Design And Implementation Of Multi-objective Flexible Workshop Production Scheduling System

The development of China's economic is in a stage of high-speed.With the concept of industry 4.0 put forward,manufacturing industry,which is an important part of China's economy,must be changed to advanced intelligent manufacturing production mode as soon as possible.In this transformation,the challenge is that most traditional manufacturing enterprises do not have intelligent and advanced technologies,which makes the transformation of enterprises difficult.Cable production workshop is a typical example of traditional manufacturing industry in our country.At present,this kind of enterprises are still in the traditional management mode based on experience,which makes the management subjective and random,and cannot be supported by accurate data to do workshop work.In case of emergency,they need to report and reissue orders layer by layer.At the same time,the production scheduling problems faced by production workshops cannot be solved perfectly,thus causing a waste of resources.Therefore,the workshop production scheduling system studied and designed in this paper are important to help enterprises to enhance their competitiveness.In this paper,a real wire and cable production workshop is taken as the research object,and according to the problems existing in the enterprise and workshop,the corresponding solutions are suggested.The main content and innovations of this paper are as follows:(1)This paper analyzes the features of cable workshop production scheduling,types of workshop scheduling problems,classification of cable products and production processes,as well as the constraint conditions of machines,etc.,and establishes a suitable mathematical model for cable workshop scheduling problems by using mathematical constraint method.(2)According to the data problems in practical application scenarios,this paper designs a corresponding coding rules and introduces the concept of matrix coding.Due to the matrix coding,the computational complexity is greatly reduced.At the same time,due to the characteristics of chromosomes in genetic algorithm,a crossover operation method of row transformation and column transformation for the matrix coding and a method of random position mutation are proposed.Subsequently,the initial population and the new population generated by the crossover operation are merged,and the population size is expanded by means of data augmentation to take the advantage of the diversity of the population.Finally,the fitness function and the termination rules of the algorithm are designed.(3)In order to deal with the multi-objective optimization problems in the real production workshop,the common strategies of multi-objective optimization are analyzed first,and each optimization objective is formulated mathematically.Based on the single-objective scheduling optimization algorithm and to take advantages of NSGA_II algorithm,a multi-objective scheduling optimization algorithm based on Pareto sorting is proposed.A concept of repository is introduced which is to store the Pareto optimal solutions obtained by each generation separately,and these optimal solutions are sorted separately to find the optimal solution in the optimization.According to the disadvantages of NSGA_II algorithm,a new calculation method of crowding distance is designed to improve the population diversity.(4)According to a variety of unexpected situations that the real production workshop may encounter,this paper proposes a production delay prediction based on Gated Recurrent Unit networks.The prediction of the average production speed of the production workshop and the change trend makes it possible for people to take effective interference.(5)Finally,a corresponding production scheduling system is designed and implemented based on the demand of enterprises.