FORECASTING THE ECOLOGICAL SITUATION USING NEURAL NETWORKS

Abstract

Atmospheric air is a vital component of the natural environment, an integral part of the human, plant, and animal habitat. Ambient air quality is the most important factor affecting health, sanitary and epidemiological situations. With industrial growth, environmental issues and environmental management are revived and take on new significance. To effectively solve these problems, it is necessary to create modern environmental monitoring systems. In this article, we have applied artificial neural networks to predict PM2.5 concentrations as determinants of smog. We used meteorological data and PM2.5 concentrations to create these networks. PM2.5 data and concentrations at several points in the city of Almaty were used as input data for training the model. The measurements were carried out over three months (February-March) from 2019–2021. The best results were shown by a recurrent neural network with long short-term memory, which has proven to be effective in predicting this type of data.