Design And Implementation In APS For Semiconductor Test Based On B/S

In recent years,China's semiconductor industry has maintained a rapid trend of development,all kinds of electronic products affect our lives profoundly.The semiconductor industry is a technology-intensive industry with high technical content,complex manufacturing processes and increasingly individualized customer needs,which means enterprises must require advanced management models and management concepts to enhance the competitiveness of enterprises.Advanced Planning and Scheduling(APS)is an informational system based on resource constraints that provides long-term planning and short-term scheduling guidance to enterprises.APS system can greatly improve the level of production and management of enterprises.And now APS system has become one direction of production management development for most manufacturing enterprises.Based on APS theory,this paper develops an advanced planning and scheduling system for a semiconductor test enterprise in China.One of the problems to be solved in this APS system is Job-shop Scheduling Problem in the semiconductor test shop.This paper summarizes the classification and characteristics of Job-shop Scheduling Problem,mainly studies the use of genetic algorithm to solve the classic Job-shop Scheduling Problem(JSP)and the Flexible Jobshop Scheduling Problem(FJSP).Also this paper introduces the steps of Genetic Algorithm in detail,and describes the coding method of chromosomes by using operation-based representation.In view of the situation that standard crossover and mutation operators are easy to produce infeasible descendants,the paper optimizes the crossover and mutation process so that it always produces viable solution descendants,which enhances the iterative performance of the algorithm.In view of the possible degeneracy of population caused by genetic manipulation in evolutionary process,this paper uses immune operator to optimize it,so genetic iteration always develops in a correct direction.In addition,this paper proves the feasibility of the genetic algorithm by experiment.In addition,through the research and analysis of the production process and production of the semiconductor test,this paper puts forward several problems that the enterprises need to focus on.The APS system is divided into functions of system management,equipment management,product management,production management,job scheduling management,integrated management and several other functional modules,and conducted a needs analysis of each module.At the same time,this paper provides some aspects of the non-functional requirements in the user interface,operating environment,scalability,operability,security and so on.In the aspect of system implementation,this paper adopts the B/S system architecture and follows the classic layered architecture model in the software field,using the Spring Framework as the application container and Spring Data JPA as the ORM framework to complete the data persistence.In addition,Velocity,AngularJS and other open source technology are used to design User Interface,and provides users a good using experience.Finally,the APS system will be deployed in the Tomcat.Through the actual operation of the system in the target enterprise,it shows that the system is fully functional and highly scalable.The APS system can be a good solution to solve Job-shop Scheduling Problem in semiconductor test industry,enhance the ability of sophisticated management of enterprises,and ultimately enhance the core competitiveness of enterprises.