Research On Multi-constraint Production Scheduling Optimization Technology For Flow-shop

Under the national strategy plan of Made in China 2025 and Industry 4.0,Intelligent manufacturing has been the main direction of development for the manufacturing industry,and intelligent information management systems have been promoted in manufacturing enterprises of different industries and sizes.As a part of the intelligent information management system,MES is one of the hotspots of intelligent manufacturing research.This dissertation studies the optimization technology of the multi-constraint flow shop scheduling problem and its application in MES.First,based on the workshop status and processing flow of the enterprise,this dissertation designs the system architecture and workflow of MES,and divides the service functions of the system into three functional modules as user management,production management and basic management.To solve the flow-shop scheduling problem with sequence dependent setup times,the modified estimation of distribution algorithm was proposed.Matrix array is employed as the probability model to denote the relative information about jobs in this algorithm.Meanwhile,the updating mechanism is improved to adapt to the new probability,and an automatically adjusting method is employed to improve the global optimization ability.Besides,the modified estimation of distribution algorithm is employed to solve the limited-buffer flow-shop scheduling problem with sequence dependent setup times,compared with other algorithms,the simulation results demonstrate the effectiveness of the modified estimation of distribution algorithm.To solve the optimization of the limited buffer capacity problem in non-blocking scheduling,this dissertation proposes a modified branch and bound algorithm.Based on a variable step length branch method,this algorithm improves its searching efficiency.When the modified estimation of distribution algorithm is combined with the modified branch and bound algorithm,it is successful to solve the limited-buffer flow-shop non-blocking scheduling problem with sequence dependent setup times.Meanwhile,the effectiveness of the modified branch and bound algorithm is demonstrated.Finally,a backstage management system with B/S network structure mode by using Spring+Spring MVC+Mybatis framework is built in this dissertation.As a prototype system of MES,the system has three functional modules: user management,production management and basic management.The optimization algorithms of the flow shop scheduling with multiple constraints is embedded in the production management functional module.The system realizes the intelligent management of the comprehensive information in workshops.There are important theoretical value and engineering application value for improving the production efficiency and the information management level of manufacturing enterprises.