scholarly journals Improved Neural Network Adaptive Control for Compound Helicopter with Uncertain Cross-Coupling in Multimodal Maneuver

2021 ◽  
Author(s):  
Fengying Zheng ◽  
Bowei Xiong ◽  
Jingyang Zhang ◽  
Ziyang Zhen ◽  
Feng Wang
Keyword(s):  
Neural Network ◽  
Complex Dynamics ◽  
Cross Coupling ◽  
Adaptive Controller ◽  
Control Input ◽  
Control Scheme ◽  
Robust Control System ◽  
Working Capability ◽  
Compound Helicopter ◽  
Parameters Perturbation

Abstract The main goal of this study was to create a robust control system that could guide or replace the pilots in tracking of commanded velocity and attitude in multimodal maneuver, while complex dynamics and uncertain aerodynamic cross-coupling among control surfaces of compound helicopter are considered. To this end, a Pi-Sigma neural network (PSNN) adaptive controller is proposed based upon the certainty-equivalence (CE) principle, where a novel Lyapunov-based weight self-tuning algorithm augmented with e-modification is designed to realize efficient uncertainty approximation and guarantee robustness of convergence process. Compared with traditional neural networks in control field, stronger generalization ability of PSNN must be balanced against weaker stability, which leads to inevitable parameters perturbation. Therefore, an incremental nonlinear dynamic inversion (INDI) framework is established to decouple original overactuated system and reject parameters perturbation in PSNN. Meanwhile, by incorporating Lagrang- multiplier method into allocation, an original incremental allocation method is designed to get globally ideal control input according to time-varying working capability of each surface. In terms of Lyapunov theorem, it is demonstrated that the closed-loop augmented system driven by the proposed control scheme is semi-global uniformly ultimately bounded (SGUUB). Finally, the simulation result validates the effectiveness of proposed control scheme.

scholarly journals A Distributed Formation Control Scheme with Obstacle Avoidance for Multiagent Systems

2019 ◽  
Vol 2019 ◽  
pp. 1-16 ◽  
Author(s):  
Xiaowu Yang ◽  
Xiaoping Fan
Keyword(s):  
Obstacle Avoidance ◽  
Multiagent Systems ◽  
Formation Control ◽  
Adaptive Controller ◽  
Control Input ◽  
Control Laws ◽  
Integral Term ◽  
Control Scheme ◽  
The Stability ◽  
Avoidance Problem

This study considers the problem of formation control for second-order multiagent systems. We propose a distributed nonlinear formation controller where the control input of each follower can be expressed as a product of a nonlinear term that relies on the distance errors under the leader–follower structure. In the leader–follower structure, a small number of agents are assumed to be the leaders, and they are responsible for steering a group of agents to the specific destination, while the rest of the agents are called followers. The stability of the proposed control laws is demonstrated by utilizing the Lyapunov function candidate. To solve the obstacle avoidance problem, the artificial potential approach is employed, and the agents can avoid each possible obstacle successfully without getting stuck in any local minimum point. The control problem of multiagent systems in the presence of unknown constant disturbances is also considered. To attenuate such disturbances, the integral term is introduced, and the static error is eliminated through the proposed PI controller, which makes the system stable; the adaptive controller is designed to reduce the effect of time-varying disturbances. Finally, numerical simulation results are presented to support the obtained theoretical results.

Control of Robot Manipulators Based on Legendre Orthogonal Neural Network

2013 ◽  
Vol 427-429 ◽  
pp. 1089-1092
Author(s):  
Pei Feng Wang
Keyword(s):  
Neural Network ◽  
Adaptive Learning ◽  
Lyapunov Stability Theory ◽  
Adaptive Controller ◽  
Tracking Errors ◽  
Orthogonal Functions ◽  
Control Scheme ◽  
Adaptive Weight ◽  
Simulation Results ◽  
Learning Law

In this paper, Legendre orthogonal functions neural network is used to achieve the control of nonlinear systems. The adaptive controller is constructed by using Legendre orthogonal functions neural network. The adaptive learning law of orthogonal neural network is derived to guarantee that the adaptive weight errors and tracking errors are bound by using Lyapunov stability theory. Simulation results are given for a two-link robot, and the control scheme is validated.

