Bio-inspired auto-adaptive framework for optimized movement of passive knee prosthesis

This research addresses the challenges faced by amputees who struggle while performing daily activities due to a missing limb. The objective is to create a bio-inspired framework that intelligently adapts to compensate for lost mobility and mimics natural walking for passive knee users. We have developed a framework that takes input power from human femur and drives the passive knee with the help of sensors and damping control mechanism. Our deep learning architecture achieved a high classification accuracy 94.44% for gait phase events. The proposed framework demonstrated optimized movement with reduced hip hikes and less fatigue, maintaining normal knee flexion (64^∘±6), and achieving a good fall prevention rate of 95%. This research presents a promising solution to improve the functionality and comfort of passive knee prostheses, significantly improving the quality of an amputee's life.