| The changes of human pulse reflect the rising or falling of gas and blood in human internal organs. TCM doctors obtain information about the human body pulse from pressure sensation organs (such as fingers) in order to diagnose the illness. Although this diagnostic process is simple and noninvasive, it is mainly based on the clinical experience of individual doctor. In addition, because of differences in sensory sensitivity and subjective factors, the development and exchange of TCM pulse diagnosis objective are limited. Therefore, the objective recognition of TCM pulse diagnosis methods and innovative research will greatly promote the research of TCM pulse diagnosis.According to the modern control theory and system information theory, the internal state of the system can be known by analyzing the output signal from the system. Therefore, combined with modern signal processing methods and traditional pulse theory, the study of output pulse signal from human complex system will be the effective way to achieve the pulse objective and scientific research.First, plentiful information contained in pulse signals is mined by time and frequency domain analysis methods according to TCM theory and the mechanism of pulse formation, and then the time-frequency domain characteristics, which are used to distinguish the different pulses, are extracted. Then, the correlation dimension, maximum Lyapunov exponent and Kolmogorov entropy, which are used as the chaos feature parameters of pulse and used to quantitatively verify that the pulse is a typical chaotic signal, are received by calculating in the reconstructed multidimensional phase space of pulse. Secondly, neural network identifier is designed to identify pulses according to pulse characteristics. The lack of echo state networks is effectively solved by the designed identifier. Based on complex network theory and Lyapunov stability theory, adaptive and clustering algorithms are used. Complex networks adaptive clustering synchronization controller is designed. And a sufficient condition is proposed to achieve global stability of cluster synchronization. And then a novel adaptive cluster synchronization echo state network identifier model is designed by taking designed complex networks adaptive clustering synchronization controller as the middle layer. The identifier model is proved that it has the robustness of input variable and feedback variable through theoretical analysis and numerical analysis. In the recognition process, global stability can be achieved by adjusting the state matrix only. At the same time, the network identifier can effectively mimic the human brain identify process for multiple information characteristics of time series that have adaptive clustering features. Finally, the validity of the identifier model is examined in the recognition of pulse time series. And compared with the traditional identifiers, the effectiveness of the new identifier is verified further. |
PDF Full Text Request