Study On Monitoring System Of Air Quality Based On Electronic System

Nowadays, indoor air quality problem has already become the focus of people's attention. Indoor air pollutants include formaldehyde, benzene, volatile organic compounds, nitrogen oxides, etc. Indoor air quality's improvement is the premise of testing indoor air pollutants, based on the air quality of testing system of electronic nose with real-time, convenient and efficient, etc.E-nose, also called artificial olfactory system, is mainly composed of a gas sensor array, a signal preprocessing unit and a pattern recognition algorithm. Applying e-nose's detection of hazardous gas, this paper investigates the hardware design of e-nose system, selection and optimization of sensor arrays and differentiating capability.Hardware's design of e-nose system for air detection. According to real-time performance, stability, and portability, it focuses on the system's control module, signal conditioning module, sampling and A/D conversion module, display module circuit design and drive programming.Selection and optimization of sensor arrays. According to metal-oxide sensor cross-sensitivity and selectivity characteristics, TGS2201, TGS2602, TGS2620 and GSBT11 are selected as metal oxide sensors for indoor air pollution,and oxygen sensor is also used to detect concentration of oxygen content because it is also an important indicator of indoor air quality. To meet actual needs, temperature sensor, humidity sensor and pressure sensor are chosen too. The optimization of the sensor array is to eliminate redundancy and correlation, and enhance the differentiating capability of the system. Based on the analysis of the common optimization method of the gas sensor array, a sensor-array optimization method is proposed based on adaptive genetic algorithm, which realizes the elimination of redundancy and correlation and achieves the optimization of sensor array by setting the significant coefficients of the sensors.Established recognition algorithms based on qualitative analysis and quantitative analysis. Principal component analysis is a successful data compression method, which can be used to reduce the dimensionality of array signals. Probabilistic neural network classifier, under certain condition that the a priori probabilities of all classes are equal, is equivalent to the Bayes classifier based on the kernel function, and can achieve minimum error in classification on optimal classifier. In the paper, the dimension of the array signal is reduced firstly, and then sent into the probabilistic neural network. The results from quantitative analysis show that the probabilistic neural network classifier can achieve the accurate classification of the formaldehyde, carbon monoxide and nitrogen dioxide. Furthermore, it discusses the quantitative analysis of harmful gases, using radial basis function neural network to identify, and the identification of formaldehyde, carbon monoxide and nitrogen oxides shows that this algorithm can obtain higher accuracy.