Discrete Self-adaptive Sine Cosine Algorithm For The Permutation Flowshop Scheduling

The flow shop scheduling problem is also called the same-order job scheduling problem,which has been widely used in the manufacturing industry.As time goes on,the market economy has accelerated its pace of changes,and the manufacturing industry has also developed in the direction of globalization.The production with multiple varieties and small batches has greatly accelerated the production of distributed manufacturing systems,thus making the problems of distributed manufacturing scheduling become a hot topic in production today.The DPFSP is one of the most common distributed manufacturing scheduling problems.It primarily focuses on assigning jobs to appropriate factories and appropriate positions on appropriate machines,so as to optimize the corresponding objectives.Theoretically,the PFSP and the DPFSP are classical combinatorial optimization problems,which have been proved to be NP-hard.In real production systems,solving this type of scheduling problem could organize resources reasonably and improve the efficiency and the yield of the enterprise.Therefore,it is of great significance and practical value to study the flow shop scheduling problem.First,the principle,characteristics and workflow of the sine cosine algorithm are simply analyzed,and an improved strategy is proposed,which called the adaptive parameter mechanism.The performance of the improved sine cosine algorithm is proved by testing it on the function sets.Moreover,the computational results are compared to some frequentlyused algorithms to present the effectiveness and superiority of the proposed adaptive sine cosine algorithm.After that,the permutation flowshop scheduling problem is introduced.In order to minimize the maximum completion time,a corresponding job-based encoding and decoding strategy is developed.And the sine and cosine optimization algorithm is discretized to address the problem.A forward movement and reverse movement based on exchange are also designed.Mobile operation transforms the adaptive module of the algorithm into a mobile strategy based on probability.For improving the fitness and diversity of the initial population,the initial population is generated by various approaches.In the existing solution space,the local search algorithm based on insertion is used to further optimize each solution in the population.Through testing the algorithm on instances with different scales and comparing the experimental results to that of other algorithms,the performance of the decentralized SCA is validated.Finally,this paper works on the distributed permutation flowshop scheduling problem,with the goal of minimizing the maximum completion time.A two-vector coding and decoding strategy is designed for this problem.Each solution contains two vectors: factory selection and job sequencing.For the factory selection code,two movement strategies of direct forward movement and direct reverse movement are developed.A local search procedure based on key factories is used to optimize the order of jobs in the key factory and improve the fitness of each individual in the current population.By performing the proposed approach on small-scale large-scale instances,computational results are compared to that of other algorithms to validate the effectiveness of the proposed algorithm.