A novel human-robot interface based on low-cost sEMG device designed for the collaborative wearable robot

Abstract

Currently, the acquisition of surface electromyographic (sEMG) signals commonly requires powerful and expensive equipment or systems. The signal analysis also requires additional equipment to process independently, which is time-consuming and the results are difficult to understand and directly apply to wearable robots. Therefore, this paper developed a sEMG signal device with a simple structure and high-performance cost ratio. The hardware architecture consists of a two-stage small-size amplifier with a total gain of 1000 and a band-pass filter from 10 to 150 HZ. Since real measured sEMG signals are usually corrupted by various types of noises, this paper designed an envelope processing to obtain the sEMG envelope signal which has stronger anti-interference ability than the raw signal, and it is also easier to apply to wearable robots. The proposed system is evaluated quantitatively by the experimental test and electrode groups analysis. In addition, the feasibility and effectiveness of the system is performed by measuring the sEMG signal of the forearm muscle group of four volunteers during motion action and extract the time domain feature to control a wearable robot. The sEMG envelope successfully manipulates the wearable robot with four different actions, whose precision is as high as a conventional sEMG device. The proposed system is promising in providing the reliable control interface of a wearable robot based on the sEMG.