Predictive analysis of passenger vehicle emissions and fuel consumption in Addis Ababa, Ethiopia using COPERT based on vehicle growth forecasting

Abstract

This study investigates Addis Ababa's passenger car growth after 2005 and its effects on the environment between 2018 and 2024. Predictive Linear Regression and Artificial Neural Networks (ANN) machine learning techniques were used in the study to forecast the growth of the fleet of passenger cars to determine related emissions and energy usage. The study incorporates environmental factors, vehicle activity patterns, and local vehicle registration data into the COPERT model. Vehicle classifications by fuel type (diesel and petrol) and Euro 1 to Euro 6 criteria were taken into account in the analysis. The overall number of passenger vehicles has increased by more than twentyfold during the last 20 years, according to the results. Pollutant emissions have increased as a result, especially from older Euro 1 to Euro 3 vehicles, and include CO₂, NO_x, CO, and VOC. Due to limited adoption of low-emission vehicle technology and growing travel demand, the energy consumption for passenger cars also exhibited steady growth. The results highlighted how policy changes stopped the increase of passenger vehicle emissions after 2024 due to the banning of internal combustion-based passenger vehicles. By offering a data-driven methodology for assessing vehicle-related emissions and guiding mitigation methods, this study supports sustainable transportation planning at the local and national levels.