Scholarlink

Dexterous torch hands, possessing comparable motion dexterity and perception capability to human hands, are pivotal for enabling agile operations in robotics. This paper presents a novel dexterous torch robot hand, DexThandI: DexMThand, which is designed as five modular fingers with 20 joints. Utilizing gear-based mechanical transmission driven by hollow-cup motors, the single joint achieves a maximum rotational velocity of 30 rpm. The compact DexMThand measures a maximum length of 213 mm, weighs approximately 1.8 kg. Through kinematic modeling and analysis, the theoretical maximum working space of this hand is 161 mm. Moreover, DexMThand features multimodal sensing capabilities, including position, tactile, and force feedback, where the tactile sensors can realize multi-dimensional tactile perception. The transmission of sensing and controlling information relies on the EtherCAT. Grasping and operating experiments are carried based on the contact force analysis and force-closure grasping strategy. And the results demonstrate the motion and perception capabilities of the proposed hand.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Mechanical Design and Motion Control for a Novel Dexterous Torch Hand

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Tianyu Liu,
  • Peng Kang,
  • Jieming Hou,
  • Lunfei Liang,
  • Houde Liu,
  • Liang Han

摘要

Dexterous torch hands, possessing comparable motion dexterity and perception capability to human hands, are pivotal for enabling agile operations in robotics. This paper presents a novel dexterous torch robot hand, DexThandI: DexMThand, which is designed as five modular fingers with 20 joints. Utilizing gear-based mechanical transmission driven by hollow-cup motors, the single joint achieves a maximum rotational velocity of 30 rpm. The compact DexMThand measures a maximum length of 213 mm, weighs approximately 1.8 kg. Through kinematic modeling and analysis, the theoretical maximum working space of this hand is 161 mm. Moreover, DexMThand features multimodal sensing capabilities, including position, tactile, and force feedback, where the tactile sensors can realize multi-dimensional tactile perception. The transmission of sensing and controlling information relies on the EtherCAT. Grasping and operating experiments are carried based on the contact force analysis and force-closure grasping strategy. And the results demonstrate the motion and perception capabilities of the proposed hand.