ABSTRACT
Improved fuel economy over the past 20 years has been confirmed by the Nebraska Tractor Test Laboratory (NTTL). According to the ASAE Standards, there was a 4.9% decline in the average annual specific volumetric fuel usage. Diesel engine fuel consumption predictions under partial and full loads, and at reduced engine speeds from full throttle, were achieved by developing new fuel consumption equations based on power and fuel consumption data from the NTTL reports. Model training, feature expansion, and preprocessing are the three most critical steps. Additional input elements could be incorporated during preprocessing using data collected from the various sources mentioned in the preceding chapter. Feature extraction makes use of fuel consumption projections. The model's accuracy was enhanced during training with the RBF-MLP-SVM. With a rate of 97.35%, the proposed method outperformed state-of-the-art technologies like MLP and SVM.
