Robust Precision Motion Control Based on Enhanced Unknown System Dynamics Estimator for High-DoF Robot Manipulators
Xinyu Jia, Jun Yang, Tian Shi, Wenxin Wang, Yongping Pan, Haoyong Yu
Abstract
This article proposes efficient and robust motion control schemes for high-dimensional robot manipulators to perform precise manipulation under unknown disturbances. An enhanced unknown system dynamics estimator (EUSDE) is formulated to estimate disturbances without relying on the acceleration signal and the inverse of the inertia matrix. Low estimation errors are achieved by mitigating the effects of measurement noise and phase lag. Based on the EUSDE, a robust control framework is also developed and its feedback terms are enhanced from different perspectives. The first scheme is to design error-driven control gains, while the second one exploits the finite-time convergence of super-twisting algorithms. We conduct various simulations and hardware experiments on a manipulator with 7 degrees of freedom. The proposed controllers exhibit high tracking accuracy and strong disturbance rejection capability in validation, and importantly show high deployment efficiency with low workload in parameter tuning.