Research And Appaction Of Wearable EMG Acquisiton Device

Recently, the human-computer interecactive technology based on bioelectricity is the frontier of the human-computer interection. The surface electromyogram (sEMG) signals, which are recorded the electrical signals from muscle fibers during voluntary contractions, are widely utilized as the input signals of human-machine interface(HMI). As it is easy to acquire and no trauma to the body, it is a research focus on HMI based on sEMG. However, these conventional sEMG devices have many limitations, such as high prices, not easy to carry. Combining sEMG based on wearable devices with HMI, a wearable sEMG acquisition system is developed, and it is possible for the HMI based on sEMG device to be gradually popular in family.This thesis is completed in the following parts:1. Hardware design of sEMG devicesFirstly, every functional unit of the sEMG acquisition circuit is designed, verified and analyzed closely. On the basis of these, the defects of the conventional sEMG are pointed out. Then this paper compares analogy filter with digital filter in the sEMG devices. We used techniques of analogy filter, amplifier, digital filter, Bluetooth, FPC and low power msop components and so on, to complete the design of wearable dual channel sEMG devices based on NRF51822 and FPC skilledlly.2. Hand gesture recognition algorithm based on the sEMG of upper limbUsing SVM algorithm, offline analysis of sEMG transmitted from wearable devices is realized, including preprocessing, feature extraction, normalization and selection of SVM kernel function. After obtaining satisfactory classification results in off-line analysis, the training model is loaded into the online control system based on SVM classifier, which can control the test equipment to complete the task according to the sEMG generated by upper limb movement. The accuracy is more than 90%.In this thesis, the wearable sEMG acquisition system is completed by containing electrode, electrode lead wire and sEMG acquisition circuit, which is small, low power and easy to use. Finally, the research on HMI based on sEMG is tested, by utilizing the wearable sEMG system.