Solving the Job-shop Scheduling Problem with a Simple Genetic Algorithm

The job-shop scheduling is concerned with arranging processes and resources. Proper schedules are very important for the manufacturers, but can cause serious problems because of the enormous solution space. Pressure from the competitive enterprises is the main reason why time is becoming one of the most important success factors. Scheduling tools allow production to run efficiently. The goal in this paper is the development of an algorithm for the job-shop scheduling problem, which is based only on genetic algorithms. Our intention is to prove, that even a relatively simple genetic algorithm is capable for job-shop scheduling. The effectiveness of the algorithm is demonstrated by solving practical problems. The first problem consists of 10×10 instances (10 jobs and 10 machines) and the second one of 20×5 instances (20 jobs and 5 machines). The scheduling efficiency is measured by the time required to complete all jobs (makespan). In case of the first and the second problem, the best obtained solution (i.e., deviation from optimal solution) was 1.2 % and 4 %, respectively. (Received in March 2009, accepted in June 2009. This paper was with the authors 1 month for 1 revision.)