Job Shop Scheduling Problem With Stochastic Processing Time And Fuzzy Due-date

Under the background of globalization, the manufacturing in China is expandingremarkably and the competition is more and more intensive. Using scientific methodsto reduce the production cost, increase the production e?ciency and customer servicelevel becomes consensus of supervisors. Manufacturing scheduling, as the basic pro-cedure of manufacture process, is becoming more and more important. Its theory andapplication is a hot topic in both academia and industry world. Traditional researchfocus mainly on deterministic problems, however, there are many uncertain factors inreal world, which make great impact to the e?ect of scheduling. Therefore, we shouldtake uncertain factors into consideration to get more e?ective and robust schedulingplan.In this thesis, we consider two kinds of uncertain factors, one is the processingtime and the other is the due-date. Past researches tend to distinguish stochastic andfuzzy factors, while we choose proper mathematica language to describe uncertainvariables, that is, processing time is stochastic and due-date is considered as fuzzy.Thus, we study the job shop scheduling problem with stochastic processing time andfuzzy due-date. The two ways are connected by cost related objective functions.The structure of this thesis is as follows. First, we review the background andliterature of job-shop scheduling problem, then we construct uncertain models on thefoundation of deterministic models. To solve the model, we design hybrid intelligentalgorithm based on genetic algorithm. Through special design of operations such as en-coding and crossover, we implement the ?ow of algorithms. Then parameters are testedand selected to make the algorithm robust and e?cient. At last, we use simulation ex-periment to compare hybrid intelligent algorithm to other five traditional schedulingmethods. The simulation result verified that the algorithm is e?ective.