Modeling Diesel and Hybrid Bus Fuel Consumption with Virginia Tech Comprehensive Power-Based Fuel Consumption Model: Model enhancements and Calibration issues

Abstract

The goal of this paper was to develop a calibration procedure and use it to estimate diesel bus fuel consumption and carbon dioxide emission levels. There are few models for estimating those values. Available models require dynamometer data to calibrate model parameters and produce a bang-bang control system (optimum control entails maximum throttle and braking input). The only diesel fuel consumption model that does not suffer from these deficiencies is the Virginia Tech comprehensive power-based fuel consumption model (VT-CPFM). VT-CPFM can be calibrated with publicly available data from the Altoona Bus Research and Testing Center. However, each bus is slightly different because it is built and tuned for the specific transit agency. Consequently, research presented in this paper enhanced the VT-CPFM for modeling diesel buses and developed a procedure for calibrating bus fuel consumption models by using in-field data. All models produced a good fit to the in-field data with a coefficient of determination ( R2) greater than .936, and the sum of the mean squared error for each quarter of a second was less than 0.002. Validation found an average error of 17.55% in total fuel consumed during the validation portion of the test. However, for tests with air-conditioning on, the average error was 10.82%.