| The introduction of Germany's "Industry 4.0" and "Made in China 2025" marks the arrival of the era of intelligent manufacturing with manufacturing as the core.The continuous improvement and change of the technical level and production process have made the competition among manufacturing companies intensified.Production scheduling is one of the core competitiveness of modern manufacturing enterprises,especially in semiconductor manufacturing enterprises.Due to the randomness of orders,the dynamics of production systems,uncertainty and imbalance,reasonable and efficient production scheduling becomes More difficult.As a kind of production scheduling,rule scheduling has the characteristics of simple operation,high flexibility,visibility and legibility.After combining with simulation,it can clearly and intuitively understand the workshop through discrete simulation technology and evolutionary algorithm optimization.The situation,and targeted improvement of production scheduling programs,in order to achieve the purpose of reducing the human and material costs of enterprises,and enhance the core competitiveness of enterprises.This paper focuses on the dynamic scheduling problem of semiconductor job shop,combined with rule scheduling and simulation,uses fuzzy logic theory to evaluate and decide the current workshop state,and dynamically adjusts the scheduling rules to the current better rules,thus optimizing the comprehensive performance of the workshop.Improve delivery time satisfaction,reduce the makespan,maximum flow time,mean flow time,total setup time,etc.under the constraints of satisfying customer needs and ensuring equipment load balancing as much as possible.Therefore,the main research contents of this paper include the following aspects:(1)Based on the integration and customization functions of Extendsim,the semiconductor shop scheduling simulation model is constructed.The simulation model is divided into functional module,control module and optimization module,and a rule base is built according to different target performance requirements.In the model,the functional module is used to simulate and analyze the production scheduling of the workshop;the scenario manager in the control module is used to change the operating environment and related parameters of the system,and the performance of each scheme in different environments are analyzed and compared;the evolution in the optimization module is used to solve the optimal scheduling rule base.(2)Aiming at the randomness and uncertainty in the dynamic scheduling of job shop,based on the rule scheduling,combined with the fuzzy logic decision theory,a multi-factor combination rule dynamic model(MFCRD)based on cycle and event-driven is proposed to realize the rolling of the job shop and real-time response to uncertain events.The workshop production is divided into multiple decision points according to the process.At each decision point,the system parameters such as delivery satisfaction,mean waiting time delay proportional coefficient,and setup time ratio are calculated as input variables of fuzzy logic decision.After the membership function discriminates the type and the output judgment of the fuzzy rule,the optimal scheduling rule in the period is finally output.(3)In the case analysis,the dynamic model and several classical scheduling rules were compared and verified from the perspective of the difference in the tightness of the due date and the occurrence of the machine failure,and the multiple plan was evaluated by the improved TOPSIS comprehensive evaluation method..The weights in the TOPSIS comprehensive evaluation method are obtained by the subjective weights in the AHP method and the objective weights in the entropy method.The subjective influence of the index importance and the diversity of the data are considered more comprehensively.Through the simulation and comparison analysis,the feasibility and effectiveness of the multi-factor combination rule dynamic model are verified. |
PDF Full Text Request