Road vehicle energy consumption model by using neuro-fuzzy approach

The aim of this work is to compare the efficiency of fuzzy logic with that of a traditional mathematical method (based on least squares) to determine the road vehicle fuel consumption model by using only four measured variables of a vehicle (bus) driven in heavy city traffic conditions. Results of fuel consumption measurements, (in actual service of city bus expressed in terms of volumetric fuel consumption per units covered distance), are presented and correlated to a set of free variables which represent the vehicle average speed (taken into account with its reciprocal), number of passengers on board (or better, the ratio between the actual number of passengers to the passenger capacity of the bus) and the actual elevation of the road that is considered as the average up-hill-slope and the average down-hill-slope over the route section. Fuzzy logic is shown to be much more efficient in correlating measured and simulated data