摘要
This paper proposes a modified dynamic surface control (DSC) for speed tracking and torsional vibration suppression for two-inertia systems with nonlinear friction. The proposed controller contains two parts: tracking controller and friction compensator. The tracking controller is designed by modifying dynamic surface control, which replaces the traditional first-order filter with a high-gain tracking differentiator (HGTD). Meanwhile, an improved prescribed performance function with error constraint is also presented and incorporated into DSC design. As for the friction compensator, the nonlinear nonsmooth friction is parameterized and then compensated using echo state neural networks (ESNs). The state observer with friction compensation is used to estimate unmeasurable load speed and torsional torque. The effectiveness of proposed control scheme is verified by simulation and experiment results.
源语言 | 英语 |
---|---|
页(从-至) | 73-85 |
页数 | 13 |
期刊 | Neurocomputing |
卷 | 203 |
DOI | |
出版状态 | 已出版 - 26 8月 2016 |