联合火力打击战法策略的智能进化方法研究

Research on Intelligent Evolution Method of Joint Fire Strike Strategy

提出了基于智能算法的火力打击战法策略优化方法。该方法突破常规人工智能算法的神经网络和强化学习范式结构,利用智能体和种群的双层构造和对抗推演中的反馈评分实现智能体的智力提升,通过群体进化和复杂系统涌现达成战法策略的优化。实验结果表明:总体效果优于常规的遗传算法和对抗进化算法,设计思想能够迁移到诸多研究领域,具备应用性和扩展性。

This paper put forward an optimization method of firepower attack strategy based on intelligent algorithm.This method breaks through the neural network and reinforcement learning paradigm structure of conventional artificial intelligence algorithms,and makes use of the double-layer structure of agent and population and feedback score in adversarial deduction to improve the intelligence of agents,and tries to optimize the tactical strategy through swarm evolution and complex system emergence.The experimental results show that the proposed method achieves the complex evolution of task planning through the emergence effect,and the overall effect is better than the conventional genetic algorithm and adversarial evolutionary algorithm.The algorithm design idea in this paper can be transferred to many research fields and has certain application and expansibility.