Research On The Production Scheduling Problems Based On Intelligent Optimization Algorithms

Manufacturing industries such as iron and steel industry,chemical industry,machinery industry,and so on are the pillar industry of the national economy.In recent years,with the development of economic globalization,manufacturing enterprises are facing more and more pressures in market competition,resources and environment.Production scheduling is the key component of production management in manufacturing enterprises.In view of the complexity of production scheduling problems,how to use advanced modeling and optimization methods to further improve the scheduling quality so as to help the enterprises to improve product quality and production efficiency,reduce production cost and energy consumption,has always been the opportunity and challenges for the current academic and manufacturing enterprises.Therefore,the research on the production scheduling problems has important theoretical significance and application value.This dissertation was focused on the typical production scheduling problems commonly existing in manufacturing industries,and developed several improved intelligent optimization algorithms for them.These researches contain not only theoretical discussion in problem modeling and solution methods,but also their applications in practical production scheduling problem.The detailed contents of this dissertation are as follows:(1)For the two-agent single machine scheduling problem,the actual processing time of a job from both the first agent and the second agent is modeled as a linearly increasing function of its starting time based on practical condition that bigger production cost is occurred if a job has a long waiting time for produciton.The objective of this problem is to minimize the total weighted number of tardy jobs of the first agent subject to the condition that the maximum lateness of the second agent is allowed to have an upper bound.A tabu search algorithm was developed to obtain near optimal solutions for large size problems,as well as a branch-and-bound algorithm for small size problems.(2)For the permutation flowshop scheduling problem,an improved genetic algorithm based on genetic algorithm and local search was developed.In the local search,a novel adaptive negiborhood was proposed,whose size could be dynamically adjusted during the search process so as to improve the exploitation ability.Computational results based on benchmark problems illustrated that the proposed genetic algorithm based on new neighborhood is superior to the other particle swarm opitmizatioin algorithms in the literature.(3)For the re-entrate permutation flowshop scheduling problem,the actual situation in the machinery manufacturing process that a job should be processed for several times on one machine is considered.This situation is different from traditional permutation flowshop scheduling problem in which a job can be processed only once on each machine.For this problem,an adaptive memetic algorithm in which the negibhorhood size can be adaptively adjusted in the local search was developed.(4)For the multi-objective permuation flowshop scheduling with sequence-dependent setup times,a multi-objective iterated local search(ILS)was proposed by extending the classical single objective ILS to multi-objective optimization.In the local search of the proposed multi-objective ILS,a multi-objective variable depth search(VDS)based on dynamic neighborhood was developed so as to improve the search efficiency by achieving a balance of the exploration and exploitation.Computational results based on benchmark problems illustrated the efficiency of the proposed multi-objective ILS algorithm.Finally,the research works were summarized,and further works needing to be done and application prospects of the production scheduling problems in the manufacturing industry were proposed and discussed.