system of intelligent workshop is established.A simulation workshop material handling experimental platform based on ROS system was built,and the two-dimensional grid map of workshop was constructed by laser Salm technology,and the accuracy of the map was verified,which made a good foundation for the construction of the system and the realization of subsequent functions.(2)Combined with the constraints of AGV material handling path in intelligent workshop,an improved ant colony algorithm is proposed.Using the proposed algorithm,the path planning is carried out in the two-dimensional grid map of the workshop.This paper analyzes the optimal path length,running time and turning times of the improved ant colony algorithm,and compares with the ant colony algorithm and A* algorithm,which proves the effectiveness of the improved ant colony algorithm used in this paper.(3)In order to solve the problem that laser guided AGV deviates from the established target due to the error of its path,an optimization algorithm based on RBF neural network is proposed to solve the problem.After the AGV deviates from the planned path,the RBF neural network algorithm is used to correct the AGV trajectory and return to the planned path.Taking lidar as the core,a sensor system composed of gyroscope,odometer and lidar is constructed as the data input to predict the position and attitude of AGV using RBF neural network.Then,the established calibration model is used to correct the path deviation,which can reach the given target in the predetermined precision requirements.Then a verification experiment is constructed to prove the path deviation of the algorithm model The validity of difference correction.(4)Aiming at the problem that the AGV using RBF neural network correction algorithm can not be real-time continuous correction.Through the establishment of a database and correction training,the corrected pose information of each time is gradually memorized and stored in the database.Firstly,the database information and RBF neural network are synchronized offline to update the correction information of AGV to update the database information.Then,only depending on the database information to get a material handling route which only depends on the database information to correct the path.