Adaptive sliding mode observer–based integral sliding mode model-free torque control for elastomer series elastic actuator–based manipulator. (May 2022)

• Record Type: Journal Article
• Authors: Wei, Yangchun; Wang, Haoping; Tian, Yang
• Is Part Of: Proceedings of the Institution of Mechanical Engineers. Volume 236:Number 5(2022)
• Page Start: 1010
• Page End: 1028
• Publication Date: 2022-05
• Subjects: Model-free control -- sliding mode observer -- integral sliding mode control -- adaptive control -- series elastic actuator; Mechanical engineering -- Periodicals; Automatic control -- Periodicals; Systems engineering -- Periodicals; 621.3
• DOI: 10.1177/09596518211064757
• Languages: English
• ISSNs: 0959-6518

Abstract

In this brief, an adaptive nonsingular terminal sliding mode observer–based adaptive integral terminal sliding mode model-free control is proposed for the trajectory tracking control of the output torque of elastomer series elastic actuator–based manipulator. Considering the tip load and its external disturbance, an elastomer series elastic actuator–based manipulator model is established. In order to realize the output torque tracking control of elastomer series elastic actuator–based manipulator, by using the characteristics of elastomer series elastic actuator, the output torque control is transformed into position control. Based on the idea of model-free control, an ultra-local model is applied to approximate the dynamic of the manipulator, and all the model information is considered as an unknown lumped disturbance. The adaptive nonsingular terminal sliding mode observer is designed to estimate the lumped disturbance, and the absolute value of the tracking error is introduced into the sliding surface to make the selection of parameters more flexible. Then, on the basis of adaptive nonsingular terminal sliding mode observer, the adaptive integral terminal sliding mode model-free control is proposed under model-free control framework. The design and analysis of both observer and controller do not rely on accurate model information. Finally, the performance of the proposed method is verified by simulation results.