A Hybrid Genetic Algorithm for No-Wait Job Shop Problems

• Main Authors: Han-Chiang Huang, 黃漢強
• Other Authors: Chao-Hsien Pan
• Format: Others
• Language: en_US
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/06501267320444578454

碩士 === 國立臺灣科技大學 === 工業管理系 === 91 === A no-wait job shop (NWJS) describes a situation where every job has its own processing sequence with the technological constraint that no waiting time is allowed between operations within any job. A common example of a no-wait production process is the production line in a steel factory where after a slab of steel enters the furnace it has to continue with no-wait until all rolling operations are completed. A NWJS problem with the objective of minimizing total completion time is a NP-hard problem. This thesis proposes a hybrid genetic algorithm (HGA) to deal with this complex problem. A genetic operation is defined by cutting out a section of genes from a chromosome and treated as a subproblem. This subproblem is then transformed into an asymmetric traveling salesman problem (ATSP) and solved with a heuristic algorithm. Subsequently, this section with new sequence is put back to replace the original section of chromosome. The incorporation of this problem-specific genetic operator is responsible for the hybrid adjective. By doing so, the course of the search of the proposed genetic algorithm is set to more profitable regions in the solution space. The experimental results show that this hybrid genetic algorithm can both accelerate convergence and improve solution quality.