PM2.5 prediction model based on multi-source heterogeneous data fusion

Abstract

The invention provides a PM2.5 prediction model based on multi-source heterogeneous data fusion, and the PM2.5 prediction model achieves the integration of information through the fusion of multi-source heterogeneous data, and builds a multi-core support vector regression machine model on the basis of the fusion data to achieve the prediction of PM2.5 concentration. The PM2.5 prediction model includes the steps: firstly, collecting two kinds of multi-source heterogeneous data including air quality data and images, and preprocessing and feature extraction are conducted on the collected data; secondly, completing multi-source heterogeneous data fusion by using a multi-kernel expansion method based on a kernel function, wherein the fusion process is mainly completed by constructing and combining Gram matrixes, and a multi-kernel expansion kernel function is derived; then, on the basis of the multi-kernel extended kernel function and the extended kernel matrix, reconstructing a multi-kernel support vector regression machine model; and finally, optimizing the model parameters by using an improved minimum sequence optimization algorithm. According to the PM2.5 prediction model, PM2.5 prediction is realized based on multi-source heterogeneous data fusion; more comprehensive and credible judgment can be obtained on the basis of realizing information fusion; and the prediction accuracy,stability and credibility are ensured.

Description

technical field [0001] The present invention is a PM based on multi-source heterogeneous data fusion 2.5 The prediction model is mainly used in related work such as air quality monitoring and early warning, and belongs to the field of environmental monitoring technology. Background technique [0002] The rapid economic development and the advancement of industrialization have brought unprecedented pressure on the ecological environment. For a long time, the hidden dangers accumulated in the economic development model that focuses on economic growth and neglects environmental protection have gradually emerged. Under the circumstance that the pollution of inhalable particulate matter and total suspended particulate matter cannot be completely solved, the long-term and sustainable development of economically developed regions such as Beijing, Tianjin and Hebei Smog poses a serious threat to people's life and production activities. In order to effectively prevent and control t...

AI Technical Summary

Problems solved by technology

However, due to unrepairable data caused by some human factors or accidents, PM based on modeling 2.5 Forecasting is more difficult

In addition, image-based PM developed in recent years 2.5 Forecasting method, although it gets rid of the shackles of large-scale equipment in data collection, but the prediction accuracy is not satisfactory

Images