Design And Application Of Air Quality Prediction Model Based On Time Series Analysis

Deterioration of air and haze occur frequently raises continuing concerns about air pollution, controling pollution levels and variation of air pollutants、analyzing and predicting trends of pollution become particularly important. Time Series named ARMA and GARCH model prediction method are based on rigorous mathematical basis of the theory, having many features, such as a simple structure, forecasting fast, easy operation. Compared with other time series analysis and forecasting methods, like regression analysis, gray model and neural network, it is more suitable for practical applications. The topic of this study is about intelligent analysis and forecast of environmental protection, and build smart environmental analysis platform and provide a reliable basis for forecast and early warning of environmental protection.Different air pollution data in NingXia was taken for the study, It depends on ARMA, GARCH model to analyze and forecast air pollution data, afterwards, establish a forecast model. At the same time, its in-depth analysis and research was applied to the environmental protection intelligence platform module. The main contents completed as follows:(1) We used mathematical statistics analysis, correlation analysis and wavelet analysis to integrate the major air pollutants in six cities of Ningxia Hui Autonomous Region, then we showed the law on seasonal variation of air pollutants, monthly changes, hourly change regularly with time scale and used data visualization method for in-depth analysis of the variation, the pollution situation and reasons.(2) For the sake of predicting air pollutant and improving the comprehensive evaluation method of air quality, we used the time series named ARMA and GARCH algorithm:Time series model includes data preprocessing, model identification, parameter estimation, prediction part analysis, error analysis, and gives a detailed analysis of the processes and methods of algorithm.(3) For the lack of systematic software implementation issues, this study took Django as the core technology to develop intelligent analysis of environmental quality platform. The system used the Python language to develop, and realized the effective combination of three major module, namely the model (Model), the template (Template) and views (Views), thus it ensured the efficient operation of the entire system as well as software upgraded, software reused and so on in the future.