An adaptive sliding-mode controller developed from an orthogonal Haar wavelet is proposed for a pneumatic servo control system experiment to overcome its non-linear and time-varying characteristics. To achieve real-time control of the pneumatic servo system, the orthogonal Haar wavelet is employed to quickly and accurately fit a non-linear function, thus bypassing the model-based prerequisite. The adaptive laws for the coefficients of the Haar wavelet series are derived from a Lyapunov function to guarantee system stability. One of the authors’ purposes is to enhance the stability, reliability and working performance of the pneumatic servo system. Hence, the H_∞ tracking technique is incorporated into the conventional adaptive sliding-mode control method [Haar wavelet-based adaptive sliding-mode controller with H_∞ tracking performance (HWB-ASMC + H_∞)] to attenuate the vibration of servo valve, which is caused by the chattering effect. The authors also show that the proposed HWB-ASMC + H_∞ is robust against approximated errors, un-modelled dynamics and disturbances, and can reduce the control chattering problem. The advantages of the proposed method include that no system dynamic models being required to achieve the controller design and no trial-and-error efforts are needed in selecting an approximation function. Consequently, practical experiments on a pneumatic servo system are successfully implemented with different position tracking profiles, which validates the proposed method.

This paper proposes an optimized virtual model reference (OVMR) control synthesis method for semiactive suspension control based on ride and vehicle handling characteristics. First, we present the semiactive Macpherson suspension system as an H _∞ robust output feedback-oriented control model. Then, by using the combination of a set of linear matrix inequalities (LMIs) and genetic algorithm (GA), the desired internal states for the tracking control problem of the semiactive suspension can be obtained via an OVMR. To achieve the H _∞ performance of ride comfort and vehicle handling against the influence of parameter uncertainties and external disturbances of the system, a robust adaptive controller is designed so that the controlled system can track the desired states generated from OVMR. The tracking control can be converted into a stabilization problem with asymptotic convergence in the sense of Lyapunov stability theorem. To validate the effectiveness of the proposed approach, the cosimulation technique is employed to bridge the gap between the mathematically well-defined system model and the optimization quality of control. It can be confirmed that the designed control system can achieve performance-effective suspension control through the confident software-in-the-loop (SITL) simulation.

The electro-hydraulic displacement-controlled system (EHDCS) performs specific non-linear and time-varying characteristics such that an exact model-based controller is complicated to be realized and the servo control is difficult to be implemented. In this study, the design method and experimental implementation of an adaptive fuzzy controller with self-tuning fuzzy sliding-mode compensation (AFC-STFSMC) are proposed which has on-line tuning ability for dealing with the system time-varying and non-linear uncertain behaviours for adjusting the control rule parameters. This control strategy employs the adaptive fuzzy approximation technique to design the equivalent controller of the conventional sliding-mode control (SMC). Furthermore, the fuzzy sliding-mode control scheme with self-tuning ability is introduced to compensate the approximation error of the equivalent controller for improving the control performance. The proposed AFC-STFSMC scheme can design the sliding-mode controller with no requirement of the system dynamic model, be free from chattering, be stable tracking control performance, and be robust to uncertainties. Moreover, the stability proof of the proposed scheme using Lyapunov method is presented. The experimental results of the position control and the path control in EHDCS with different strokes and external disturbance forces show that the proposed AFC-STFSMC approach can achieve excellent control performance and robustness with regard to parameter variations and external disturbance.

The use of bodyweight unloading force control on a treadmill with therapist manual assistance for gait training imposes constraints on natural walking. It influences the patient’s training effect for a full range of natural walks. This study presents a prototype and a safety controller for a mobile rehabilitation robot (MRR). The prototype integrates an autonomous mobile bodyweight support system (AMBSS) with a lower-limb exoskeleton system (LES) to simultaneously achieve natural over-ground gait training and motion relearning. Human-centered rehabilitation robots must guarantee the safety of patients in the presence of significant tracking errors. It is difficult for traditional stiff controllers to ensure safety and excellent tracking accuracy concurrently, because they cannot explicitly guarantee smooth, safe, and overdamped motions without overshoot. This paper integrated a linear extended state observer (LESO) into proxy-based sliding mode control (ILESO-PSMC) to overcome this problem. The LESO was used to observe the system’s unknown states and total disturbance simultaneously, ensuring that the “proxy” tracks the reference target accurately and avoids the unsafe control of the MRR. Based on the Lyapunov theorem to prove the closed-loop system stability, the proposed safety control strategy has three advantages: (1) it provides an accurate and safe control without worsening tracking performance during regular operation, (2) it guarantees safe recoveries and overdamped properties after abnormal events, and (3) it need not identify the system model and measure unknown system states as well as external disturbance, which is quite difficult for human–robot interaction (HRI) systems. The results demonstrate the feasibility of the proposed ILESO-PSMC for MRR. The experimental comparison also indicates better safety performance for the ILESO-PSMC than for the conventional proportional–integral–derivative (PID) control.