A Flywheel Energy Storage System Suspended by Active Magnetic Bearings Using a Fuzzy Control with Radial Basis Function Neural Network

Nguyen, Van-Sum; Tran, Hong-Van; Lai, Lai-Khac; Nam, Nguyen Thi-Hoai; Giang, Tran Huu-Chau; Truong, Le-Phuong

doi:10.1007/978-3-319-17314-6_38

Cited by 2 publications

(1 citation statement)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Modern control theory for nonlinear system such as fuzzy control (Sung et al, 2012) and sliding mode control (SMC) (Aydın and Gu¨rleyen, 2016) are also research focuses in AMB control. Additionally, the combinations of control methods with intelligent algorithms such as neural network (Nguyen et al, 2016), genetic algorithm (Zhong and Palazzolo, 2015), and particle swarm optimization (PSO) (Wei et al, 2015), integrate respective advantages and enhance the control effect.…”

Section: Introductionmentioning

confidence: 99%

Sliding mode control with deep learning method for rotor trajectory control of active magnetic bearing system

Yao

Chen

2018

Transactions of the Institute of Measurement and Control

View full text Add to dashboard Cite

Active magnetic bearing (AMB) is competent in rotor trajectory control for potential applications such as mechanical processing and spindle attitude control, while the highly nonlinear and coupled dynamic characteristics especially in the condition of rotor large motion are obstacles in controller design. In this paper, a controller of AMB is proposed to achieve rotor 3D trajectory control. First, the dynamic model of the AMB-rotor system containing a nonlinear electromagnetic force model is introduced. Then the DCNN-SMC (deep convolutional neural network - sliding mode control) controller is proposed. Sliding mode control is used to achieve the tracking control with high robustness and responsiveness, and a deep convolutional neural network based on deep learning method is designed to compensate the uncertainties of the system. Finally, simulation of a 5-degree of freedom (DOF) system on various trajectories demonstrates evident control effect of the proposed controller in precision and significant effect of DCNN based on deep learning method in compensation control.

show abstract