Smart prediction and optimization of air quality index with artificial intelligence

Air pollution, a critical environmental issue, necessitates urgent action. It originates from both human activities, like industrial emissions and vehicle pollution, and natural events such as sandstorms, leading to increased atmospheric pollutants such as sulfur dioxide (SO₂), nitrogen dioxide (NO₂), ammonia ion (NH₄⁺), black carbon, ozone, and fine particulate matter (PM_2.5). Leveraging China's extensive air quality monitoring data, artificial intelligence (AI) was used in this study to enhance air quality prediction and management. The study aims to utilize the vast air monitoring data more effectively by developing advanced air quality assessment methods and AI models. An AI-based method presented in this study was applied to train extensive air quality data, enabling an intelligent air quality index (AQI) that swiftly and accurately reflects air quality status, to assess impacts on sensitive groups, and to predict future trends. This smart prediction and optimization (SPO) approach not only utilizes existing monitoring network data efficiently but also offers precise future air quality forecasts, providing valuable strategies for pollution prevention and air quality improvement. Data on various pollutants were collected from four regions in China between August 2021 and July 2022, using diverse modeling techniques and machine learning methodologies. The models achieved a high accuracy level of around 99 %, indicating the significant portion of air quality that falls into the unhealthy category, especially impacting sensitive groups and reflecting the adverse atmospheric conditions in the studied regions.