Air quality prediction method based on time sequence convolution network algorithm

Abstract

The invention relates to an air quality prediction method based on a time sequence convolutional network algorithm, and aims to solve the problem that an accurate prediction model is difficult to establish by an existing air quality prediction method. The method comprises the following steps: 1) acquiring historical data from a plurality of observation stations in one region and preprocessing the historical data; 2) taking historical data of the observation station to be predicted in a past first time period as a training set, taking a historical data mean value of the other observation stations at the same moment in a past second time period as a verification set, and enabling the sample number ratio of the training set to the verification set to be 6: 4-8: 2; normalizing the data of the training set and the verification set, and converting the normalized data into a three-channel format; 3) establishing a TCN model composed of causal convolution, expansion convolution and five layers of residual blocks; setting model hyper-parameters, and training the TCN model by using the training set and the verification set; and 4) taking historical data of several hours before the current moment as input, and performing reasoning by utilizing a training result to obtain air quality prediction values of several hours in the future.

Description

technical field [0001] The invention relates to an air quality prediction method based on a time series convolutional network algorithm. Background technique [0002] In recent years, the problem of air pollution has become a hot issue in society. In terms of air pollution control, real-time grasp of air pollution and mid- and long-term trend forecasts for air quality can provide a reliable basis for air pollution control. [0003] Traditional air quality prediction mainly includes numerical model method and statistical forecasting method. Since air quality is affected by the coupling of multiple factors, the influencing mechanism is very complicated. Therefore, it is difficult to establish an accurate prediction model by using the traditional single air quality prediction method. Contents of the invention [0004] The purpose of the present invention is to solve the problem that the existing air quality prediction method is difficult to establish an accurate prediction mo...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the problem that the existing air quality prediction method is difficult to establish an accurate prediction model, and provides an air quality prediction method based on time series convolutional network algorithm

Images