Huang, How-WeiHow-WeiHuangLu, Ching-HuChing-HuLuLI-CHEN FU2020-05-042020-05-042007https://scholars.lib.ntu.edu.tw/handle/123456789/489084https://www.scopus.com/inward/record.uri?eid=2-s2.0-44449095212&doi=10.1109%2fCOASE.2007.4341833&partnerID=40&md5=91f5be6b348cb0c1c9af099cb46fccd8In this paper, we focus on the lot dispatching and vehicle dispatching problems in a 300mm wafer fab. We propose a two phase lot dispatching and vehicle dispatching integrated algorithm. The first phase is the offline optimization phase and the second phase is the online dispatching phase. In the offline optimization phase, we focus on the lot dispatching problem. We propose a scheduler which can search for appropriate weights of several rules and combines them into one mixed rule. Genetic algorithm (GA) is used to And the optimal combination of these heuristic rules. And the mixed rule is used in online dispatching phase. In the online dispatching phase, we focus on the vehicle dispatching problem. We take the real time traffic information into consideration and we apply Markov decision model to model the estimate of the cost of traversing a congested edge. Compared with the traditional shortest path method that searches for the shortest distance path, in our traffic model we search for the path with least expected travel time. © 2007 IEEE.Decision making; Genetic algorithms; Markov processes; Online systems; Optimization; Congested edge; Heuristic rules; Vehicle dispatching integrated algorithm; Schedulingconference paper10.1109/COASE.2007.43418332-s2.0-44449095212