Supplementary Reinforcement Learning Controller Designed for Quadrotor UAVs
Authors:
Open Access
- 25 February 2019
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 7, 26422-26431
Abstract
The control problem for quadrotor UAVs is difficult and challenging due to the complex nonlinear dynamics and ever-changing disturbances. In this paper, a supplementary controller based on reinforcement learning (RL) is proposed to improve the control performance of quadrotor UAVs. The proposed RL method is constructed by an actor-critic structure and some improved technologies, e.g. Q-learning, temporal difference (TD) and experience replay. With the proposed method, the speed and stability of training can be improved greatly. On one hand, the supplementary controller can work together with the traditional controller on line, which can guarantee the stability of the system. On the other hand, the model uncertainties and external disturbances could be restrained through the on-line RL training. Lyapunov theory is used to prove the convergence of RL controller’s weights theoretically. Finally, three simulations are provided to illustrate the effectiveness of the proposed controller.Keywords
Funding Information
- Fundamental Research Funds for the Central Universities (FRF-GF-17-B46)
- National Natural Science Foundation of China (61520106009, 61533008, 61703037)
This publication has 24 references indexed in Scilit:
- Adaptive tracking control of a quadrotor unmanned vehicleInternational Journal of Precision Engineering and Manufacturing, 2017
- Adaptive Trajectory Tracking for Quadrotor MAVs in Presence of Parameter Uncertainties and External DisturbancesIEEE Transactions on Control Systems Technology, 2017
- Gr-GDHP: A New Architecture for Globalized Dual Heuristic Dynamic ProgrammingIEEE Transactions on Cybernetics, 2016
- Output feedback observer-based dynamic surface controller for quadrotor UAV using quaternion representationProceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, 2016
- Q-Learning-Based Vulnerability Analysis of Smart Grid Against Sequential Topology AttacksIEEE Transactions on Information Forensics and Security, 2016
- Data-Driven Optimal Consensus Control for Discrete-Time Multi-Agent Systems With Unknown Dynamics Using Reinforcement Learning MethodIEEE Transactions on Industrial Electronics, 2016
- Model-based reinforcement learning for approximate optimal regulationAutomatica, 2016
- Neural‐network‐based online optimal control for uncertain non‐linear continuous‐time systems with control constraintsIET Control Theory & Applications, 2013
- Backstepping Approach for Controlling a Quadrotor Using Lagrange Form DynamicsJournal of Intelligent & Robotic Systems, 2009
- PID vs LQ control techniques applied to an indoor micro quadrotor2004