Prediction of real-world CO2 emissions of passenger cars — Project proposal

Motivation

As the climate change becoming a global problem over the years, reducing co2 emission became a primal goal. Passenger cars are a significant contributor to co2 emissions, making it essential to understand and predict real-world emissions accurately. This project aims to predict real-life co2 emissions of passenger cars using given laboratory test information and real-world user-submitted or automatically recorded data.

Data

We’ll be using a combined data where we merge European Environment Agency’s Datahub, and user data collected by us from the web.

Related work

Fuel consumption and CO2 emissions from passenger cars in Europe — Laboratory versus real-world emissions [1]

This paper is the inspiration behind our project, where authors compare real-world emissions versus laboratory results. They do not make any predictions though.

From laboratory to road: A 2014 update of official and “real-world” fuel consumption and CO2 values for passenger cars in Europe [2]

This paper was given in references of the first paper. We will be mostly using this paper’s data section for data collection where we need real-world data of the cars.

Segment-Based CO₂ Emission Evaluations From Passenger Cars Based on Deep Learning Techniques [3]

Authors of this paper uses deep learning techniques to evaluate inter-class and intra-class differences in co2 emissions.

A deep neural network-based model for the prediction of hybrid electric vehicles carbon dioxide emissions [4]

In this paper, authors combine already used DP based energy management systems and deep neural networks (DNN). Their DNN’s are called DNNs-PM which refers to deep neural network-based prediction model. Their claim is this model can be used to optimize design parameters with respect to driving missions.