Unmanned aerial vehicle safety path planning method based on maximum entropy multi-agent reinforcement learning
The invention discloses an unmanned aerial vehicle safe path planning method based on maximum entropy multi-agent deep reinforcement learning, and the method comprises the steps: building a reinforcement learning air combat simulation environment without human-computer interaction on the basis of a...
Saved in:
Main Authors | , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
19.04.2024
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention discloses an unmanned aerial vehicle safe path planning method based on maximum entropy multi-agent deep reinforcement learning, and the method comprises the steps: building a reinforcement learning air combat simulation environment without human-computer interaction on the basis of a pre-assumed condition for the collaborative path planning problem of an unmanned aerial vehicle group, and completing the initialization setting of parameters; introducing a two-dimensional unmanned aerial vehicle kinetic equation; constructing a six-tuple of a partial observable Markov decision process of an unmanned aerial vehicle group collaborative path planning problem to obtain a POMDP model; based on a multi-agent soft actor commentator algorithm, through interaction of unmanned aerial vehicles and an air combat simulation environment, training agents to solve an unmanned aerial vehicle group collaborative path planning POMDP model strategy in the air combat simulation environment, and obtaining trained agen |
---|---|
Bibliography: | Application Number: CN202410070432 |