| With the proposal of industry 4.0,industrial Internet and intelligent manufacturing,manufacturing enterprises have more and more clearly realized that the development direction of enterprises in the new era has become to empower the traditional manufacturing industry by using technologies such as Internet of things,artificial intelligence and big data.And it changes the traditional manufacturing from taking scale,cost and benefit as the first priority to taking customized experience and innovative delivery as the core of competition."If you want to do a good job,you must first sharpen your tools".Robots and their applications that deal with small quantities of work with high accuracy and flexibility have become a crucial point.Taking the intelligent factory workshop as the background and aiming at flexible,accurate,intelligent and small batch industrial production,this thesis studies the task allocation and path planning of material handling with robot as the main part.The specific research contents are as follows:Firstly,based on the analysis of robot control mode,the control mode of material handling robot suitable for intelligent factory is studied.According to the characteristics of intelligent factory production,a hybrid control mode is proposed,in which the central control center and robots are responsible for different tasks in task allocation and path planning.Secondly,a robot task assignment method based on auction algorithm is proposed.In view of the shortcomings of the traditional auction algorithm,such as more round trips and more total system consumption,a combined auction task allocation method based on K-means clustering algorithm is proposed,and the selection of the initial clustering center is improved to reduce the mobile path of the robot and the total system consumption;Aiming at the problem that the workshop needs to add new tasks in real time,an improved dynamic auction algorithm is proposed,which also reduces the round-trip path,The simulation results show that this method can get better allocation results than the traditional auction algorithm.Thirdly,taking the robot as the main body,the path planning problem of single robot is studied.The path planning process is divided into global planning and local planning.Aiming at the problem of increasing the amount of algorithm calculation in complex environment,a two-way a * search algorithm is proposed to plan a global optimal path,which improves the search efficiency;Aiming at the problems of multiple turning points and large turning angle of the path searched by a * algorithm,the redundant points of the path are removed and smoothed in combination with Bezier curve to obtain a smooth and optimal path conducive to robot turning.Fourthly,considering the collision and other conflicts caused by multi robot handling,the path planning problem of multi robot handling is studied.Based on the research of single robot path planning,a method of setting robot priority is proposed.The robot with low priority invokes the improved artificial potential field method proposed earlier to avoid obstacles.Simulation results on MATLAB show that this method can successfully solve the conflict between machines.In view of the dynamic obstacles,local path planning using artificial potential field method,by adding virtual target point to solve the traditional artificial potential field method is easy to fall into local minimum value problem,at last the improved A * algorithm is combined with artificial potential field method,the simulation verify the method can effectively planning out A better path and managed to avoid dynamic obstacles.Finally,the task allocation method and path planning method are combined to form a complete overall scheme of material handling;The simulation results show that the scheme can make the robot carry out reasonable task allocation and plan the optimal path,and finally reach the target location without collision,which verifies the effectiveness and feasibility of the overall scheme. |
PDF Full Text Request