Classification of lower extremity motions for exoskeleton robot control

in this paper, an algorithm for classifying human lower extremity movements was proposed and developed for the control of exoskeleton robots. The developed algorithm converts the movements of the knee and hip joints into two-dimensional data and classifies them into motions such as walking, running, and sitting. As a result of the experiment, it showed the highest performance when the movements of the knee and hip joints were judged at the same time, and the performance level was maintained even when the input size was reduced for real-time performance. The developed algorithm is expected to be installed in an exoskeleton robot operating in real time in the future.