A method of aircraft formation control based on deep reinforcement learning

Through a deep reinforcement learning method, a Markov model and DQN algorithm are established, and a neural network is trained to achieve precise control of multi-aircraft formations, and an autonomous collision avoidance strategy is designed, which solves the problem of collision avoidance in formation flights and improves the autonomy and intelligence of formation control.

CN115576353BActive Publication Date: 2025-03-18BEIJING INST OF TECH
2 Cites 0 Cited by

Patent Information

Application Number
CN202211288418.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-10-20
Publication Date
2025-03-18
Estimated Expiration
2042-10-20

AI Technical Summary

Technical Problem

The prior art has failed to effectively solve the collision avoidance problem of aircraft during formation formation during multi-aircraft formation during multi-aircraft formation, especially when there are many aircraft, the collision avoidance logic design is complex and difficult to implement.

Method used

Using the aircraft formation control method based on deep reinforcement learning, the neural network is trained to learn formation behavior, realize precise formation control, and design independent collision avoidance strategies to avoid collisions.

Benefits of technology

It realizes precise control of aircraft formation, reduces the collision risk during formation formation, improves the autonomy and intelligence of formation control, and is suitable for multi-aircraft formations.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The present invention provides an aircraft formation control method based on deep reinforcement learning. The method takes into account the random disturbance of the environment to the aircraft, establishes an aircraft formation motion model, then maps the formation problem into a Markov decision process, and constructs a deep reinforcement learning network to learn the formation behavior of the aircraft. The method mainly learns based on the feedback given by the environment after the aircraft executes the action. Therefore, after multiple learning and explorations, the aircraft can obtain the correct formation control action. This method does not require an accurate mathematical model of the aircraft. Therefore, it is feasible and effective when there is random interference from the environment in the model. Dividing the formation formation process into three sections can effectively reduce the state space dimension of each stage and increase the success rate of network training. After the formation control algorithm based on reinforcement learning is obtained through training, it can be applied to start the formation from different initial positions without the need for network training, which greatly expands the application scope of the formation controller.
Need to check novelty before this filing date? Find Prior Art