Reseach And Implementationan Embedded Electronic Nose System Based On ARM9

Electronic nose has developed rapidly in recent years. With the deep research of the computer technology, sensor technology, and human olfactory mechanism, the electronic nose in social development has been more and more concerned. China’s rapid industrial development has been showing the urgent need for highly intelligent instrument to solve many problems in the field of medical testing, environmental monitoring, food hygiene and national defense construction.Domestic detection and analysis of volatile gases use mostly off-the-shelf commercial electronic nose to do the study of algorithms and applications, little research on design and manufacture of electronic nose.Due to the complexity of equipment for gas chromatographic analysis, the method is difficult to achieve on-line or on-site detection analysis.A single gas sensor can be used only to single-component gas sensing and Can not effectively detect volatile gas mixture containing a variety of ingredients also can not meet the rapid development of science and technology needs. And foreign electronic nose equipment is expensive. Therefore, the study of the electronic nose system has broad application prospects.The objective of the study of this subject is embedded electronic nose equipment to study the key technologies for embedded electronic nose.Designed and implemented an electronic nose system.The system using ARM9for control chip, powered by the Linux operating system, combined with a variety of peripheral hardware module and the corresponding control procedures, constitutes a complete electronic nose system. The electronic nose device with simple operation is accurate and rapid to detect variety of gas in real time.the chip of ARM9selects Samsung S3C2440A and Linux operating system version is2.6.30.4.The hardware of the systerm consists of ARM9core control module, automatic air intake device, closed chamber, the sensor array, data acquisition modules and USB data storage module. Through a variety of gas sensors to test, six different characteristics of gas sensitive sensors are slected to form the core of the electronic nose:cross gas sensor array. The system software is divided into the machine software and PC software. The electronic nose equipped with a Linux operating system of Linux kernel version2.6.30.4, In Linux, drivers and applications of the corresponding modules, as well as touch screen control interface are complete. The real-time sampling control software, and electronic noses data analysis software constitute the PC software. Real-time sampling of the control software can be achieved on the gas inlet, sampling data, real-time monitoring, data storage and other functions.The PC software, control software, real-time sampling to constitute recognition software and electronic nose data analysis. Real-time sampling of the control software can be achieved on the gas inlet, sampling data, real-time monitoring, data storage and other functions.The electronic nose data analysis and recognition software written by Matlab software integrating a variety of eigenvalue extraction algorithm can achieve the purpose to achieve the purpose of distinguishing different smell by choosing a different characteristic values of the sample and classification pattern recognition algorithms.Finally, the electronic nose system is sampled in three different Chinese herbal mint, Bupleurum and motherwort. Eight training samples in each species of medicinal herbs import the odor-recognition software. Seven kinds of eigenvalues and PCA+LDA classification algorithm are selected in the software. By using Mahalanobis distance four unknown samples of each type of Chinese herbal medicines, a total of12test samples are identified and the correct rate is100%.