City monitoring station air quality prediction method based on online multi-core regression

Abstract

The present invention relates to a city monitoring station air quality prediction method based on online multi-core regression. The method comprises: firstly, based on historical data, extracting multi-field features of a city monitoring station, such as a forecasting meteorological feature, a real-time meteorological feature, a traffic feather, local and peripheral city air pollutant feature and the like; then, based on the extracted features, training a multi-core regression model, and performing online adjustment on the multi-core regression model by using new data; and finally, based on the adjusted model, predicting the air quality of the monitoring station hour by hour in a future period of time. The method can be used for accurately and efficiently predicting the air quality of the city monitoring station and has a guiding effect on environment protection and public life.

Description

technical field [0001] The invention relates to the field of air quality monitoring, in particular to an air quality prediction method for urban monitoring sites based on online multi-core regression. Background technique [0002] Air is the material on which living things on earth depend. Air quality is closely related to people's daily life and plays an important role in the comprehensive evaluation of urban environment. However, with the development of human civilization and economy, air pollution is becoming more and more serious. How to improve air quality and accurately predict air quality is becoming more and more important. According to the air quality prediction results, people can take corresponding measures (such as wearing masks, avoiding going out, etc.) to avoid being harmed by air pollutants. On the other hand, environmental protection is the cause of the whole society, and the degree of public participation in environmental protection is an important indica...

AI Technical Summary

Problems solved by technology

However, the traditional air quality prediction method only considers the characteristics of related fields such as meteorology, traffic, and local air pollutants, and does not consider the impact of the air quality status of surrounding cities on the predicted city, thus affecting the accuracy of the prediction results.

On the other hand, traditional air quality prediction methods use relevant features based on real-time meteorological data in the model training stage, and relevant features based on forecasted meteorological data in the prediction stage, and both real-time and forecasted meteorological data are related to air quality. give simultaneous consideration, otherwise it will affect the effectiveness of the model

[0005] From the perspective of the model, due to the inherent defects of the production model (such as Markov model, etc.) with label bias and independence assumptions, the prediction accuracy is not ideal; while the discriminative model (such as decision tree, support vector Although it is simpler than the production model, due to its black-box operation, it cannot clearly show the relationship between the data, so it cannot reflect the characteristics of the training data itself, and thus has a negative impact on its predictive ability.

Although the conditional random field model not only has the advantage that the discriminant model is easier to learn, but also considers the transition probability between context labels like the production model, but it is the same as the traditional production model and the discriminant model. Batch type Learning method, when there is new data, it needs to be retrained based on all data

Due to the high cost of retraining, it is difficult to update the model in time

Although online single-kernel regression can overcome the above-mentioned shortcomings of the batch processing model, it often fixes a kernel function before the learning task. If the data flow changes unstable over time, it will lead to unsatisfactory prediction results.

Images