A UAV Pursuit-Evasion Strategy Based on DDPG and Imitation Learning

Algorithm 1

The UAV pursuit strategy using DDPG.

Training algorithm for UAV strategy based on DDPG

initial experience pool D with memory size M

initial the Eval networks of Actor network and Critic network: and

forepisode=1 to MaxEpisodedo

 initialize OU-Noise

 initialize the state of pursuit-UAV and evasion-UAV in set range randomly,

obtain the initial state of simulation environment

 fort=1 to MaxStepdo

  select action of pursuit-UAV where is the action constraint processing process

  select maneuver strategy for evasion-UAV

  input the control signal into the UAV integrate to get the next state of UAV, and calculate the environment state

  obtain the immediate reward from the environment

  store experience sample in D

  randomly sample form D to get a sample set of BatchSize

  update the Eval network parameter of the Critic

  update the Eval network parameter of the Actor

  update the Target network parameters and of Critic network and Actor network by (20)

  if the episode end condition is satisfied, break

 end for

end for