Adaptive discrete sliding mode control of an electro-hydraulic actuator system

摘要

This thesis presents system identification and development of an adaptive robust control strategy based on discrete sliding mode control (DSMC) with zero phase error tracking control (ZPETC) for an electro-hydraulic actuator (EHA) system. A linear type actuation of the EHA system using a single-ended cylinder controlled by a servo valve was considered in the experimental design. In the system identification process, EHA system was modelled using parametric linear time varying equations with parameters that were identified using recursive and non-recursive identification techniques. An identification process that recursively computes the dynamic model was performed using recursive least square with varying forgetting factors and the estimated linear model was validated through statistical approaches. From the identification process, a non-minimum phase model of EHA system with a high sampling time was obtained. To formulate the control algorithm for the EHA system, a robust feedback control theory with feedforward structure was employed to overcome the non-minimum phase problem in EHA system. The algorithm was also subjected to model uncertainty and non-linear characteristics. As a result, a new robust controller with an integrated design scheme based on DSMC and ZPETC was developed using a reaching law technique where parameters of the controller had been analytically determined. Subsequently, the new adaptive control strategy was improved by enhancing DSMC and ZPETC that are adaptable with variations in the parameters of EHA system. In simulation and experimental studies, an optimal linear-quadraticregulator (LQR) and a proportional-integral-derivative (PID) were implemented in the position tracking control as comparisons with the proposed robust controller. A comprehensive performance evaluation with quantitative measures of the tracking performance is presented and the results show that the robust system performance was achieved with DSMC under different operating system conditions. The findings also demonstrated that the new adaptive DSMC with ZPETC structure has reduced the control effort and gave a better performance in terms of tracking accuracy as compared to the conventional DSMC, LQR and PID controllers.