A practical approach for job-shop scheduling problems using genetic algorithm

Proposes an intuitive, yet efficient approach, which is based on a genetic algorithm (GA), for solving job-shop scheduling problems. Aiming at practical use in real manufacturing, the approach is designed in such a way that it elegantly simulates the actual organization of job shops and is efficient in finding a good schedule. It has been proved to perform better than other heuristic methods with a number of established job-shop problem instances. In the meantime, due to its domain-independent design, it can be easily extended to address such complex constraints as non-zero ready time, due time, sequence-dependent setups, machine downtime, etc. Also, it is capable of system objectives other than makespan, such as cost. A discussion of such extensions and corresponding conclusions are given in this paper.