INTELLIGENT SYSTEM FOR FUEL CONSUMPTION PREDICTION USING MACHINE LEARNING

Abstract

Modern vehicles are increasingly dependent on Electronic Control Units (ECUs) that regulate and monitor critical subsystems, such as the engine, transmission, and braking systems. With the integration of the Internet of Things (IoT) and advanced sensor technologies, ECUs generate vast amounts of real-time data, including vehicle speed, engine output, throttle position, gear selection, acceleration, engine load, and fuel consumption data. Leveraging these data, the present study investigated the application of machine learning (ML) techniques for two key objectives: (i) real-time prediction of fuel consumption and (ii) classification of driving behavior profiles categorized as aggressive, moderate, and economical.The research methodology comprised data preprocessing, feature selection, and cross-validation to ensure the accuracy, robustness, and generalizability of the results. Among the evaluated models, Ridge Regression demonstrated strong predictive capability owing to its effectiveness in addressing multicollinearity within high-dimensional data. The experimental results revealed enhanced accuracy in both fuel consumption prediction and driver profile classification, validating the feasibility of machine learning approaches in this domain.The findings highlight the potential of ML-driven vehicle data analytics for advancing fuel efficiency, lowering operational costs, and mitigating environmental impacts, such as carbon emissions. This study contributes to the development of intelligent transportation systems and smart mobility solutions, where real-time predictive modeling can support eco-driving strategies, adaptive vehicle control, and sustainable fleet management.