Abstract
This paper describes the implementation of an ant colony algorithm (ACA), applied to a combinatorial optimization problem called job shop scheduling problem (JSSP). At first, a rather good solution is generated in negligible computation time and then, the trail intensities areinitiated based on this solution. Moreover, the trail intensities are limited between lower and upper bounds which change dynamically in a new manner. It is noteworthy that in initializing, updating as well as limiting the trail intensities, the goal is to guide the search towards the neighborhood around the best solution found. This paper outlines the algorithm’s implementation and performance when applied to job shop scheduling. The computer simulations on a set of benchmark problems are conducted to assess the merit of the proposed algorithm compared to some other heuristics in the literature. The solutions were of good quality and demonstrated the effectiveness of the proposed algorithm.