Diversity Improved Genetic Algorithm for Weapon Target Assignment

There are many optimization problems in military applications, among which the weapon target assignment (WTA) problem is the most typical and the most widely studied problem. Plenty of evolutionary algorithms-based methods are studied for resolving it. However, the quality of the solutions of WTA still has a lot of room for improvement. We propose a prominent method called diversity genetic algorithm (DGA) which has three significant components to handle WTA. A hybrid crossover strategy combining two operators is introduced to improve DGA’s exploration performance. Lévy flight mutation is used to control the mutation percentage of offspring chromosomes, which could improve DGA’s exploitation. Besides, an enhanced mechanism is put forward based on the fitness of best solutions and Logistic chaotic mapping, which balances the performance of DGA. Five representative algorithms and twelve classical benchmark testing instances are adopted to evaluate DGA. Experiment results indicate that DGA has superior ability and suitable time cost.