Enhancing Emitter Prediction in Heavy Duty Vehicles Using Recurrent Convolutional Neural Network

Predicting emissions from Heavy Duty Vehicles (HDV) accurately is crucial for environmental monitoring and regulatory compliance. These vehicles, particularly diesel trucks, significantly contribute to urban air pollution, Particulate Matter (PM), and Nitrogen Oxide (NOx) emissions. Effective vehicle emissions control and detection are crucial for protecting public health. Regular vehicle inspections typically conducted every six months or annually, aim to identify and filter out high-emission vehicles at testing stations. This research compares the performance of Faster Mask Recurrent Convolutional Neural Networks (FMRCNNs) to existing in predicting vehicle emissions, aiming to improve prediction accuracy and enable more effective monitoring and control. This study presents a novel approach using FMRCNN that leverages convolutional layers for spatial feature extraction and recurrent layers for modeling temporal dependencies. The FMRCNN model is thoroughly evaluated against CNN and LSTM models using a comprehensive dataset of heavy-duty vehicle emission measurements. The results show that the FMRCNN model significantly outperforms CNN and LSTM models in prediction accuracy and computational efficiency, making it a superior choice for emission prediction tasks. These findings demonstrate the potential of FMRCNN to improve emission predictions for HDV, providing a more accurate and reliable tool for environmental monitoring and helping to ensure compliance with regulatory standards. FMRCNN can reduce harmful emissions and promote a cleaner environment by enhancing emission prediction capabilities.