Design of Drive Mechanism for Finger Rehabilitation Robot

Abstract

The incidence of hand dysfunction caused by factors such as stroke and occupational injuries is increasing, severely impacting patients' daily living abilities. High-repetition training based on Continuous Passive Motion (CPM) theory can promote neuroplasticity; however, traditional manual rehabilitation suffers from issues such as high dependence, high cost, low efficiency, and poor compliance. Robot-assisted training, leveraging its high repetition, precise quantitative assessment, and programmable modes, has become a key direction for enhancing rehabilitation efficacy and reducing the healthcare burden. This paper focuses on three mainstream drive mechanisms: pneumatic drive, rigid linkage drive, and cable-driven drive. The pneumatic drive utilizes pneumatic pressure to deform flexible chambers, offering good compliance and high safety, but suffers from shortcomings such as insufficient control accuracy and slow response. The rigid linkage drive transmits power based on planar linkages, featuring high control accuracy, good ergonomics, and ease of achieving fine movements; however, multiple degrees of freedom lead to complex and bulky structures, affecting wear comfort and compliance. The cable-driven drive transmits power from remote motors via flexible cables, offering advantages such as lightweight design, good compliance, and strong adaptability; some designs utilize differential mechanisms to achieve self-adaptive grasping, but face challenges in friction control, elastic compensation, and efficient bidirectional driving. Analysis and comparison suggest that current designs need to strike a balance between safety, performance, and cost: pneumatic and cable-driven drives offer better compliance and adaptability, while the rigid drive demonstrates more pronounced advantages in precision.