Trip Based Modeling of Fuel Consumption in Modern Heavy-Duty Vehicles Using Artificial Intelligence

Heavy-duty trucks contribute approximately 20% of fuel consumption in the United States America (USA). The economy heavy-duty vehicles (HDV) is affected by several real-world parameters like road parameters, driver behavior, weather conditions, and vehicle etc. Although modern comply with emissions regulations, potential malfunction engine, regular wear tear, or other factors could affect performance. Predicting per trip based on dynamic on-road data can help automotive industry to reduce cost time for testing. Data modeling easily diagnose reason behind a knowledge input parameters. In this paper, an artificial neural network (ANN) was implemented model predicting total instantaneous very few key such as engine load (%), speed (rpm), (km/h). Instantaneous predict patterns optimized fleet operations. work, used collected at frequency 1Hz during testing West Virginia University Center Alternative Fuels Engines Emissions (WVU CAFEE) using portable monitoring system (PEMS). performance evaluated mean absolute error (MAE) root square (RMSE). further from trip. study shows that networks performed slightly better than machine learning techniques linear regression (LR), random forest (RF), high R-squared (R2) lower error.