An Improved Moth-flame Optimization Algorithm For A Lot-streaming Hybrid Flow-shop Scheduling Problem

The Hybrid Flow-shop Scheduling Problem(HFSP)has been the focus of academic and manufacturing enterprises.The actual production system often encounters many kinds of largescale production scheduling problems,which involve the division of the lot of jobs and the selection of all sublots of production sequence and machine.Reasonable lot division and scheduling can effectively improve the production efficiency and reduce the high energy consumption caused by long time standby.In this thesis,single-objective and multi-objective HFSP problems based on lot-streaming are studied and solved by using the moth algorithm.The specific research work is as follows:In order to save energy consumption,a mixed integer programming model for lotstreaming HFSP is established.Two different decoding strategies are designed based on“minimum energy consumption”,and an effective improved moth-flame optimization algorithm(IMFO)is proposed to solve the problem.Cauchy variation is carried out for the best flame to avoid falling into local optimum prematurely,and cross and reverse operation is carried out for the poor flame to improve the quality of solution.Finally,the effectiveness and superiority of IMFO algorithm are proved by comparing with CPLEX and other six effective algorithms.A clustering-based moth-flame optimization algorithm(CMFO)is proposed to solve this problem for multi-objective lot-streaming HFSP,which is considered to minimize both maximum completion time and total energy consumption.The algorithm uses an adaptive adjustment strategy of flame number decreasing with population recursively to enhance the global development and local search ability of the algorithm.K-mean clustering algorithm was used to cluster the population,and the excellent moth in each cluster is used as flames to guide the local searching ability of moths in this class.At the same time,Cauchy variation is carried out for the best flame in each classification to enhance the development ability of the algorithm,and cross and reverse operation is carried out for the poor flame to improve the quality of the population.The parameters of CMFO algorithm were determined by orthogonal experiment.Finally,the comparison experiments between the improved strategy and other three multiobjective algorithms proves that CMFO algorithm has great advantages in computing efficiency and solving quality.