Modelling of IoT based Vehicular Emission through Regression Analysis

Authors

  • S. Magesh, Chandar G

Abstract

Emissions from vehicles such as cars, motorbikes, trains, aeroplanes account for the major share of rationale towards global warming. The various air pollutants that come from vehicles include CO, CO2, NO, sulphur dioxide and many other particulate substrates. In any vehicle class, the emissions and pollutants primarily depend on engine displacement, engine type, fuel type, vehicle age, vehicle speed, urban metric etc. In this paper, we employed regression analysis to model vehicle emissions based on engine displacement, fuel economy, fuel cost for 6000 miles, fuel cost for 12000 miles and urban metric. Also, based on the analysis using a representative dataset, we have concluded that power model fits the data better than an exponential model for the factor of engine displacement and exponential model fits the dataset better than power model with respect to the factor of fuel economy. The values of the correlation coefficient clearly indicate that the city fuel economy influences the emissions and pollutants than engine displacement. We employed transformation principles to convert non-linear power and exponential models into linear models. We have also suggested that polynomial regression techniques can be employed in future with respect to obtaining more accurate emission models.

Downloads

Published

2020-05-12

Issue

Section

Articles