Intelligent prosthetic arm force control based on sEMG analysis and BPNN classifier

Abstract

An intelligent prosthetic arm force control method based on surface electromyography signal (sEMG) is proposed. The time-frequency characteristics of sEMG signal under different maximum voluntary contraction (MVC) forces were analyzed. The muscle force's signal was extracted and normalized. Root mean square (RMS) value and median frequency (MF) are used as time domain characteristic and frequency domain characteristic respectively. The relationship between sEMG signal eigenvalues and muscle forces was recognized and analyzed by BP neural network (BPNN) classifier. The control model of the relationship between sEMG signal characteristic and output force of intelligent prosthetic arm was established, which satisfied the muscle force identification better. An experimental system based on light weight intelligent upper arm prototype combined with pressure sensor was designed and accuracy of intelligent prosthetic arm force control method was verified by comparing the output force of human upper arm and the output force of intelligent prosthetic arm.