New Integration-Enhanced Newton Algorithm for Real-Time Tracking Control of Robot Manipulators

Lin Wei,Long Jin,Rui Zhang,Xiuchun Xiao,Mei Liu,Jingwen Yan,Jiliang Zhang
DOI: https://doi.org/10.1109/ddcls.2019.8908844
2019-01-01
Abstract:It is of great significance to enhancing and maintaining accurate real-time output for tracking control which has been diffusely confronted in robotics. Note that, for most traditional algorithms, utilizing a neural controller for convergence and a filter for dealing with noises is a habitual method. However, there always exists interplay of the neural controller and filter that prevents these existing traditional algorithms from keeping stable when encountering tracking control problem, which results in large lagging errors. To this end, a novel integration-enhanced Newton (IN) algorithm aided with the superior integration control is proposed and inherently robust to a variety of noises, which is proved thoroughly by corresponding theoretical analyses and further employed to realize tracking control of robot manipulators successfully.
What problem does this paper attempt to address?