Shop Scheduling Problem Based On Improved Cuckoo Search Algorithm

With the development of economy and science,the competition between enterprises is more and more fierce.It is of great importance to reasonably schedule and workshop plan,which can improve the production efficiency,lower the production costs and make the delivery on time.Shop scheduling problem are typical combinatorial optimization problems.It's hard to get effective solution with traditional mathematical method for large scale problems.Cuckoo Search(CS)is an intelligent optimization algorithm which was proposed in recent years.CS has been applied to many fields,but it is seldom applied in shop scheduling problem.Therefore,in this paper,for two typical workshop scheduling problems:Flow-shop scheduling problem(FSP)and Job-shop scheduling problem(JSP),some improved CS are proposed and several examples are used to verified the effectiveness of these improved algorithms.The main research work is as follows:(1)In order to solve the problems of CS falling into local minima easily and converging slowly at the last stage.The hybrid Genetic algorithm and Cuckoo Search(GACS)has been proposed.In the proposed method,by integrating the selection and crossover theory of GA into the iteration process of CS.The results of different standard test functions verify that the improved algorithm not only can be used to solve continuous function optimization problem,but also can be used for solving the simple FSP.(2)For the solution of the unbalance workload in mixed-model automobile assembly line,an optimization model,in which the limit of violating assemble frequency is minimized,is constructed.The GACS is used to solve the model,the result shows that the production sequence which is got by improved CS has greatly increased the production of balance,and it provides the theoretical basis for production scheduling problems on car mixed flow assembly line.(3)For the problem of continuous function optimization,GPSCS was proposed based on gauss mutation and neighborhood search,and the construction of neighborhood employs PS method.For the problem of JSP,GHNSCS was proposed based on gauss mutation and neighborhood search,and the construction of neighborhood employs exchange,insert and Nowicki method.Some typical continuous test functions and JSP used for standard tests were respectively used to test these two improved CS.The simulation results illustrated the effectiveness of the improved CS algorithms.(4)Considering the Job-shop dynamic scheduling problem with emergency order inserting,the maximum completion time and scheduling stability are set as optimization targets.The method of weighting is used to deal with multi-objective problem.In order to solve this problem,the GHNSCS is used to solve the problem.