International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 11 Issue: 06 | Jun 2024
www.irjet.net
p-ISSN: 2395-0072
Air Pollution Prediction System in Smart Cities Using Data Mining Technique Debopriya Manna1, Rohan Mondal1, Arpan Sanyal1, Ahana Biswas1, Hritam Roy1 and Subhajyoti Barman2 1B. Tech, Dept. of Computer Science and Engineering, Techno India University, Kolkata, West Bengal, India. 2 Assistant Professor, Dept. of Computer Science and Business Systems, Techno India University, Kolkata, West
Bengal, India. ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - In every nation, Air Pollution has grown to be a
contaminating the indoor and outdoor air. Health and climate are being compromised due to air pollution. It is becoming a major threat to the environment. The major contributor to air pollution is fine particulate matter which is leading to lung cancer, acute and chronic respiratory disorders, heart attacks, and strokes.
serious problem. Among mankind's most serious issues, pollution in the air is one of the major contributors to health and climate change. Fears over rising air pollution, that could harm human health, the health of every living creature, and the advancement of the world economic growth, are shared by both the government and the general population. For this reason, air pollution forecasting has become crucial. To address this issue, several deep learning models were applied. AQI was calculated to conduct this study. Various pollutioncausing gases like nitrogen dioxide, sulphur dioxide, carbon monoxide and ozone and particulate matter like PM10 and PM2.5 were studied. The vast quantity and diversity of data collected by air pollution monitoring stations across different cities have made air pollution forecasting an important topic. This research incorporates an LSTM (Long-Short Term Memory), ARIMA, Prophet Linear Regression, and Polynomial Regression. The dataset is primarily comprised of different pollution-causing components data collected from the Central Pollution and Control Board (CPCB). Our project aims to investigate the weather in several Indian cities and record their AQI level. Using additional variables, we attempt to determine the extent of air pollution. Eventually, we base our findings on the historical pattern of the graph. Based on the previous year’s pollution data, we attempt to forecast how the weather will evolve over the following several years. Different time series models were studied and the bestsuited model was decided based on different parameters. This paper utilises ARIMA, SARIMA, Prophet, Long short-term memory (LSTM), Linear Regression and Polynomial Regression. These models aim to find the future trend for the upcoming months. Root Mean Square Error (RMSE) was used as performance metrics to evaluate the models, along with it Mean Absolute Percentage Error (MAPE) was also utilised to assess the models.
Vehicles, domestic energy use for heating and cooking, and other sources are the main contributors to outdoor pollution, power generation, agriculture/waste incineration, and industry. The primary air pollutants are PM 10 with a diameter of less than 10 microns and PM 2.5 is more hazardous as it has a diameter of less than 2.5 microns. Sulphur Dioxide (SO2) and PM 2.5 are caused due to unburned fuel and also by processed byproducts. Due to fuel combustion Nitrogen Dioxide (NO2), Ozone (O3) and Carbon Monoxide (CO) are produced. CO is the most dangerous one and it is also known as a silent killer. It deprives the brain and heart of oxygen that is required by the body to function, by entering our blood cells directly and replacing the oxygen in our body. When there is an increase in the pollutant levels it leads to humans losing consciousness and vomiting. When the exposure to these harmful pollutants is too long, it can eventually damage the brain cells in the body or can cause death. The government of India uses PM 10 and PM 2.5 as the major criteria for Air Quality Index (AQI) calculation. The calculation of AQI comprises a minimum of three parameters out of which one must be either PM 2.5 or PM 10. The calculation of sub-indices requires 16 hours of data. For the Calculation of AQI, the Sub-indices for individual pollutants at a monitoring location are calculated using their 24-hour average concentration value (8 hours in the case of CO and O3) and health breakpoint concentration range. The sub-index that will be worst, is the AQI for that location.
Key Words: LSTM, ARIMA, RMSE, MAPE, AQI.
Equation :
1. INTRODUCTION
Isi = [((Cobs – Cmin) (Imax – Imin)) / Cmax – Cmin] + Imin
In the last few years, there have been constant changes in the environment which led to degrading air quality due to the presence of various harmful air pollutants. Air pollution is the presence of harmful pollutants in the air that are
© 2024, IRJET
|
Impact Factor value: 8.226
Where, Isi = Sub-index value of observed pollutant
|
ISO 9001:2008 Certified Journal
|
Page 783