Optimization algorithm of gait recognition based on EMG signal

Abstract

At present, exoskeleton robot technology is becoming more and more mature, and its application scenarios have covered many aspects. In the field of exoskeleton robot technology, one of the most important research technologies is gait recognition technology. This article studies the human gait phase recognition algorithm based on sEMG of human lower limbs. After using KNN, improved KNN, SVM and dagsvm algorithms for gait recognition, this paper proposes and uses knn-dagsvm algorithm and improved knn-dagsvm algorithm. Finally, according to the recognition results, it is concluded that the average recognition rate of gait phase recognition obtained by using improved knn-dagsvm fusion algorithm is the highest, which is 95.00%..