Research And Simulation Of Flexible Job-shop Scheduling Optimization Based On Hybrid Algorithm

Manufacturing Execution System(MES)is a powerful tool for manufacturing enterprises to realize informationization and intelligence.Therefore it not only needs to have the function of controlling closed-loop information,but guide and optimize the production route of the work order to make quick decisions on production anomalies with the real-time information of the entire workshop production.At present,the design of scheduling module in the manufacturing execution system mainly pursues low computational complexity with simple algorithms(for example,the shortest delivery date algorithm),which often leads to the optimization of the scheduling scheme is not good enough.Therefore,this paper takes the typical flexible job shop in the manufacturing enterprise as the research object,and studies the efficient hybrid optimization algorithm to get the optimal scheduling solution of the work order.The main contents and results of the research in this paper are as follows:1.Design an improved genetic algorithm with Powell called GA-Powell algorithm,and select the minimum of maximal completion time or the minimum of maximal machine load as the single optimization goal.Firstly considering the chromosomal of genetic algorithm in the flexible job-shop scheduling problem is special,the traditional Powell optimization algorithm is improved to avoid the infeasible solution,which improves the robustness and search efficiency of the algorithm.Secondly the Kacem instances and the Brandimarte instances are used to test the GA-Powell algorithm,and compare them with other algorithms to prove the effectiveness of the algorithm.2.Considering the three optimization objectives of minimum of maximal completion time,total machine load and maximal machine load at once,an improved NSGA-III algorithm with tabu search called NSGA-III&TS algorithm is proposed.In addition,a judgment strategy of solution quality is designed with the Pareto dominant relationship between solutions to improve the performance of hybrid algorithm the improved tabu search algorithm having a new neighborhood structure.Finally the Brandimarte instances are used to test the NSGA-III&TS algorithm and the improved NSGA-III algorithm without the local search strategy,which is in order to prove the validation of the local search strategy;the Kacem instances and the DPdata instances are used to test the NSGA-III&TS algorithm,and compare them with the existing algorithm.As a result,the effectiveness of the NSGA-III&TS algorithm is demonstrated.3.Design and develop a flexible job-shop dynamic scheduling system,whose scheduling module is based on the hybrid algorithm studied in this paper and take a transmission equipment manufacturing enterprise in Nanjing as an example object in order to use it`s data to test the functions of the system.At last,the results show that: the system can re-schedule for uncertain factors to make the original scheduling scheme corrected in time,and ensure the production of the order.Last but not least,the final scheduling scheme is displayed in the form of Gantt chart,and it is interactive for users to hover and zoom through the mouse,which can acquire the specific production information of a certain block and the scheduling information within a certain period of time.4.Establish a flexible-job shop model with Plant Simulation,which adds a certain failure rate to the machine and increases the logistics transit time between different machines on the part processing path.Therefore the performance of the scheduling scheme of the GA-Powell algorithm in such cases is studied,and the advantages and disadvantages of the scheduling scheme with the NSGA-III&TS algorithm are compared.