Skoči na glavni sadržaj

Izvorni znanstveni članak

https://doi.org/10.31803/tg-20231016114033

Forecasting Daily Air Quality Index and Early Warning System for Estimating Ambient Air Pollution on Road Networks Using Gaussian Dispersion Model with Deep Learning Algorithm

Asha Unnikrishnan orcid id orcid.org/0000-0003-2998-9126 ; Department of Computer Science, Sree Saraswathi Thyagaraja College, Palani Road, Pollachi, Tamil Nadu 642107, India *
S. Rajeswari ; Department of Computer Science, Sree Saraswathi Thyagaraja College, Palani Road, Pollachi, Tamil Nadu 642107, India

* Dopisni autor.


Puni tekst: engleski pdf 2.727 Kb

str. 549-559

preuzimanja: 62

citiraj


Sažetak

The rapid growth of the vehicle population is a major factor in heavy air pollution and public health issues. Traffic-related air pollutants (TRAPs) on roads are often much higher than ambient values, leading to high exposure levels in vehicles. This research proposes a hybrid forecasting model for early detection and early warning systems (EWS) of road networks during real-world travels. Data is collected from Kannur, Calicut, Palakkad, and Coimbatore using real-time sensors, including surrounding discussion information, activity information, vehicle speed, and stopping events. The study predicts ambient air quality (AAQ) levels on the road network using the Gaussian Dispersion model (GDM) and measures the risk sensitivity of PM10 and PM2.5 in selected regions. This helps formulate powerful prevention strategies and prevent negative health impacts. The air pollution module for predicting concentration has an innovative hybridization model that combines an improved cuckoo search (CS) and differential evolution (DE) algorithm with a stacked LSTM model to increase forecasting accuracy of six major environmental pollution levels. This model predicts the AAQ level and is effective and robust for warning one day before the pollutant event occurs based on the risk level of an ambient air pollutant from the RN.

Ključne riječi

Air Quality Index (AQI); Ambient Air Pollution (AAP); EWS; GDM; Hybrid Forecasting Method; LSTM; Road Networks

Hrčak ID:

321395

URI

https://hrcak.srce.hr/321395

Datum izdavanja:

5.12.2024.

Posjeta: 201 *