Hybrid Genetic Algorithm For Solving Production Scheduling Problem

The job shop scheduling problem is a well-known NP problem,which is limited by various conditions.As the scale of the problem increases,the difficulty of finding the optimal solution increases.It is a difficult combination problem.Limited by the various constraints of the actual production environment,how to effectively arrange the processing order of each part will directly affect the production efficiency.The appropriate production scheduling algorithm can correctly and effectively plan the enterprise resources.Arranging the processing order and processing time of the workpiece rationally and using existing resources to meet the basic requirements of production scheduling properly by optimizing production scheduling instructions,in order to obtain the optimization of total production time,which has important theoretical significance for the actual production of enterprises.In this paper,aiming at optimizing production schedule and shortening production cycle,this paper studies the batch scheduling problem of multi-process and multi-processing paths.Based on genetic algorithm and ant colony algorithm,a hybrid production scheduling algorithm involved global search and local search is proposed.First,this paper described the production scheduling problem in detail,and described the characteristics of each type of scheduling.It briefly introduced the principle and basic framework of genetic algorithm and ant colony algorithm and also deeply understood genetic operators and parameter selection of genetic algorithm.At last,it introduced the pheromone updating mechanism of ant colony algorithm.Further on the basis of the production scheduling problem,the mathematical model is abstracted.According to different parts of the same machine and different parts of the same part,the corresponding processing time and waiting time are respectively obtained.At the same time,the genetic algorithm is improved by genetic algorithm.A dynamic genetic operator based on iteration number is proposed,which further enhances the convergence performance and search ability of genetic algorithm.This paper improved the optimization method of ant colony algorithm and the ability to find excellence of ant colony algorithm.Then the two algorithms are mixed,the genetic algorithm generates the initial population,and the improved coding conversion makes the information carried by the initial population more perfect,provides the initial pheromone for the ant colony algorithm,and performs the smaller granular fusion solution of the two algorithms.The most appropriate combination of time,mutual guidance to update the global optimal solution,so that the total production cycle is minimal.The two algorithms are fused to overcome the problem that the genetic algorithm has low search efficiency and is easy to fall into local optimum,and the ant colony algorithm is difficult to grasp the global information.Finally,the superiority of hybrid genetic algorithm is analyzed.The simulation of MATLAB simulation program is combined with the scheduling standard example to analyze the performance of different algorithms.The search efficiency of hybrid algorithm is improved,and the convergence performance of hybrid algorithm is improved.Time weights lead to different optimization options.The operation results of the system meet the requirements of production scheduling,which proves the feasibility and practicability of the hybrid genetic algorithm.