Author : Po-Hsun Wu
Publisher :
ISBN 13 :
Total Pages : 0 pages
Book Rating : 4.:/5 (14 download)
Book Synopsis Applying the PPO Algorithm to Fixed-Wing UAV Attitude Control by : Po-Hsun Wu
Download or read book Applying the PPO Algorithm to Fixed-Wing UAV Attitude Control written by Po-Hsun Wu and published by . This book was released on 2023 with total page 0 pages. Available in PDF, EPUB and Kindle. Book excerpt: This thesis applies the reinforcement learning (RL) to design the attitude controller for fixed-wing unmanned aerial vehicle. In contrast to traditional linear control theory, which are limited to linearization, or complex nonlinear control theory solving processes, reinforcement learning utilizes an artificial neural network as the control law and applies RL algorithm to train the network by adjusting its parameters. The results demonstrate that the controller designed using the RL framework exhibits advantages over traditional control laws. Within the RL architecture, there are four main elements: policy, algorithm, reward function, and environment. In this thesis, a combination of a quadratic performance index and a time varying coefficient is used as the reward function. The categorize policy is employed as the controller, and the linearized UAV dynamic model and the fourth order Runge-Kutta method are utilized as the training environment. The Proximal Policy Optimization algorithm (PPO algorithm) is used to train the policy. Upon completing the training, a neural network controller is obtained. The results demonstrate that the trained controller successfully achieves the goal of controlling the UAV's attitude.