Scholarlink

<p>A prosthetic hand must provide adequate grasp speed and enough grip force for everyday use. However, this is very difficult for choosing motors that produce sufficient power to give high grasp speed and significant grip force because of limits in size and weight. In light of this, this study suggests a novel design for a prosthetic hand that uses a dual-motor actuator for actuation method. This type of actuator comprises the two motors that can function independently at high speeds under no load and high force under load to a differential mechanism. For a prosthetic hand to function at loads which are not required for high-power motor that is too big and heavy, one motor is utilized to create high torque, and another is used to create fast speed. Create a deep learning ensemble framework to forecast every grip’s best motor combination (speed vs. torque). Use a sizable dataset that includes a variety of grasping scenarios in Activities of Daily Livings (ADLs) to train the ensemble model. Improve system performance over time by using a hybrid optimization strategy. The hybrid optimization algorithm that has been suggested combines the Hunter-prey and Spider-wasp optimization techniques (HSWHP). The prosthetic hand has the following capabilities when it comes to gripping motions: With non-back drivable mechanisms, this may deliver the maximum grip force with 44&#xa0;N, sustain the maximum grip force with 38&#xa0;N, and have an average closing time of 0.5&#xa0;s. Additionally, a device may exert greater grip force with a single motor than research prosthetic hands.</p>

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

Single-DOF prosthetic hand with enhanced grasp capability using ensemble deep learning and hybrid optimization

  • 期刊论文

  • 发表日期：

    2025年8月22日

查看全文
  • J Josephine Hope Hailma,
  • S. Prakash

摘要

A prosthetic hand must provide adequate grasp speed and enough grip force for everyday use. However, this is very difficult for choosing motors that produce sufficient power to give high grasp speed and significant grip force because of limits in size and weight. In light of this, this study suggests a novel design for a prosthetic hand that uses a dual-motor actuator for actuation method. This type of actuator comprises the two motors that can function independently at high speeds under no load and high force under load to a differential mechanism. For a prosthetic hand to function at loads which are not required for high-power motor that is too big and heavy, one motor is utilized to create high torque, and another is used to create fast speed. Create a deep learning ensemble framework to forecast every grip’s best motor combination (speed vs. torque). Use a sizable dataset that includes a variety of grasping scenarios in Activities of Daily Livings (ADLs) to train the ensemble model. Improve system performance over time by using a hybrid optimization strategy. The hybrid optimization algorithm that has been suggested combines the Hunter-prey and Spider-wasp optimization techniques (HSWHP). The prosthetic hand has the following capabilities when it comes to gripping motions: With non-back drivable mechanisms, this may deliver the maximum grip force with 44 N, sustain the maximum grip force with 38 N, and have an average closing time of 0.5 s. Additionally, a device may exert greater grip force with a single motor than research prosthetic hands.