Fuzzy Distributed Two-stage Hybrid Flow Shop Scheduling Based On Teaching-learning-based Optimization Algorithm

Under the background of deepening economic globalization and the continuous implementation of Chinese enterprises’ "going global" strategy,mergers and acquisitions,cooperative production and multi-plant production are becoming increasingly common among enterprises,making distributed manufacturing a common and important manufacturing mode.Like centralized manufacturing processes,distributed manufacturing processes are inevitably uncertain.Uncertainty is an important characteristic,which is difficult to be ignored in actual manufacturing process.As an important part of distributed manufacturing,distributed scheduling research should also be closely combined with the uncertainty of manufacturing process,so as to enhance the application value of research results and provide theoretical and methodological support for efficient distributed manufacturing.In this paper,Sequence dependent setup time is considered.SDST)fuzzy Distributed Two-stage Hybrid Flow Shop Scheduling Problem(FDTHFSP),Two new teaching optimization algorithms are designed for single objective optimization and multi-objective optimization respectively,and the effectiveness of the algorithm is verified by experimental simulation.The main research work of this paper is as follows:(1)This paper introduces the research background and significance of distributed hybrid flow shop scheduling,summarizes the research status of distributed hybrid flow shop scheduling at home and abroad and the application of TLBO algorithm in production scheduling,and briefly summarizes the shop scheduling theory,intelligent optimization algorithm and reinforcement learning Q-learning algorithm.(2)A Q-learning-based teaching-learning-based optimization(QTLBO)algorithm is proposed to minimize the maximum completion time for SDST FDTHFSP.The algorithm includes teaching stage,learning stage,teacher self-study stage and student self-study stage.Q-learning algorithm consists of 9 states,4 actions,rewards and action selection strategies,which is used to dynamically adjust the TLBO algorithm structure.A large number of experimental and computational results show that QTLBO has a strong search advantage in solving single target FDTHFSP.(3)A Diversified teaching-learning-based optimization(DTLBO)algorithm is proposed for FDTHFSP,which takes SDST into account and takes maximum completion time and total consistency as its objective.The algorithm divides all classes into two classes according to the class quality.For each class,a genotype difference index is proposed by using different search times and the combination of global search and neighborhood search.According to this index,temporary classes with multiple teachers are constructed and temporary class search is realized.Experimental results show that DTLBO is a competitive method for solving multi-objective FDTHFSP.