Design And Implementation Of Acquisition And Processing System For Lower Limb SEMG

When the human body is doing exercise, the skin surface above the muscles will produce a weak biological signal, which is called surface electromyography. The biological signals can be a true reflection of the body of the movement, and the way to extract the signal is simple, painless, and noninvasive. Using the scientific methods, researchers extracted the useful information from the surface electromyography, which reflects the muscle’s degree of fatigue and limb’s movement intentions. Thus the biological signals can be used in the limb rehabilitation training, intelligent prosthesis and medical diagnostics and have high research value in research areas.The body is an excellent conductor and filled with a variety of electro-magnetic interference noise, the EMG amplitude of the body’s surface is usually below a mill volt, and the amplitude of the interference noise is usually more than a mill volt, so the EMG extraction from the surface of human skin is often lost among the interference signals. In order to extract biological information, the useful signals need to be enlarged and unwanted noise need to be filtered out firstly. For the existing problems in the study of the sEMG, this paper proposed a system acquiring and processing sEMG which can be simply implemented; and the real-time and stability of this system will be tested when it works with the robot using for lower limb rehabilitation training.This paper will mainly research on the following four parts:(1) On the foundation of learning of the research status about sEMG at home and abroad, this paper described the characteristics and mechanism of EMG and the sEMG’s characteristics and the relationship joint movement and muscles of the human lower limb.(2) The hardware circuit of the instrument using for acquiring sEMG is designed, elaborating systematically the overall structure of the hardware, describing the principles of circuit design and composition of sub-module, introducing the component of the hardware in detail.(3) The software processing sEMG is designed and implemented, describing the overall design of the software’s structure and the development environment; proposing and implementing a simple efficient algorithm using for exacting signal’s feature; finishing the classification and recognition of the movement intention by using the existing development package of support vector machine.(4) The test environment of the system and the experimental test results are given; describing the tectonic structure of the lower limb rehabilitation training robot, working mechanism; using the TCP protocol to achieve the error-free communication between the EMG processing system and robot’s control system.As part of Key Program for International Cooperation Projects of Hubei province of China, whose name is multi-DOF Robot for Lower Limb Replacement and Rehabilitation, this paper will achieve the system using for acquiring and processing sEMG from the surface of human lower limb, and the achieved system will control the movement of the robot platform by limber movements.