Extreme learning machine-based super-twisting repetitive control for aperiodic disturbance, parameter uncertainty, frict

  • PDF / 6,454,775 Bytes
  • 13 Pages / 595.276 x 790.866 pts Page_size
  • 71 Downloads / 186 Views

DOWNLOAD

REPORT


(0123456789().,-volV)(0123456789().,-volV)

S.I. : EXTREME LEARNING MACHINE AND DEEP LEARNING NETWORKS

Extreme learning machine-based super-twisting repetitive control for aperiodic disturbance, parameter uncertainty, friction, and backlash compensations of a brushless DC servo motor Raymond Chuei1 • Zhenwei Cao1 Received: 26 February 2019 / Accepted: 18 April 2020 Ó Springer-Verlag London Ltd., part of Springer Nature 2020

Abstract This paper presents an extreme learning machine-based super-twisting repetitive control (ELMSTRC) to improve the tracking accuracy of periodic signal with less chattering. The proposed algorithm is robust against the plant uncertainties caused by mass and viscous friction variations. Moreover, it compensates the nonlinear friction and the backlash by using extreme learning machine based super-twisting algorithm. Firstly, a repetitive control is designed to track the periodic reference and compensate the viscous friction. Then, a stable extreme learning machine-based super-twisting control is constructed to compensate the aperiodic disturbance, nonlinear friction, backlash and plant uncertainties. The stability of ELMSTRC system is analysed based on Lyapunov stability criteria. The proposed algorithm is verified on a brushless DC servo motor with various loading, backlash and friction conditions. The simulation and experimental comparisons highlight the advantages of the proposed algorithm. Keywords Aperiodic disturbance  Backlash  Chattering  Coulomb friction  Extreme learning machine  Neural network  Nonlinear control  Parameter uncertainty  Periodic signal  Repetitive control  Super-twisting control  Viscous friction

1 Introduction Repetitive control (RC) was firstly proposed by Inoue et al. [1] in 1980s based on the internal model principle [2] to track the periodic signals and reject the periodic disturbances. RC can be applied on hard disk drives, robot manipulators, induction motors, power electronic devises, and more. Hence, RC has been studied by many scholars recently [3–7]. However, the conventional RC system is a neutral-type delay system. It may cause the whole system unstable, since the plant is strictly proper in most of the engineering applications. Hence, Hara et al. [3] proposed a modified RC (MRC) which inserts a low-pass filter in the delay line to stabilise the strictly proper plant.

& Raymond Chuei [email protected] 1

Faculty of Science, Engineering and Technology, Swinburne University of Technology, Melbourne, Australia

There is a limitation that RC cannot compensate and may even amplify the aperiodic disturbances [7]. Hence, the integration of RC with sliding mode control (SMC) is necessary, since the SMC is robust against aperiodic disturbances, and parameter variations [8–10]. However, the SMC produces undesired chattering due to the nonlinearity of the sign function. In order to reduce the chattering, the super-twisting control (STC) algorithm [11] was applied on super-twisting-based repetitive control (STRC) [12], which requires only informat