Distributed near-end strategy optimization method based on cognitive behavior knowledge and application thereof

Abstract

The invention discloses a distributed near-end strategy optimization method based on cognitive behavior knowledge and application thereof, and the method comprises the following steps: employing cognitive behavior knowledge to establish a cognitive behavior model of an Agent, introducing the cognitive behavior model into deep reinforcement learning, and constructing a deep reinforcement learning framework based on cognitive behavior knowledge; based on the deep reinforcement learning framework, providing a distributed near-end strategy optimization algorithm based on cognitive behavior knowledge; and quantitatively designing a guide mode of the cognitive behavior model for Agent strategy updating, and achieving continuous learning of the Agent on the basis of cognitive behavior knowledge. According to the method provided by the invention, cognitive behavior knowledge can be effectively utilized, and strategy updating is carried out on the basis, so that the learning efficiency of the Agent is improved.

Description

technical field [0001] The invention relates to the technical field of deep reinforcement learning, in particular to a distributed proximal strategy optimization method based on cognitive behavioral knowledge and its application in air combat maneuver decision-making. Background technique [0002] In recent years, deep reinforcement learning has found widespread applications in video games, traffic light control, robotics, and more. However, the reinforcement learning algorithm requires a large number of samples and continuously optimizes the agent's strategy through trial and error. However, the huge sampling complexity limits the application of reinforcement learning algorithms in practical problems, and the use of existing cognitive behavioral knowledge to accelerate agent policy learning is an effective means to solve the above problems. [0003] Human beings and learning agents are quite different in cognition level, cognition mode and behavior mode. It is very difficu...

AI Technical Summary

Problems solved by technology

However, the reinforcement learning algorithm requires a large number of samples and continuously optimizes the Agent's strategy through trial and error.

However, the huge sampling complexity limits the application of reinforcement learning algorithms in practical problems, and the use of existing cognitive behavioral knowledge to accelerate agent policy learning is an effective means to solve the above problems

Images