【正文】 f the quality of black tea in India Bipan Tudu to do related research. Based on neural network in dealing with the main problems of sensor array data: the quality of black tea, tea from different tea gardens picked by different climatic conditions, farm type, season and planting methods are quite different, and therefore, the entire database can not be readily be met difficult to train a neural network model. The study proposed to enhance the learning characteristics of RBF neural network pattern recognition algorithms to classify the smell of black tea, and get a better recognition effect. In addition, the Task Force also uses an electronic nose to predict the optimum fermentation time for black tea. Wang Jun, et al Zhejiang University to study electronic nose applications in various types of food, such as Longjing tea, wheat, eggs, etc., have achieved better results. Luo Dehan such as China Science and Technology University of Electronic Nose to the counterfeit cigarettes identification, Huazhong University of Science and Technology, Xie Changsheng electronic nose to identify 17 different brands of domestic vinegar.MedicineWith the development of electronic nose technology and its application in the field of medicine is being increasingly popular. ERThaler in electronic nose applications early, in order to verify whether it applies to the field of medicine do exploratory test. The experiments showed that the electronic nose can better distinguish the cerebrospinal fluid and plasma, high accuracy, you need a small amount of sample. The experiment took the first step to explore the feasibility of electronic nose applications in the field of medicine. Clinically, the electronic nose for evaluation of halitosis。 2) Determination of the sample and data analysis: sampling operation unit the initialization of the sensor array is exposed to the odor body, when the contact with the surface of the active material of the volatile pounds (VOC) and the sensor (ultrasonic sensor), to produce the transient response. Such a response is recorded and transmitted to the signal processing unit for analysis, and stored in the database of a large number of VOC pattern parison, identification, and to determine the type of odor。s sense of smell. However, with the development of science and technology, more authoritative some universities in the world have developed electronic nose has a wide range of applications, most notably to the number of the University of Hamburg, Germany, has absolute authority in the sensor field in the world today. Electronic nose is the response pattern to identify the odor of the gas sensor array electronic systems, it can be in a few hours, days or even a few months time within a continuous, realtime monitoring of the specific location of the odor condition. Identifying odor main mechanism of the electronic nose is that each sensor in the array has a different sensitivity of the measured gas. The core of the device of the electronic nose gas sensor. Gas sensors based on the principle of different type of metal oxide, electrochemical, and conductive polymer type, quality, photoionization type can be divided into many types. Currently the most widely used is a metal oxide.How it works: Electronic nose is mainly posed of three parts of a gas sensor array, signal preprocessing and pattern recognition. Presented in front of an active material of the sensor, an odor sensor chemical input is converted into an electrical signal by a plurality of sensor response to an odor they constitute the sensor array to the odor of the response spectrum. Obviously, the various chemical ponents in the odor sensitive materials play a role in this response spectrum for a broad spectrum of odor response spectrum. To achieve the odor qualitative or quantitative analysis, the sensor signal must be appropriate pretreatment (to eliminate noise, feature extraction, signal amplification, etc.), using a suitable pattern recognition analysis method for processing thereof. Theoretically, each odor will have its characteristic response spectrum, according to its characteristic response spectrum can distinguish between small same odor. While gas sensors constituting the array to measure the crosssensitivity of a variety of gases, by a suitable analytical method, the mixed gas analysis. The electronic nose is the use of various gas sensing device has a response to this characteristic but different from each other, with the data processing methods to identify a variety of odor, odor quality analysis and evaluation of plex ponent gases. The main mechanism of the electronic nose identify each sensor in the array has a different sensitivity of the measured gas, ., high response One gas may be generated on a sensor, while the other sensor is a low response。 It is a statistical analysis method to grasp the principal contradiction of things, you can parse out the main influencing factors from diverse things, to reveal the nature of things, simplify plex problems. PCA as a linear feature extraction technology, designed to exploit data dimensionality reduction idea that the calculation of the main ponents and highdimensional data is projected onto a lowerdimensional space, multiindicators into a few indicators, so as much as possible to show the information contained in the original data. PCA in electronic noses for an objective analysis of the differences between the samples.cluster analysis (CA)Cluster analysis is a kind of subjects were divided into relatively homogeneous groups of statistical analysis techniques. From a statistical point of view, the cluster analysis is a way to simplify the data through data modeling. CA is classified based on the relationship between the number of the individual or variable the objectivity strong, but various clustering methods can only be achieved under certain conditions the local optimum。 also used to assessment of the unpleasant smell on the pill, the patie