Project Scheduling Using Hybrid Genetic Algorithm with Fuzzy Logic Controller in SCM Environment  

作　　者：Mitsuo Gen KwanWoo Kim Genji Yamazaki 

机构地区：[1]Graduate School of Information Production & Systems, Waseda University,Department of Intelligent Systems Tokyo Metropolitan Institute of Technology,Department of Intelligent Systems Tokyo Metropolitan Institute of Technology Kitakyushu 808-0135, Japan , Hino-city, Tokyo 190-0065, Japan, Hino-city, Tokyo 190-0065, Japan

出　　处：《Tsinghua Science and Technology》2003年第1期19-29,共11页清华大学学报（自然科学版（英文版）

摘　　要：In supply chain management (SCM) environment, we consider a resource-constrained project scheduling problem (rcPSP) model as one of advanced scheduling problems considered by a constraint programming technique. We develop a hybrid genetic algorithm (hGA) with a fuzzy logic controller (FLC) to solve the rcPSP which is the well known NP-hard problem. This new approach is based on the design of genetic operators with FLC through initializing the serial method which is superior for a large rcPSP scale. For solving these rcPSP problems, we first demonstrate that our hGA with FLC (flc-hGA) yields better results than several heuristic procedures presented in the literature. We have revealed a fact that flc-hGA has the evolutionary behaviors of average fitness better than hGA without FLC.In supply chain management (SCM) environment, we consider a resource-constrained project scheduling problem (rcPSP) model as one of advanced scheduling problems considered by a constraint programming technique. We develop a hybrid genetic algorithm (hGA) with a fuzzy logic controller (FLC) to solve the rcPSP which is the well known NP-hard problem. This new approach is based on the design of genetic operators with FLC through initializing the serial method which is superior for a large rcPSP scale. For solving these rcPSP problems, we first demonstrate that our hGA with FLC (flc-hGA) yields better results than several heuristic procedures presented in the literature. We have revealed a fact that flc-hGA has the evolutionary behaviors of average fitness better than hGA without FLC.

关 键 词：resource-constrained project scheduling problem (rcPSP) priority rule method (PRM) hybrid genetic algorithm (hGA) fuzzy logic controller (FLC) 

分 类 号：TP273.4[自动化与计算机技术—检测技术与自动化装置]