Evolutionary game decision-making method for high dynamic UAV swarm attack-defense considering uncertain factor influences

Unmanned Aerial Vehicle (UAV) swarm attack-defense scenarios have the characteristic of highly dynamic evolution, which consequently brings challenges for the existing decision-making methods. For highly stochastic processes and dynamic combat environments scenarios, we propose a UAV swarm attack-defense decision-making method based on uncertain factor influences evolutionary game. First, the attack-defense interest relationships among offensive UAV swarm, defensive UAV swarm, and base are analyzed, and a UAV swarm attack-defense game framework is established. Second, an improved replicator dynamic equation considering uncertain factors is proposed, and the stability of equilibrium points is further analyzed. Subsequently, situation assessment functions, which can evaluate the strategies chosen by the attack-defense both sides, are designed to adjust the payoff matrix of UAV swarm attack-defense game model. The strategy combination that reaches Nash equilibrium is further solved, so that the offensive UAV swarm and defensive UAV swarm can adaptively select the strategies. Finally, experimental simulations are conducted to verify the impact of random disturbances of varying intensities on the attack-defense decision-making process. Additionally, the proposed decision-making method can realize the adaptive selection of both attack and defense strategies. The video (Size: 15.2 MB) of high dynamic UAV swarm attack-defense based on the proposed method has been uploaded to https://github.com/mengmeng1023/UAV-Swarm-Attack-Defense .