A Data-Based Simulator for Predicting Vehicle Fuel Consumption

A hybrid system, combining on-road testing and statistical modeling, has been developed which permits calculation of a vehicle's fuel consumption over any driving cycle. An instrumentation system permits collection of the necessary on-road and laboratory test data within a relatively short period of time and, once calibrated, the model requires considerably less CPU time for execution than engineering simulation models while providing results within 5% of on-road measured fuel uses. The model has been run in conjunction with optimization routines for deriving minimum fuel use strategies for a variety of common driving maneuvers. Results for a 1979 Ford Fairmont wagon are given.