scholarly journals Robust Quadrotor Control through Reinforcement Learning with Disturbance Compensation

2021 ◽  
Vol 11 (7) ◽  
pp. 3257
Author(s):  
Chen-Huan Pi ◽  
Wei-Yuan Ye ◽  
Stone Cheng
Keyword(s):  
Neural Network ◽  
Reinforcement Learning ◽  
Control Strategy ◽  
External Disturbance ◽  
Control Agent ◽  
Network Control ◽  
Outdoor Environment ◽  
Disturbance Compensation ◽  
Tracking Accuracy ◽  
Control Scheme

In this paper, a novel control strategy is presented for reinforcement learning with disturbance compensation to solve the problem of quadrotor positioning under external disturbance. The proposed control scheme applies a trained neural-network-based reinforcement learning agent to control the quadrotor, and its output is directly mapped to four actuators in an end-to-end manner. The proposed control scheme constructs a disturbance observer to estimate the external forces exerted on the three axes of the quadrotor, such as wind gusts in an outdoor environment. By introducing an interference compensator into the neural network control agent, the tracking accuracy and robustness were significantly increased in indoor and outdoor experiments. The experimental results indicate that the proposed control strategy is highly robust to external disturbances. In the experiments, compensation improved control accuracy and reduced positioning error by 75%. To the best of our knowledge, this study is the first to achieve quadrotor positioning control through low-level reinforcement learning by using a global positioning system in an outdoor environment.

Real Time Neural Network Control of a Twin Rotor System

2011 ◽  
Vol 271-273 ◽  
pp. 616-621 ◽  
Author(s):  
Jih Gau Juang ◽  
Ting Kai Liu
Keyword(s):  
Neural Network ◽  
Real Time ◽  
Pid Controller ◽  
External Disturbance ◽  
Computing Time ◽  
Rotor System ◽  
Network Control ◽  
Attitude Tracking ◽  
Control Scheme ◽  
Twin Rotor

Real time intelligent control scheme that utilizes neural network to perform PID controller to control a twin rotor system is proposed. An experimental propeller setup called the twin rotor multi-input multi-output system (TRMS) is used in this study. The use of single neuron in PID control reduces system computing time that makes on-line learning and real time control feasible. The proposed control scheme can drive the TRMS to follow desired attitudes. The pitch angle and the azimuth angle in the condition of cross-coupled between vertical and horizontal axes is considered. Furthermore, the control scheme can overcome external disturbance. Simulation results show that the new approach can improve attitude tracking performance and the neural network-PID controller can work in real time.

PIDNN control for Vernier-gimballing magnetically suspended flywheel under nonlinear change of stiffness and disturbance

2020 ◽  
pp. 095965182097757
Author(s):  
Jiqiang Tang ◽  
Mengyue Ning ◽  
Xu Cui ◽  
Tongkun Wei ◽  
Xiaofeng Zhao
Keyword(s):  
Neural Network ◽  
Attitude Control ◽  
Magnetic Force ◽  
Magnetic Bearing ◽  
Network Control ◽  
Gradient Descent Method ◽  
Neural Network Control ◽  
Proportional Integral Derivative ◽  
Tracking Accuracy ◽  
Proportional Integral

Vernier-gimballing magnetically suspended flywheel is often used for attitude control and interference suppression of spacecrafts. Due to the special structure of the conical magnetic bearing, the radial component generated by the axial magnetic force and the change of the magnetic air gap will cause the nonlinearity of stiffness and disturbance. That will lead to not only poor stability of the suspension control system but also unsatisfactory tracking accuracy of the rotor position. To solve the nonlinear problem of the system, this article proposes a proportional–integral–derivative neural network control scheme. First, the rotor model considering the nonlinear variation of disturbance and stiffness parameters is established. Then, the weight of neural network is adjusted by the gradient descent method online to ensure the accurate output of magnetic force. Finally, the convergence analysis is carried out based on the Lyapunov stability theory. Compared with the general proportional–integral–derivative control and the radial basis function neural network control, the simulation results demonstrate that the proposed method has the highest tracking accuracy and excellent performance in improving stability. The experimental results prove the correctness of the theoretical analysis and the validity of the proposed method.

Recurrent Higher Order Neural Network Control for Output Trajectory Tracking with Neural Observers and Constrained Inputs

2010 ◽  
pp. 286-311 ◽  
Author(s):  
Luis J. Ricalde ◽  
Edgar N. Sanchez ◽  
Alma Y. Alanis
Keyword(s):  
Neural Network ◽  
Tracking Error ◽  
Network Control ◽  
Design Parameters ◽  
The Neural Network ◽  
Control Scheme ◽  
Output Trajectory ◽  
Error Dynamics ◽  
Adaptation Law ◽  
Constrained Inputs

This Chapter presents the design of an adaptive recurrent neural observer-controller scheme for nonlinear systems whose model is assumed to be unknown and with constrained inputs. The control scheme is composed of a neural observer based on Recurrent High Order Neural Networks which builds the state vector of the unknown plant dynamics and a learning adaptation law for the neural network weights for both the observer and identifier. These laws are obtained via control Lyapunov functions. Then, a control law, which stabilizes the tracking error dynamics is developed using the Lyapunov and the inverse optimal control methodologies . Tracking error boundedness is established as a function of design parameters.

scholarly journals High Precision Adaptive Robust Neural Network Control of a Servo Pneumatic System

2019 ◽  
Vol 9 (17) ◽  
pp. 3472 ◽  
Author(s):  
Chen ◽  
Tao ◽  
Liu
Keyword(s):  
Neural Network ◽  
Adaptive Control ◽  
Robust Control ◽  
Network Control ◽  
Parametric Estimation ◽  
Pneumatic Cylinder ◽  
Unmodeled Dynamics ◽  
Neural Network Control ◽  
Tracking Accuracy ◽  
Neural Network Controller

In this paper, an adaptive robust neural network controller (ARNNC) is synthesized for a single-rod pneumatic actuator to achieve high tracking accuracy without knowing the bounds of the parameters and disturbances. The ARNNC control framework integrates adaptive control, robust control, and neural network control intelligently. Adaptive control improves the precision of dynamic compensation with parametric estimation, and robust control attenuates the effect of unmodeled dynamics and unknown disturbances. In reality, the unmodeled dynamics of the complicated pneumatic systems and unpredictable disturbances in working conditions affect the tracking precision. However, these cannot be expressed as an exact formula. Therefore, the real-time learning radial basis function (RBF) neural network component is considered for better compensation of unmodeled dynamics, random disturbances, and estimation errors of the adaptive control. Although the bounds of the parameters and disturbances for the pneumatic systems are unknown, the prescribed transient performance and final tracking accuracy of the proposed method can be still achieved with fictitious bounds. Asymptotic tracking performance can be acquired under the provided circumstance. The comparative experiments with a pneumatic cylinder driven by proportional direction valve illustrate the effectiveness of the proposed ARNNC as shown by a high tracking accuracy is achieved.

Sign in / Sign up

Export Citation Format

Share Document