Research On Multi-objective Dynamic Scheduling Optimization In Flexible Job Shop

The flexible job shop is a rigid device that is currently used by many manufacturing plants for lean production,agile manufacturing,and mass customization.The improvement of many production modes is based on the flexible job shop.Flexible job shop scheduling for intelligent rationalization is the support for the transformation of its production model.Through scheduling scheduling of production tasks on the shop floor,production goals can be maximized.This paper analyzes the research status of objective functions,constraints and research methods of both job shop scheduling problems and flexible job shop scheduling problems.Based on this,for the parallel scheduling problem of the flexible job shop,on the basis of the "bottleneck ideology" of the constraint theory,a heuristic algorithm of "minimum pre-production" was proposed after constructing "virtual orders" based on historical data,and combining machine tools with products from the perspective of balancing machine load rates.Then,a static flexible job shop production scheduling target with maximum completion time,maximum machine load,maximum advance/the cost of tardiness penalties,and minimum load of machine with maximum load was proposed.A matrix decoding method was introduced to double code,crossover,mutate and correct the Chromosomal gene under the constraints of the production environment with machine constraints and process constraint,and after modifying fitness function,an improved genetic algorithm is used to solve the optimization.Finally,the objective function of dynamic flexible job shop scheduling optimization is set supported by the research results of parallel machine scheduling and static flexible job shop scheduling based on historical data,and through rolling window technology,coding according to the process priorities in the re-scheduling window,and combining event and cycle-driven strategy to generate generator,we conduct flexible job shop dynamic scheduling based on improved genetic algorithm.All of the above studies used data cases for simulation experiments in the end,and was proved effective and feasible of their methods in the optimization process based on the calculation results.