Research On AGV Path Planning Strategy In Mixed-Flow Production Workshop

At present,the global manufacturing industry is developing rapidly,and the demand for product diversification and customization is increasing day by day.Mixed-flow production has become the mainstream workshop production mode.In this production mode,material transportation efficiency is an important factor affecting production efficiency.Automated Guided Vehicle(AGV)is a highly intelligent vehicle,which is widely used in the transportation system of workshops because of its advantages of safety and stability.The normal driving of AGV depends on path planning,and its planning effect will directly affect the transportation efficiency of AGV.Therefore,path planning strategy has always been a hot research direction in the field of AGV.In this paper,aiming at reducing AGV moving cost and improving transportation safety,the global path planning and local cooperative obstacle avoidance involved in AGV path planning in mixed-flow production workshop are studied in detail.The main contents are as follows:(1)The demand of AGV path planning in mixed-flow production workshop is analyzed,and the overall scheme of AGV path planning strategy is formulated.Considering the characteristics of the workshop scenes involved,the grid method is selected to construct the map model of AGV workspace.On this basis,the global path planning algorithm and multi-AGV obstacle avoidance strategy are compared and analyzed.Ant Colony Optimization(ACO)is proposed as the basic path planning algorithm,and the time window algorithm is selected to solve the AGV path conflict problem.(2)Research on global path planning strategy of AGV.Ant colony algorithm has some problems in AGV path planning,such as long convergence time and local extremum.This paper proposes an improved algorithm combining grey wolf optimization with ant colony optimization.Taking advantage of the easy realization of grey wolf algorithm,the optimal solution is introduced into the pheromone model of ant colony algorithm to solve the blind search problem caused by the unclear pheromone difference in the initial stage of ant colony algorithm.Then,the heuristic information is modified,the rotation angle constraint is added to the heuristic function to reduce the redundancy of the path,and the heuristic factor is updated adaptively,and the proportion of the two in the calculation is dynamically weighed.Furthermore,the pseudo-random strategy is used to select the path,which accelerates the convergence of the algorithm and introduces the conversion rate to adjust the balance between certainty and randomness.The simulation results show that the improved algorithm has good optimization performance and obvious advantages in improving the path quality.(3)Formulate collision avoidance strategies for multi-AGV path planning.Based on the study of global path planning strategy,the time window algorithm is applied to detect multi-AGV paths.According to the difference of conflict types,two different collision avoidance methods,namely,re-planning method and waiting method,are adopted to determine the occupation order of conflict path resources,combined with the difference of vehicle priority.Using the proposed method,each AGV in the experiment can effectively avoid path collision and successfully complete the task,which provides an effective scheme for improving the operation efficiency of AGV system.