Comparison of scheduling efficiency in two/three-machine no-wait flow shop problem using simulated annealing and genetic algorithm

This paper considers two/three-machine no-wait flow shop scheduling problems with makespan minimization. Inherited with the NP-hard problem nature, this scheduling problem is solved by using a Simulated Annealing (SA) and Genetic Algorithm (GA) instead of mathematical programming. These two well-known search algorithms are often used to solve complex combinational optimization problems. In order to compare the performance of the two algorithms, we use solution quality (under the same computational time) and computation efficiency (under the same solution quality) as the measuring criteria. From the example problem, we found that SA is superior to GA in both solution quality and computation efficiency under identical terminating conditions. The performance of SA and GA decreased upon increasing the number of jobs. Our main contributions are to compare SA and GA under limited resources, which would be more consistent with the real world.