Research On Flexible Scheduling Modeling And Optimization Method Combined With AGV

Flexible job shop scheduling problem(fJSP)is an extension of traditional job shop scheduling problem(JSP),which has the characteristics of computer complexity,dynamic randomness and multi constraints.The intelligent and automatic production mode puts forward higher requirements for the material transportation in the workshop.The application of AGV in the flexible system can reduce the artificial pressure,shorten the transportation time and improve the production efficiency.Flexible scheduling modeling and optimization combined with AGV has become a research hotspot.In this paper,the network modeling of fJSP with AGV is an effective way to simplify the understanding and feature representation of the problem.In this paper,we propose a cooperative hybrid EA(ChEA)to solve the large-scale fJSP with AGV,whose goal is to minimize the makespan.The encoding and decoding process of fJSP is complex.It is simulated as the representation based on three-stage random key,it can prevent the algorithm from getting into the global optimal deadlock.In view of the complexity of large-scale problems,this paper uses the idea of "divide and conquer",and adopts an effective set based random grouping paradigm to combine variable space.The solution space is decomposed into small-scale variable space to achieve coevolutionary optimization;after comparing the performance of several evolutionary algorithms,in the global search part,particle swarm optimization algorithm(PSO)based on Gaussian distribution and local optimal individuals is used as the evolutionary algorithm;PSO algorithm realizes global search at the individual level to find a better solution,local search is to find the local optimal solution by deleting and inserting nodes on the key path of disjunctive graph.Compared with the local search based on the key path moving one operation,deleting two operations creates more space for inserting key operations.The optimization experiments of fJSP and fJSP with AGV are carried out by using the ChEA proposed in this paper.Experimental results show that compared with the most advanced algorithm,this algorithm has strong competitiveness,robustness and stability.