基于MADDPG算法和群体捕猎行为的多无人机协同围捕决策方法OA

Utilizing UAVs for intercepting other UAVs is a cost-effective and attritional counter-UAV approach.In order to improve interception success rates,a UAV adopted by multi-UAV to round the ene-my UAVs up in cooperation is a practical counter-UAV strategy.On the basis of high-precision target de-tection information,in view of the problem that how to improve the distributed cooperative decision-mak-ing ability is still a critical factor in determining the engineering applicability of cooperative rounding-up,this paper proposes a multi-UAV cooperative rounding-up decision method based on the MADDPG algo-rithm,incorporating characteristics of collective hunting behavior.Both the kinematic model and Markov decision process model for the multi-UAV cooperative rounding-up problem are established.The MAD-DPG algorithm is enhanced through the implementation of a stochastic exploration strategy and priori-tized experience replay mechanism to improve exploration capability and learning efficiency.The simula-tion results demonstrate that the proposed algorithm can effectively learn cooperative rounding-up strate-gies,and is superior to the two baseline algorithms in key metrics in rounding-up time and at success rate.