An Adaptive Nonsingular Fast Terminal Sliding Mode Control Method With Time-Delay Estimation for Pneumatic Soft Actuator Without Specific Model

Abstract

Due to the complex nonlinear behavior and external sensitivity of pneumatic soft actuators (PSAs), it is difficult to design control methods to achieve their control objectives by establishing their precise specific models. This article takes a PSA with bellow-type folds, called pneumatic bellow actuator (PBA), as the object, and proposes an adaptive nonsingular fast terminal sliding mode (NFTSM) control method with time-delay estimation (TDE) to achieve the trajectory tracking control of the PBA without relying on its specific model. The model framework of the PBA is first discussed, but the specific model is unknown, and all the unknown model information is defined as the lumped unknown dynamics. The TDE is applied to obtain the estimation of the lumped unknown dynamics using only the measured system states. Based on the estimation information, an adaptive NFTSM control method is proposed. This control method is nonsingular, continuous, adaptive, and accurate. Since the step of establishing the specific model of the PBA is omitted, and the model framework used in this article is universal for many PSAs, this control method exhibits universal applicability for controlling other PSAs that have difficulty obtaining specific models. The effectiveness is validated through stability analysis and PBA tracking experiments under different loads.