Weapon-target Assignment of Ballistic Missiles Based on Q-Learning and Genetic Algorithm
Quan Cheng, Derong Chen, Jiulu Gong
Abstract
There are two methods to handle the weapon target assignment (WTA) problem: treat it as a single-agent multi-step decision-making problem or a multi-agent single-step decision-making problem, but both have the problem of low computational efficiency. In order to improve the computational efficiency of the algorithm, we combine above two methods and propose a two-stage optimization algorithm based on Q-Learning and genetic algorithm (QL-GA). We first use Q-Learning with high exploration efficiency to explore excellent solutions through a few iterations. Then, we use the optimal solution explored by Q-Learning as the initial population of the genetic algorithm (GA), and use GA to find the optimal solution with a small population size. The experimental results show that the average running time of the proposed algorithm is decreased by 2.96s and 13.42s compared with Q-Learning and GA under the same experimental background, which verifies that our algorithm has high computational efficiency. At the same time, this algorithm also has better performance in global optimality.