Extremal optimization for a dye vat scheduling problem

This paper models a dyeing scheduling process with sequence dependent changeovers for minimization of costs. To obtain an optimal vat arrangement, we propose an improved extremal optimization (EO) embedded a heuristic rule. In iterations of the EO, each product´s fitness is evaluated and a relative bad product is selected and reassigned. This reassignment impacts the related products that are in the same dye vat with the selected one and the related vats that contain the related products, so that all relevant products and vats are reassigned using the heuristic rule. The proposed improved EO exhibits good performance for solving the dye scheduling problem tested by the numerical experiments. The simulations on two groups of data in different size show that the EO gets better dye vat scheduling scheme in shorter time compared with genetic algorithm.