Method for predicting PM2.5 concentration of regional air

Abstract

The invention discloses a method for predicting the PM2.5 concentration of regional air. The method comprises the steps that firstly, training sample data of a support vector machine regression model to be trained are constructed through historical data, then the trained support vector machine regression model is obtained through the training sample data, and the trained support vector machine regression model is treated as a PM2.5 concentration prediction model; then a particle swarm optimization algorithm is combined with the PM2.5 concentration prediction model, through the continuous optimization and iteration of the particle swarm optimization algorithm, input parameters of the PM2.5 concentration prediction model are reconstructed continuously through the particle positions till the final global polarity of a particle swarm is obtained after iteration is completed, an input parameter of the PM2.5 concentration prediction model is reconstructed with the position of the particle corresponding to the final global extreme value of the particle swarm, and when the input parameter is input into the PM2.5 concentration prediction model, the obtained output is considered as the PM2.5 concentration. The method has the advantages that the dimensionality of the input parameters of the PM2.5 concentration prediction model can be lowered, and the PM2.5 concentration prediction accuracy can be improved.

Description

technical field [0001] The invention relates to a PM2.5 concentration prediction method, in particular to a regional air PM2.5 concentration prediction method. Background technique [0002] With the rapid advancement of industrialization, the phenomenon of atmospheric smog is becoming more and more serious. PM2.5 is one of the main culprits causing smog. Its particle size is small, can be suspended in the air for a long time and spread, and can carry toxic Harmful substances enter the respiratory tract and lungs, posing a direct threat to human health. For this reason, it is of great significance to predict the concentration of PM2.5 based on the historical monitoring data of the atmosphere. However, there are many factors affecting the concentration of PM2.5, such as: instant industrial emissions, the concentration of pollutant particles in the atmosphere, meteorological conditions, seasons and sunlight radiation, and it is challenging to predict its concentration. [000...

AI Technical Summary

Problems solved by technology

Whether it is an SVR model or an artificial neural network model, the time series characteristics of variables need to be considered in the prediction of pollutant concentration, and the time lag factor must be considered in the input parameters of the model, so that the input parameters of the model have high dimensions and the training sample data is insufficient. The prediction accuracy of the model is not high

To this end, some scholars proposed a genetic-neural network (GA-ANN) model, GA-ANN uses the neural network (ANN) model to predict and model the concentration of atmospheric pollutants, and uses the genetic algorithm (GA) to predict the The index selects 0 or 1 for the input parameters, which reduces the dimension of the input parameters and avoids the overfitting of the model to a certain extent, but at the same time loses useful information of some input parameters, and the prediction accuracy is still not high

Images