The Research Of Electronic Nose Detector On Surface Acoustic Wave

Since the World WarⅠ, a large number of Chemical Warfare Agents(CWAs) were used in the military technology. Furthermore, in past decades, CWAs were also used by terroristic organizations, which seriously endanger the development of ecology, as well as threaten human’s safety and property. So far, the International Convention of prohibiting CWAs have not successfully prohibited some countries and organizations to possess and manufacture CWAs, because of their characters of the easy manufactue, low cost, high lethality and a wide range of damage. Therefore, a handheld real-time detection approach is highly needed, which has fast response, high sensitivity and simple operation. Under certain circumstances, the electronic nose technology of trace gases based on Surface Acoustic Wave(SAW) sensor array has advantages of small size, fast response, easy to carry, thus is superior to other detection methods.The paper is based on this background, a SAW sensor array was fabricated by spray coating of selective films on SAW devices. Meanwhile, the responses to different vapors and concentrations were tested and analyzed. After combining the pattern recognition algorithm and dsPIC30F6014 A microcontroller, the classification of nerve agent stimulant(DMMP), mustard agent stimulants(2-CEES and DCP) and interfering vapors(acetone, dichloromethane, methanol and ethanol) was eventually realized. In addition, the LCD on dsPIC30F6014 A microcontroller can show the vapor types, along with the concentrations, simultaneously.This paper includes the background of this topic, the principle of SAW gas sensor, the platform of this vapor detection system, the analysis of pattern recognition algorithm, and the algirithm realization based on dsPIC30F6014 A microcontroller. The content of this research include the following aspects:(1) Design of SAW gas sensor array: This paper introduced the structure of the SAW gas sensor array based on two-channel SAW resonator mixing circuit. Spray fluoropolyol(FPOL), polycyanopropylmethylsiloxane(PCPMS) and polyepichlorohydrin(PECH) sensitive films onto SAW resonators, combine the coated SAW sensors and a bare device and then constructed a sensor array.(2) Tests and analysis vapor sensitivities of the SAW sensor array: Three simulations and four common vapors were tested and analyzed from 10 to 300 mg/m3 systematically. The results domonstrated that FPOL-SAW has surperior selectivity for nerve agent and PECH-SAW has relatively high selectivity for mustard.(3) Analysis on vapor pattern recognition system: Use the Statistical Product and Service Solution(SPSS) for pattern recognition of the gathered data with Principal Component Analysis(PCA) algorithm and Artifical Neural Networks(ANN) algorithm, respectively. The PCA model is simple, and can visual represented by char, while ANN is easy to understand. Both two algorithms can successfully distinguish agents and common vapors.(4) The algorithm realization based on dsPIC30F6014 A microcontroller for electronic nose: Realize circuit control and gas recognition algorithm on microcontroller, and the paper introduced and achieved Timer\Counter, Predict interference signal, Acquire gas response signal, Gas recognition, Serial transport and LCD module, respectively. Finally, verify the electronic nose algorithm.