Simulation Environment for Grasp Force Control
收藏arXiv2025-09-30 收录
下载链接：
https://sites.google.com/view/rl-force-ctrl
下载链接
链接失效反馈官方服务：
资源简介：
该数据集提供了一个模拟环境，用于生成真实的正常力，以训练机器人抓握过程中的连续力控制策略。此外，数据集还包含了在11种不同刚度值下的评估，并比较了采用域随机化技术的不同策略之间的性能。总体而言，模拟评估包含了8800次试验。该任务主要关注抓握力控制以及模拟到真实环境的转移评估。
This dataset provides a simulated environment for generating realistic normal forces to train continuous force control strategies during robotic grasping. In addition, the dataset includes evaluations under 11 distinct stiffness values, and compares the performance of different strategies adopting domain randomization techniques. Overall, the simulated evaluation campaign encompasses 8800 trials in total. This task primarily focuses on grasping force control and sim-to-real transfer evaluation.
提供机构：
Authors of the paper
