Research On An Intelligent Scheduling Method For Mixed Line Production

With the improvement of the socioeconomic level people's quality of life,the change of market environment and user demand,the modern manufacturing mode has continuously evolved into a multivariety and small-batch mixed-line production one.In order to adapt to this situation,more and more manufacturing companies are applying mixed-line production mode in flexible job shops.Meanwhile,after the "Made in China 2025" strategy was put forward,the artificial intelligence-based shop scheduling method has become a new research hotspot.Therefore,it is of great significance to establish the model of flexible job shop scheduling problem model for mixed line production and to research on artificial intelligence-based shop scheduling methods for guiding modern enterprise production and implementing intelligent scheduling.In the first half of this paper,a flexible job shop scheduling problem model for mixed line production is proposed.In the latter part,a new workshop scheduling method is proposed after combining shop scheduling technology and reinforcement learning technology.The main research contents of this paper are as follows:Firstly,in view of the shortcomings in the current mixed line production research,a flexible job shop scheduling model for mixed line production is proposed.Aiming at the model,the corresponding multi-segment genetic algorithm is designed.Then,the multi-segment genetic algorithm is used to solve the flexible job shop scheduling model for mixed-line production to verify the correctness of the model.Secondly,in order to improve the intelligence of production scheduling,in connection with the shortcomings of current artificial intelligence scheduling methods,based on MDP,a reinforcement learning model in shop scheduling are designed,which are FJSSP-MDP model and FJSSP-MDP model for mixed production.Thirdly,for the designed reinforcement learning model of job-shop scheduling,a reinforcement learning based scheduling method and a disturbance processing mechanism are designed.Then,the scheduling analysis and case simulation are carried out to verify the feasibility of the proposed method.Finally,a prototype system for job-shop scheduling is designed.By running a scheduling case,the feasibility of the prototype system is verified.