Warehouse Robot Path Planning And Scheduling Optimization

With the rapid development of Internet technology and e-commerce today,with the introduction of Industry 4.0 and Intelligent Manufacturing 2025,traditional industrial manufacturing and logistics warehousing systems are undergoing major changes.AGV,as the most important and most widely used automated logistics warehousing system A transport robot where every advancement and change is noticed and has an impact on the industry.There are many problems in the traditional AGV operation system that need to be improved,such as the problem of path planning efficiency and the problem of coordinated scheduling between multiple AGVs.This paper takes three-dimensional warehousing as the research object,optimizes AGV path planning and multi-AGV scheduling planning,and conducts verification tests on the simulation platform based on Unity3 D.Aiming at the problems of premature convergence and long time-consuming in traditional AGV path planning,an improved genetic algorithm with bidirectional search was proposed.First,an environment map is established,and the real environment space is abstracted to obtain a highly abstract environment map,which can meet the requirements of the AGV path planning algorithm running on the map.Secondly,the simulated annealing algorithm is introduced to transform the genetic operator,which improves the selectivity and diversity of the population.In addition,the fitness function is modified,and the indicators of path distortion and load are added to improve the path of the mobile robot and reduce collisions,so that the safety of driving in a shorter time can be guaranteed.Then add and set the rules for two-way search.Finally,Matlab modeling and simulation software is used for simulation verification,and the time-consuming time is used as the evaluation method.The results show that the improved algorithm is better than the classical genetic algorithm and the adaptive genetic algorithm.In traditional warehousing,it is generally a single operation mode,with poor resource allocation and low efficiency.In the scheduling optimization problem of multiple AGVs,the genetic algorithm has the advantage of being adaptable and satisfying most problems.Aiming at the shortest time-consuming multi-task in warehouse,an improved multi-swarm algorithm is proposed.Firstly,the initial solutions of multiple populations are arranged for multiple populations,and the population is encoded with real numbers,and the determination formula of the initial number is given.Second,the fitness function that limits the time is designed.Next,the design of the adaptive genetic operator is carried out.In the process of evolution,the size of the genetic operation will change due to the change of the degree of adaptation,and the calculation formula of the probability of crossover and mutation is given.Finally,it is verified by Matlab modeling and simulation software to verify the applicability and effectiveness of which method should be placed.After introducing two improved methods based on genetic algorithm,this paper builds a virtual simulation platform based on Unity3 D to simulate the warehouse environment.Combined with the previous research content,the two improved algorithms are run in the virtual environment,which proves the rationality.and feasibility.The results show that the improved algorithm has good performance.Run the traditional genetic algorithm in the Unity3 D simulation system,compare and prove the rationality and feasibility.The use of the Unity3 D virtual simulation platform also achieves visibility and better expresses the algorithm strategy.