A Stochastic Approach to the Analysis of Pollutant Emissions and Fuel Consumption of Motor Vehicles Driven in Real Traffic

In real road traffic, combustion engines of motor vehicles operate in dynamic conditions. Under such conditions, significant time variability in the values describing engine operations is observed, especially in terms of rotational speed and torque. Therefore, it is possible to model such conditions as probabilistic and to treat the properties of combustion engines in these conditions as stochastic processes. This paper presents a stochastic approach to the analysis of pollutant emission and fuel consumption test results of a motor vehicle driven in real traffic conditions. The empirical data were obtained from tests conducted on a car with a spark-ignition engine equipped with mobile on-board measuring equipment. The scope of the investigations covered the domains of time, frequency and process values. In the time domain, statistical characteristics of the processes were analyzed to explore potential correlations between them. In the frequency domain, the power spectral density of the processes was determined. In the process values domain, the emphasis was placed on examining the probability density of processes. A large diversification of the determined characteristics was found, in particular for vehicle velocity, engine operating states and the processes of pollutant emissions and fuel consumption. For practical reasons, the results of the correlation studies were particularly valuable, as they enabled assessment of the effects of taking action to reduce emissions of various pollutants.