Improving Heuristics and the Hybrid Genetic Algorithm for minimizing the maximum completion time of the Wafer Probing Scheduling Problem (WPSP)

• Main Authors: Yu-Lin Tsai, 蔡育燐
• Other Authors: Wen-Lea Pearn
• Format: Others
• Language: en_US
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/f5vh36

碩士 === 國立交通大學 === 工業工程與管理系所 === 92 === The wafer probing scheduling problem (WPSP) is a practical version of the parallel-machine scheduling problem, which has many real-world applications including the integrated circuit (IC) manufacturing industry and other industries. WPSP carries the objective to minimize the total machine workload, which might lead to unbalanced workloads among the parallel machines and be unaccepted for the shop floor supervisors. Therefore, we consider WPSP with the objective to minimize the maximum completion time and formulate the WPSP with minimum makespan as an integer-programming problem. To solve the WPSP with minimum makespan effectively, we proposed the improving heuristics, which add the expected machine load into savings and insertion algorithms for solving problems repeatedly. Besides, we also provide hybrid GA including initial population by WPSP algorithms and sub-schedule preservation crossover to solve the considered problem. To evaluate the performance of the two proposed approaches under various conditions, the performance comparison on a set of test problems involving four problem characteristics are provided. The computational result shows that improving heuristics are better than hybrid GA in scheduling solutions and velocities of WPSP with minimum makespan. When hybrid GA is using initial population by improving heuristics, it can make further improvement for the best solution of improving heuristics in some situations.