Fuel consumption estimate based on meso traffic characteristics and typical road environment

Vehicular fuel consumption and the consequent environmental problems in metropolis attract more and more attention. However, how to monitoring the distribution of fuel consumption dynamically in the urban road network is still a thorny problem to be solved. This paper developed a fuel consumption model based on combination of multiple meso driving characteristics (e.g. average speed, average acceleration). The combination of driving characteristics is represented in the form of feature vector in this paper. Feature vectors with same fuel consumption correspond to a driving pattern. Fuel consumption is calculated based on the distribution of driving patterns. The experimental results show that this method can achieve a high accuracy rate. Considering that typical road environment can centralize the distribution of driving patterns to improve the accuracy, this paper takes the intersection with traffic lights as the example to analyse the effects of its special traffic characteristics on fuel consumption.