Research On Integrated Optimization Of Cell Reconstruction And Scheduling For Semiconductor Manufacturing

With the development of information technologies such as 5G,Artificial Intelligen ce and the Internet of Things,the demand for semiconductor products has expanded dra matically.At present,semiconductor manufacturing presents a mixed production mode of large-scale,small batch and customization,which brings great challenges to the prod uction management and coordination control of enterprises.The cellular capacity limit ed flexible flow shop(C-CLFFS)integrates the characteristics of traditional hybrid flow shop and cell manufacturing system.It has higher production flexibility and can effecti vely solve the problem of production resource allocation caused by demand uncertainty,so as to realize rapid response to customer demand.Therefore,this paper focuses on the assembly and test production of semiconductor to carry out the research on cell reconst ruction and cross-cell production scheduling with multi-resource constraints under unce rtain market environment.For the optimization of cross-cell production scheduling with stable demand,this p aper considers the machine sharing,the product delivery time difference and resource c onstraints of human,transporter and machine,establishes a mathematical model to mini mize tardiness batches,transportation time,workers load extreme deviation and materia l handling time.In order to solve the model,a two-layer INSGA-HLS algorithm was de signed based on the standard NSGA-III algorithm that included a mixed initialization m ethod,improved genetic operators,a mixed local search method and an elite storage me chanism.In the case of T factory,the parameter setting,operation improvement and ove rall performance of INSGA-HLS algorithm are analyzed firstly;then,the effects of due date difference and constrained resources on scheduling indexes are explored.It is prov ed that considering the difference of due date,cross-cell production,difference of worke rs and quantity limitation of transporters are helpful for enterprises to quickly meet orde r demands and balance resource allocation.To solve the problem of unreasonable resource allocation caused by demand fluctu ation,the joint optimization of cell reconfiguration and cross-cell production scheduling is carried out.This study further introduces the dynamic cell layout,considers the const raint effect of various resource initial states on production scheduling,and establishes th e integrated optimization model of cell reconfiguration and cross-cell scheduling under demand uncertainty.In order to solve the model,a hybrid cell reconstruction mechanism is embedded on the basis of the two-layer coding INSGA-HLS algorithm to design a th ree-layer coding based INSGA-HLS algorithm.In the case of T factory,comparing with cross-cell production scheduling under the same initial resource state,it is proved that t he research on integrated optimization of cell reconfiguration and cross-cell scheduling i s helpful to realize reasonable allocation of multiple resources,respond to market dema nd quickly and improve workshop production efficiency;then,it is verified that conside ring the initial state of resources can further ensure the rationality of resource allocation and the effectiveness of the production scheduling scheme.