Research On Flexible Job Shop Scheduling In W Company

In recent years,military manufacturing industry has developed rapidly under the influence of economic and political environment and situation at home and abroad.Enterprises are facing multi-dimensional requirements and challenges such as order delivery schedule,customer satisfaction and low-cost sustainable development.Based on the systematic point of view,how to make full use of existing resources and arrange production efficiently and reasonably plays an important role in the survival and development of enterprises.Taking all these into consideration,this thesis takes a production workshop of W company as the research object,carrying out the following work:1.With the analysis of the current situation and problems of workshop,some improvement objectives were formulated.At the same time,on the basis of summarizing,analyzing and summarizing the research and development status of flexible workshop scheduling problem at home and abroad,the related basic scheduling algorithms and technologies were studied,and combined with case analysis,the algorithm principle,advantages and disadvantages and related program codes were mastered,which lays a foundation for later technical research work.2.Aiming at the multi-objective scheduling requirements of W company,a flexible job shop scheduling optimization technology based on improved genetic algorithm was proposed.The technology was carried out to take the the total completion time,the maximum delay time and the minimum total processing cost as the goal,with Pareto optimization.The two-stage coding and decoding method of equipment and process is designed,and the population initialization strategy based on the combination of part process time and equipment processing cost is proposed.At the same time,to rank the individuals and determine the parent population for crossover and mutation,an improved non-inferior frontier classification method is used.The improved elite retention strategy based on the external archive set is adopted to determine the new population,and the corresponding retention probability can be added to the individuals of different non-dominated levels according to the evolution of the population to generate the new population.Finally,the algorithm is tested using the actual production data of W company as an example.The experiment showed that the technology in this paper has obvious advantages over the traditional algorithm.3.With the problem of frequent emergency order insertion and heavy scheduling task in W company,an emergency order insertion scheduling optimization technology based on equipment fuzzy clustering was proposed.Combining the random failure rate and random maintenance rate factors of equipment to perform fuzzy clustering on the workshop equipment,this technology established the system equipment capacity gradient model,used the gray correlation analysis method to determine the priority of the order insertion order,and determined the insertion point of the order insertion order in the production site.On this basis,the improved genetic algorithm is used to optimize the scheduling.When the new population is determined by selection,crossover and mutation,according to the different priority of the inserted order or task,the processing equipment with excellent comprehensive ability is preferentially selected to generate a new population,so as to realize the fine matching and full utilization of equipment capacity and order urgency.The simulation experiment was carried out with the actual production and operation data of W company.The results showed that this technology can effectively solve the problem of efficient job shop scheduling in emergency order insertion environment.4.Based on the above work,the corresponding software module is developed by using the GUI development tool of Matlab software,which provides relevant tool software support for the job shop scheduling problem of W enterprise,and the software function is realized by combining the production data.Both the software function and its practical result has achieved its expectant goal.