Due to the electric scooter with nonlinear uncertainties, e.g., nonlinear friction force of the transmission belt, the linear controller was made to disable speed tracking control. In order to overcome this problem, a novel hybrid recurrent wavelet neural network (NHRWNN) control system is proposed to control for a permanent-magnet synchronous motor-driven electric scooter in this study. The NHRWNN control system consists of a supervised control, a recurrent wavelet neural network (RWNN), and a compensated control with adaptive law. According to the Lyapunov stability theorem and the gradient descent method, the on-line parameter training methodology of the RWNN can be derived by using adaptation laws. The RWNN with the on-line learning ability can respond to the system's nonlinear and time-varying behaviors according to different speeds in the electric scooter. The electric scooter is operated to provide disturbance torque with nonlinear uncertainties. Finally, performance of the proposed NHRWNN control system is verified by experimental results.
LIN, CHIH HONG
"A novel hybrid recurrent wavelet neural network control of permanent magnet synchronous motor drive for electric scooter,"
Turkish Journal of Electrical Engineering and Computer Sciences: Vol. 22:
4, Article 18.
Available at: https://journals.tubitak.gov.tr/elektrik/vol22/iss4/18