首页|Compliant Control of Flexible Joint Toward Prescribed Performance With Gaussian Kernels
Compliant Control of Flexible Joint Toward Prescribed Performance With Gaussian Kernels
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
IEEE
It remains a challenge to improve the accuracy of impedance rendering while ensuring stability under strong impacts during human-robot interaction. In this work, we aim to render the desired impedance for the flexible joint under an admittance control scheme with prescribed performance function (PPF). Specially, Gaussian kernels are introduced as the slack terms for PPF, so that the control stability can be maintained in the presence of abrupt external torques. Meanwhile, a narrower error envelope is yielded when such torques are absent, which also improves the fidelity of the desired impedance model. To achieve the prescribed tracking performance of the inner position loop, a two-stage backstepping control is proposed by defining two first-order composite error surfaces bridged by a second-order dynamic surface. This promulgates the minimum number of backstepping stages under the available state feedback, thus avoiding “explosion of terms.” In addition, dual-adaptive neural networks are incorporated into the backstepping control to compensate for the matched and unmatched disturbances. Real-time experiments are conducted to validate the appeal of the proposed method.
Hongyu Wan、Silu Chen、Xiangjie Kong、Xianbei Sun、Chin-Yin Chen、Jinhua Chen、Chi Zhang、Guilin Yang
展开 >
Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, China|College of Materials Science and Opto-Electronic Technology, University of Chinese Academy of Sciences, Beijing, China
NETL
NSTL
IEEE