Dual Adaptive Neural Network Controller for Underactuated Systems

Robotica ◽  
2021 ◽  
pp. 1-18
Author(s):  
Seyed Hassan Zabihifar ◽  
Hamed Navvabi ◽  
Arkady Semenovich Yushchenko
Keyword(s):  
Neural Network ◽  
Adaptive Controller ◽  
Underactuated Systems ◽  
Adaptive Neural Network ◽  
Neural Network Controller ◽  
Network Controller ◽  
Control Scheme ◽  
Equivalent Control ◽  
Inertia Wheel Pendulum ◽  
The Given

SUMMARY A new stable adaptive controller based on a neural network for underactuated systems is proposed in this paper. The control scheme has been developed for two underactuated systems as examples. The Furuta pendulum and the Inertia Wheel Pendulum (IWP) have been examined in this paper. The presented approach aims to address the control problem of the given system in swing up, stabilization, and disturbance rejection. To avoid oscillations, two adaptive neural networks (ANNs) are implemented. The first one is used to approximate the equivalent control online and the second one to minimize the oscillations.

Intelligent control for Hammerstein nonlinear systems with arbitrary deadzone input

2015 ◽  
Vol 39 (4) ◽  
pp. 567-578 ◽  
Author(s):  
Bi Zhang ◽  
Zhizhong Mao ◽  
Tingfeng Zhang
Keyword(s):  
Neural Networks ◽  
Nonlinear Systems ◽  
Intelligent Control ◽  
Adaptive Controller ◽  
Control Input ◽  
Adaptive Controllers ◽  
Multiple Neural Networks ◽  
Control Scheme ◽  
Robust Adaptive ◽  
Robust Adaptive Controller

In this paper, a new intelligent control scheme based on multiple models and neural networks is proposed to adaptively control a class of Hammerstein nonlinear systems with arbitrary deadzone input. This approach consists of a linear robust adaptive controller, multiple neural networks-based nonlinear adaptive controllers and a switching mechanism. Since the control input is derived from a modified certainty equivalent principle, the manner in which the closed-loop stability is established forms the main contribution. To show the usefulness of the developed results, three simulation examples, including a direct current motor subject to a nonlinear friction, are studied.

The Application of Learning Impedance Control to Exoskeleton Arm

2012 ◽  
Vol 463-464 ◽  
pp. 900-904
Author(s):  
Jie Liu ◽  
Yu Wang ◽  
He Ting Tong ◽  
Ray P.S. Han
Keyword(s):  
Neural Network ◽  
Impedance Control ◽  
Rbf Neural Network ◽  
Robotic Manipulators ◽  
Iterative Learning ◽  
Artificial Muscles ◽  
Experimental Setup ◽  
Control Input ◽  
Control Scheme ◽  
Pneumatic Artificial Muscles

In this paper, we discuss the application of learning impedance control scheme to exoskeleton arm driven by pneumatic artificial muscles (PAM), for assisting in the rehabilitation of patients who suffer from debilitating illness. An iterative learning impedance control problem for robotic manipulators is analyzed, proposed and solved. The target impedance reference modifies a desired trajectory according to the force signals and position signals of the joint. The desired control input of learning impedance control was estimated by radial basis function (RBF) neural network incorporated experience database. The curves of experiment result on the experimental setup show that the algorithm is successful also in the application of exoskeleton arm.

Neural network based optimal adaptive attitude control of near-space vehicle with system uncertainties and disturbances

2018 ◽  
Vol 233 (2) ◽  
pp. 641-656 ◽  
Author(s):  
Rongsheng Xia ◽  
Mou Chen ◽  
Qiangxian Wu
Keyword(s):  
Neural Network ◽  
Attitude Control ◽  
Feedforward Control ◽  
Space Vehicle ◽  
Adaptive Method ◽  
Optimal Controller ◽  
Control Input ◽  
Near Space ◽  
Control Scheme ◽  
System Uncertainties

In this paper, a neural network based optimal adaptive attitude control scheme is derived for the near-space vehicle with uncertainties and external time-varying disturbances. Firstly, radial basis function neural network (RBFNN) approximation method and nonlinear disturbance observer (NDO) are used to tackle the system uncertainties and external disturbances, respectively. Subsequently, a feedforward control input under backstepping control frame with RBFNN and NDO is designed to transform the optimal tracking control problem into an optimal stabilization problem. Then, a single online approximation based adaptive method is used to learn the Hamilton–Jacobi–Bellman equation to obtain the corresponding optimal controller. As a result, the compound controller consists of feedforward control input and optimal controller which can ensure that the near-space vehicle attitude angles are able to track reference signals in an optimal way. Lyapunov stability analysis method is used to show that all the closed-loop system signals are uniformly ultimately bounded. Finally, simulation results show the effectiveness of the proposed optimal attitude control scheme.

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.

Sign in / Sign up

Export Citation Format

Share Document