Research On Job Shop Scheduling Problems With Multi-Time And Its Engineering Application

With intensified competition of global market, customer demands are transformed to the personalization and diversification, enterprises increasingly concern on how to design a reasonable schedule scheme to shorten production cycle, reduce work-in-process inventory and deliver just-in-time. With the further research of Job-shop Scheduling Problem (JSP), the optimism of JSP with only processing time of job can't meet to the practical demand of job shop production activity, while the setup times, the time of job routing between machines, the time of machine occupied and failure time of machine have great impact to JSP. The JSP with multi-time is paid more and more attention by the researchers and technologists, and it is becoming the focus and emphasis of research.In this thesis, the multi-time elements are analyzed and studied, then the time model in production procedure is constructed, in which the unify definitions of the setup time, routing time and waiting time are given. Based on this, the JSP modeling with multi-time and the optimism object of minmax makespan is conceived and theoretically analyzed, and the GA algorithm is contrived to solve this problem. The test results show that the average improvement is 6.5% for setup time,9.1% for the routing time,6.9% for both setup time and routing for the performance to classical JSP, where the average improvement is more 2.6% for routing time than for setup time, the more 3.3% for sequence dependent setup time than for sequence independent setup time.Flexible job shop is extendibility of classical job shop, which involves in machine assigned and the sequence of operation, and the Flexible Job-shop Scheduling Problem (FJSP) is more complicated. Specially, the multi-time is more impact to FJSP, for the uncertainty of the setup time and routing time. On the base of the time model, the code presentation of GA algorithm for flexible job shop is discussed, and the GA algorithm with A/B code based the job coding operation and integer coding machine is devised to solve the FJSP, and the FJSP modeling with multi-time and the optimism object of minmax makespan is conceived and theoretically analyzed. The test results show that the average improvement is 17.4% for setup time,33.6% for the routing time,22.6% for both setup time and routing for the performance to FJSP, where the average improvement is more 16.2% for routing time than for setup time, the more 0.23% for sequence dependent setup time than for sequence independent setup time.The press shop is the one typical application of job shop in practical production, and the Press shop Scheduling Problem (PSP) usually is simplified as JSP or FJSP. The researches of PSP are not meeting the practical application for the setup time and routing time in production procedure are not considered. On the basis of above researches, according to the production character of the automobile press shop, especially, to overcome the problems of PSP as JSP or FJSP, for example, the setup time and routing time are hard to obtain, the PSP modeling with multi-time and the optimism object of minmax makespan is conceived and theoretically, the PSP based Operation Constraints and Parallel Machine Model (OCPMM) is conceived and the new hybrid GA algorithm is given to solve this problem. Result shows that the PSP based on OCPMM model has high application value for the parameters are less, less uncertainty, more easily obtained and more feasible.Based on the above research findings, the scheduling system of press shop is designed and developed for the MES system covering the automilbe four shops. This system is applied in practical production in press shop of Anhui Jianghuai Automile Co. Ltd, and is well-tried.