The Method Of Sensor Detection And Recognition Of Flammable Liquid Residues In Fire Scene

Fire is a kind of serious disruptive accident which has caused malignant influence to national security and social stability. In the artificial arson, flammable liquid is often used to set fire. The fire scene is usually severely damaged and the situation is complicated. Due to the lack of the corresponding fire scene investigation equipment available, investigators used to depend on the experience and accelerant detection canines to find the cause of fire, the traditional survey method has many defects and cannot meet the needs of the current situation in fire investigation. However, the application of advanced fire investigation equipment is the key factor affecting the fire cause determination and also an important means of enhancing the strength of the investigation. So for the actual situation of fire scene investigation, it has a very important significance of developing detection equipment for arson scene flammable liquid and improving the efficiency and accuracy of on-site detection of flammable liquids.This paper designs and realizes a portable and real-time detection system for flammable liquid residues. In terms of hardware, we mainly designed a multifunctional and sensitive sensor array to gasoline, diesel and other common flammable liquid, and designed an efficient sampling and injection method. Based on the advantage of embedded technology, combined with chromatographic technique and sensor array, the residual gas after the combustion of flammable liquids was detected. It realized portable, real-time, accurate detection.On this basis, we analysis and process the response signals with the feature extraction and pattern recognition algorithms and finally give recognition results. First of all, in order to eliminate the influence caused by the pressure of carrier gas, concentration and other external factors, we normalize sample data. Then, the main mode characteristics of the sensor array are extracted by PCA, LDA and MFA. Finally, the classifiers distinguish the category of test sample according to the extracted features. The experimental results show that, the portable and real-time detection system for flammable liquid residues can classify flammable liquids and residues, and can obtain a good recognition result.