Study On Algorithms Based On The Human Metabolism Of Volatile Compounds To Determine Signs Of Life

Odor compounds released by the body’s metabolism as much as 200 kinds of which compounds include alcohols, ketones, aldehydes, phenols, ethers, cool, acids, hydrocarbons, halogenated hydrocarbons, aromatic substances etc. More compounds contain information is also very rich, by the composition and properties of the human body odor, for human life, criminal investigation, health care, disaster relief and other fields have important significance. The current artificial olfactory system for main bionic devices achieve the recognition of the human body odor. It is collected by the data pre-processing and pattern recognition signal composed of three parts. Due to limitations in materials science and recognition algorithm, the current disaster rescue bionic artificial olfactory system is still in development, exploration stage.In the search to identify aspects of the research, you need to be studied composition and odor compounds released by diffusion law body, signs of life again based on the detected concentration information indirectly determine the unknown environment. Based on Fluent software compound concentration field part of the body to release the numerical simulation to study the regional and diversity of compounds; and conducted by Matlab mathematical modeling, simulation of the ideal high-dimensional, nonlinear, dynamic density information; due to the concentration of the odor compounds released information about the human body has a dynamic, high-dimensional, non-linear, non-stability characteristics, resulting in the identification algorithm complexity. In this paper, two algorithms were reviewed from the statistical analysis and intelligence analysis; then from basic interpolation algorithms, regression analysis, the ratio analysis, fuzzy analysis, curve fitting methods such laws compound concentrations and the information contained in the basic research carried out; Finally, neural network analysis and HTM experimental verification algorithms.This article first chapter purpose and significance of the current odor recognition algorithms are reviewed, and a description of the research situation, finally clarifies the task. The second chapter, by extracting the body’s metabolism and release characteristic odor compounds sensing applications by means of identification methods were studied. The third chapter, combined with knowledge of aerodynamics and Fluent software Matlab body odor smell compound concentration diffusion field simulation in the second half sealed the ideal environment. Chapter IV, combined with the basic algorithm analysis to identify the odor compound basis. Chapters V and VI, respectively, combined with neural network algorithm and HTM algorithm simulation information.