A bi-population immune algorithm for weapon transportation support scheduling problem with pickup and delivery on aircraft carrier deck

Abstract

The weapon transportation support scheduling problem on aircraft carrier deck is the key to restricting the sortie rate and combat capability of carrier-based aircraft. This paper studies the problem and presents a novel solution architecture. Taking the interference of the carrier-based aircraft deck layout on the weapon transportation route and precedence constraint into consideration, a mixed integer formulation is established to minimize the total objective, which is constituted of makespan, load variance and accumulative transfer time of support unit. Solution approach is developed for the model. Firstly, based on modeling the carrier aircraft parked on deck as convex obstacles, the path library of weapon transportation is constructed through visibility graph and Warshall-Floyd methods. We then propose a bi-population immune algorithm in which a population-based forward/backward scheduling technique, local search schemes and a chaotic catastrophe operator are embedded. Besides, the random-key solution representation and serial scheduling generation scheme are adopted to conveniently obtain a better solution. The Taguchi method is additionally employed to determine key parameters of the algorithm. Finally, on a set of generated realistic instances, we demonstrate that the proposed algorithm outperforms all compared algorithms designed for similar optimization problems and can significantly improve the efficiency, and that the established model and the bi-population immune algorithm can effectively respond to the weapon support requirements of carrier-based aircraft under different sortie missions.