An Intelligent Air Quality During COVID-19 Prediction and Monitoring System Using Temporal CNN-LSTM

Abstract

Recently, air pollution is becoming a very big issue in metropolitan cities due to the availability of vehicles. This air pollution has spread various breathing-related diseases to people vigorously in recent years. People are affected much more severely due to the lack of knowledge and updates about the air pollution level in the specific area and dates. To safeguard the people and to provide enough knowledge about the air pollution level, many researchers have developed many air qualities during COVID-19 in prediction systems using various classifiers. Even though no system resolves this issue due to the rapid growth of the population, automobile usage, and industries. To fulfil the current requirements, this paper proposes an Intelligent Air Quality During COVID-19 Prediction and Monitoring System (IAQPMS) for predicting and monitoring the air quality during COVID-19 in a specific city in every season. The proposed system uses a newly proposed data grouping algorithm called Rule-based Data Grouping Algorithm (RDGA) for grouping the data based on the different seasons by applying the suitable rules and the newly developed ensemble deep classifier (EDC) that combines the standard bidirectional LSTM (Bi-LSTM) and the existing CNN with temporal constraints (T-CNN) to perform the prediction process. The PM2.5 observation in Beijing, China forecasted the air pollution level seasonally for the next 3 years and also frequently updated the society on the status.The proposed system conducted various experiments and also proved to be better than the available prediction systems in terms of Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) values.