Investigation of road transport-based greenhouse gas prediction models and the use of ıntelligent transportation systems for emission reduction

The study analyzed road transport-related CO₂ emissions in Dilovası, Kocaeli, using artificial intelligence (AI)-based models, highlighting the transportation sector's major contribution and the need for effective mitigation strategies. Two models, the Adaptive Neuro-Fuzzy Inference System (ANFIS) and the Artificial Neural Network (ANN), were developed using vehicular and meteorological data to predict nitrogen oxide (NOₓ) and carbon monoxide (CO) emissions and were compared with the results of the Calculations of Emissions from Road Transport (COPERT 4) model. The ANFIS model achieved high accuracy, with mean squared error (MSE) values of 0.0003 for NOₓ and 0.0000 for CO, root mean squared error (RMSE) values of 0.0178 and 0.0092, and mean absolute error (MAE) values of 0.0115 and 0.0055. In contrast, the ANN model produced higher errors with MSE values of 0.0027 and 0.0009, RMSE values of 0.0523 and 0.0310, and MAE values of 0.0297 and 0.0156. Determination coefficients (R²) were 0.9993 and 0.9997 for ANFIS, compared with 0.6366 and 0.8506 for ANN. The results, consistent with previous studies, confirmed the superior performance of ANFIS and demonstrated that integrating AI-based modeling with Intelligent Transportation Systems offers an effective approach for emission reduction and sustainable transport management.