Research On Dynamic Maintenance Strategy Of Semiconductor Wafer Fabrication

Semiconductor production line is a class of complex manufacturing system with re-entrant being its feature.Semiconductor manufacturing system has the characteristics of product diversity,high integration and high cost.In recent years,the semiconductor manufacturing industry is developing rapidly and its competition is fierce.Equipment can be maximized value by reasonable maintenance policies.As a result,the earnings and competitiveness of companies will be improved.In this article,the MDP(markov decision process)model about production and maintenance of the above system is established.Considering the variability in the choice of maintenance behavior and the randomness of state transfer,the dynamic maintenance strategy of the equipment of semiconductor production line is studied.Finally,the integration of maintenance time and maintenance behavior that maximizes the benefit is obtained.In this article,the dynamic maintenance strategy is applied to a typical semiconductor manufacturing process,which is Mini-fab model.The scheduling rules of each unit are analyzed in further study in the mod el.The coupling problem between production and maintenance is optimized,and the optimal dynamic maintenance strategy is obtained.Firstly,this article introduces the research background and significance.And it analyses of the research status at home and abroad from the aspects of industry and academia.The research content and theoretical framework are described in detail.Secondly,this article expounds the related theory of reliability analysis and maintenance,as well as the type and the index of production scheduling,which provides theoretical support for the following research.Then,considering the variability in maintenance behavior choice and the randomness of state transfer,the mathematical model of dynamic maintenance strategy of semiconductor equipment based on MDP model is established.The MDP model is used to simulate the state transition in the process of maintenance of equipment.And the efficiency of equipment is taken as the objective function,which obtains a comprehensive plan for the maintenance of time and behavior.Based on the characteristics of the mathematical model,the particle swarm optimization algorithm with genetic crossover factor is introduced to obtain the optimal dynamic maintenance strategy.Finally,the dynamic maintenance strategy is applied to the Mini-Fab model,and the production process of the system is further considered.It researches the joint selection of maintenance and processing and the optimal scheduling rules of each unit,so as to obtain the integrated strategy of production scheduling and dynamic maintenance plan.