A Study of Prediction Based on Regression Analysis for Real-World Co2 Emissions with Light-Duty Diesel Vehicles

• Published in: International journal of automotive technology Vol. 22; no. 3; pp. 569 - 577
• Main Authors: Cha, Junepyo, Park, Junhong, Lee, Hyoungwook, Chon, Mun Soo
• Format: Journal Article
• Language: English
• Published: Seoul The Korean Society of Automotive Engineers 01.06.2021; Springer Nature B.V; 한국자동차공학회
• Subjects: Acceleration; Automotive Engineering; Carbon dioxide; Diesel fuels; Engineering; Exhaust gases; Light duty vehicles; Light emission; Regression analysis; Regression models; Reliability; Test vehicles; Traffic speed; Trip estimation; Vehicles; 자동차공학; Regression Analysis; Light-Duty Vehicle (LDV); Real-World CO
• ISSN: 1229-9138; 1976-3832
• DOI: 10.1007/s12239-021-0053-z

The objective in present study is to develop a regression analysis model to estimate real-world CO 2 emissions of light-duty diesel vehicles considering domestic road conditions. For regression analysis variables, OBD data such as vehicle speed, acceleration, engine speed (rpm), and engine power were used. Regression analysis results were compared with CO 2 emissions measured using PEMS on the test routes of the real driving emissions-light duty vehicles (RDE-LDV). In results, the vehicle speed and air/fuel data from the OBD signals maintained a linear relationship with the GPS and exhaust gas flowmeter-based vehicle speed and exhaust flow data. All determination coefficients were ≥0.99, indicating that the OBD data provided by the test vehicle in this study exhibited strong reliability. To investigate the accuracy of the regression equation estimated using the trip variables of the OBD data, the driving variables were substituted into the equation to obtain CO 2 estimations and the real CO 2 emissions measured using PEMS were compared. A strong linear relationship was observed between the regression equation-based CO 2 estimations and real CO 2 measurements. The determination coefficient was approximately 0.93, supporting the reliability of the estimation results.