We extract the chaotic characteristics of multiple physiological signals in this paper, and the relevant algorithms for extracting chaotic characteristics are given respectively. Based on different algorithms we respectively calculate four kinds of chaotic characteristic parameters from multiple physiological signals, such as maximum Lyapunov exponent, the correlation dimension, approximate entropy and complexity, to make them connect with the connotation of physiological signals not through the emotions. If we establish chaotic characteristic matrix by the extracted chaotic characteristic parameters, use the classifier to map the chaotic characteristics to a given category, so the problem of emotion analysis based on the multiple physiological signals is converted to the problem of classification in pattern recognition, which lay a foundation for better emotion recognition.
展开▼
