Anti-unmanned aerial vehicle task allocation method based on reinforcement learning

Abstract

Aiming at the problem that single interception equipment of a current anti-unmanned aerial vehicle system cannot effectively suppress unmanned aerial vehicles in temporary task allocation, the invention discloses an anti-unmanned aerial vehicle task allocation method based on reinforcement learning. The method comprises the following steps: initializing an improved DQN algorithm which refers to predicting a Q value by adopting a state at the current moment relative to a DQN algorithm; completing the training and optimization of the intelligent agent through an improved DQN algorithm, and storing the network parameters after the training of the intelligent agent is completed; inputting the unmanned aerial vehicle state information S into a reinforcement learning module, and outputting a suboptimal solution X, namely an initial allocation strategy, through reinforcement learning; optimizing the suboptimal solution generated by reinforcement learning through an evolutionary algorithm to generate an optimal solution of target allocation; and decoding the optimal solution to obtain an anti-unmanned aerial vehicle task allocation scheme. According to the method, the agent interception performance trained through the improved DQN algorithm is more accurate, and the task allocation is more efficient and applicable.

Description

technical field [0001] The invention belongs to the technical field of anti-UAV task distribution, and relates to an anti-UAV task distribution method based on reinforcement learning. Background technique [0002] In recent years, with the continuous development and improvement of technologies in the fields of communication and industry, the number of drones is experiencing explosive growth and has been widely used in both military and civilian fields. They are widely used in aerial photography, agricultural production, plant protection, express transportation, traffic monitoring, disaster rescue, surveying and mapping, power inspection and many other fields. Among them, the field of secure communication and attack detection has attracted special attention, and more and more researchers have begun to focus on it. [0003] At present, most countries in the world regard drones as traditional flying targets, and generally adopt traditional air defense weapon systems to ensure ...

AI Technical Summary

Problems solved by technology

At present, the all-element anti-UAV system that can systematically and effectively protect and suppress UAVs has not yet been put into practical application.

Images