Flexible job shop scheduling with novel imperialist competitive algorithm

In this paper we propose an improved imperialist competitive algorithm (ICA) for FJSP scheduling problem with the objective of makespan minimization. The proposed algorithm has been improved by a hybrid of ICA with Simulated Annealing algorithm (SA) after diagnosing of standard ICA disadvantages. The result confirms the improvement in optimization performance of novel ICA in comparison with standard ICA and also demonstrates that the novel ICA is efficacious algorithm for FJSP. Since ICA operates in continuous domains, we present a new method to exchange the continuous domains to discontinuous domains. Appropriate design of the parameters increases algorithm efficiency; therefore we calibrate the parameters of this algorithm using the Taguchi method