| Seru seisan is a flexible manufacturing system,which reversed from traditional assembly line.It replaces expensive aumation equipment with simple tools and thinks highly of multiskilled workers.Because seru system can achieve efficiency and flexiblility at the same time,under present environment of volatile market and fierce competition,more and more practitioners and researchers have taken note of seru seisan.Unfortunately,current research about seru is not sufficient enough,and many problems remain to be solved.Hence,in this paper,three main problems are discussed.Firstly,a seru loading problem with non-zero setup time is considered,the setup time is supposed to be a value that is in direct proportion to the number of different operations between two adjacent products.A heuristic algorithm based on enumerating all the product allocation results is proposed.Secondly,a seru loading problem with worker assignment is taken into account,the worker assignment decision is made before product allocation decision,and a heuristic algorithm by enumeration is also designed for sloving this problem.Finally,a seru loading problem allowing lot-splitting is discussed in this paper,and the worker assignment decision is still under consideration.Because the worker assignment decision is made by first line managers while the product allocation decision is made by production planning department,the problem presents the characteristic of hierarchy.In that case,a bi-level model is then formulated.Considering the bi-level programming problem is NP-hard,an SA-GA algorithm(simulation annealing and genetic algorithm)is designed to solve it.The GA is to solve product allocation problem under certain worker assignment,then the optimal product allocation result is fed back to upper level model,and SA is for solving upper level model of worker assignment.The effectiveness and the scalability of the SA-GA algorithm are verified by numerical examples,and the results are compared with a GA-GA algorithm. |
PDF Full Text Request