CME-EPC A coarse-mechanism embedded error prediction and compensation framework for robot multi-condition tasks

Contribution

This study presents a coarse mechanism-embedded error prediction and compensation (CME-EPC) framework for robotic multi-conditions tasks. Unlike the task-dependent characteristics of traditional mechanism models, the knowledge of coarse mechanism models is embedded in the CME-EPC in the form of data. Global error prediction and compensation are realized with the help of the proposed small-sample labeled, clustering-guided balanced domain adaptation by combining 10% measurement knowledge and all coarse mechanism knowledge. Four tasks in three working conditions were designed to verify the effectiveness of the framework. The CME-EPC framework was compared with six other methods, with the results suggesting significantly improved performance in error prediction. In addition, the compensation based on the prediction results also significantly reduces the robot error in all working conditions, with the best error reduction of 94%. Further analysis confirms that CME-EPC is robust against mechanism model uncertainty and stable in performance. All these conclusions fully illustrate the engineering application potential of CME-EPC.

Graphic Abstracts

The graphic abstract is shown below

Recommended citation: T. Zhang, F. Peng, X. Tang, R. Yan, R. Deng, CME-EPC: A coarse-mechanism embedded error prediction and compensation framework for robot multi-condition tasks, Robot. Comput.-Integr. Manuf., 86 (2024) 102675, https://doi.org/10.1016/j.rcim.2023.102675. https://doi.org/10.1016/j.rcim.2023.102675