There is no abstract available for this article.
This work was supported by National Key Research and Development Program of China (Grant No. 2017YFB1302300), National Natural Science Foundation of China (Grant Nos. 6150020696, 61503060), Sichuan Science and Technology Major Projects of New Generation Artificial Intelligence (Grant No. 2018GZDZX0037), and Fundamental Research Funds for the Central Universities (Grant No. ZYGX2015J148).
Appendix A.
[1] Atkeson C G, Moore A W, Schaal S. Locally weighted learning for control. Lazy Learn, 1997, 11: 75--113. Google Scholar
[2] Nguyen-Tuong D, Peters J R, Seeger M. Local Gaussian process regression for real time online model learning. In: Advances in Neural Information Processing Systems, 2009. 1193--1200. Google Scholar
[3] Yue M, Ning Y, Yu S. Composite following control for wheeled inverted pendulum vehicles based on human-robot interaction. Sci China Inf Sci, 2019, 62: 50206 CrossRef Google Scholar
[4] Li Z, Ma W, Yin Z. Tracking control of time-varying knee exoskeleton disturbed by interaction torque. ISA Trans, 2017, 71: 458-466 CrossRef PubMed Google Scholar
[5] Kazerooni H, Racine J-L, Huang L, et al. On the control of the Berkeley lower extremity exoskeleton (BLEEX). In: IEEE international conference on Robotics and automation (ICRA), 2005. 4353--4360. Google Scholar
[6] Song G, Huang R, Cheng H, et al. Learning continuous coupled multi-controller coefficients based on actor-critic algorithm for lower-limb exoskeleton. Sci China Inf Sc, 2020, doi: 10.1007/s11432-018-9779-6. Google Scholar
[7] Kelly R. PD Control with Desired Gravity Compensation of Robotic Manipulators. Int J Robotics Res, 1997, 16: 660-672 CrossRef Google Scholar
[8] Huang R, Cheng H, Guo H. Hierarchical learning control with physical human-exoskeleton interaction. Inf Sci, 2018, 432: 584-595 CrossRef Google Scholar
Figure 1
(a) Control diagram of ACL strategy; (b) comparison of ACL, HIL and SAC strategies with nMSE (rad) on HUALEX.