Simulation Analysis And Recognition Of Pulse Signal Based On The Electricity Network Of Cardiovascular

Pulse diagnosis as the core of the Traditional Chinese Medicine (TCM) four diagnostic, which is the essence of TCM. Pulse signal usually refers to the pressure fluctuation in human radial artery,when the volatility conducts from the left ventricle to the whole human body as carriers of human vascular,which inevitably gets the physiology and pathology of human circulatory.In the analysis of the pulse signal, we usually use the time-domain analysis method for the quantitative analysis of the part with the physiological significance in pulse figure.With the rapid development of the computer technique, many experts and scholars begin to use the method of frequency-domain and time joint with frequency-domain to analyze the pulse signal, which also make a lot of valuable information. Additionally, based on the double elastic cavity model,with the method of human cardiovascular computer simulation model to analyze the contact of the pulse wave with the physiological and pathological is also an effect method,which can get more information.Now I research the pulse signal with the method of setting up the human cardiovascular system using the electricity network model. I divide the modeling process into two parts, Part Ⅰ is the cardiac excitation source model, which refers to time-varying parameters,Part Ⅱ is the cardiovascular arterial tree hybrid electric network model. Part Ⅰ mainly set up the five order lumped parameters of cardiovascular system, which use the Runge-Kutta method to solve differential equations and gets the blood flow in the left ventricle as the input source of the Part Ⅱ model. Part Ⅱ uses the power electronic simulation SimPowerSystem components in the Matlab circuit to set up the model, which reduces the enormous computational cost, owing to the fact that the solving process is integrated with the graphical language in SimPowerSystem.In this paper, Part Ⅱ totally set three model for different lumped form:the dual elastic chamber with left hand expansion mix model, which includes the abdominal aorta and the thoracic aorta with their major branches and left hand fine model; the three elastic chamber with left hand expansion mix model, which includes small aorta, middle aorta, big aorta, total peripheral vascular bed and left hand fine model; fifteen units with left hand expansion mix model, which includes brachiocephalic aorta, ascending aorta, thoracic aorta, abdominal aorta, visceral cycle, renal cycle, upper and lower body cycle, vena cava and left hand fine model.The whole study using Part Ⅰ and Part Ⅱ combines with the optimization algorithm, mix genetic algorithms with simulated annealing algorithms, to get the most appropriate variability for the model,which achieves the voltage signal wave in the radial artery of model matching the pulse signal perfectly, which include three kinds of type:normal pulse signal, wiry pulse signal, slippery pulse signal, totally provided by the Shanghai University of Traditional Chinese Medicine. Through the modeling solution, we can get a group of hemodynamic parameters, which can reflect the physiological and pathological information of the sample pulse signals. Additionally, I analyze the hemodynamic parameters associated with the pathological information and achieve the automatic classification and recognition for three kings of pulse signals with the multi-classification support vector machines, which all get good results. In addition, using the cardiovascular electric network model makes the research of the pulse morphology influence factors and pulse shape analysis of the three parts in the radial artery.