Machine-Learning-Based Digital Twins for Transient Vehicle Cycles and Their Potential for Predicting Fuel Consumption

Tomanik, Eduardo; Jiménez-Reyes, Antonio J.; Tomanik, Victor; Tormos, Bernardo

doi:10.3390/vehicles5020032

Cited by 6 publications

(5 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Using the created algorithm, the author obtained a vehicle fuel consumption error of 1.4% compared to the actual value obtained in real tests. Tomanik et al [22] used artificial neural networks to assess vehicle fuel consumption using data drawn from the OBD system (vehicle speed, rpm, torque, fuel consumption rate, acceleration). For several of the conducted tests, the results obtained using the model and in the experiment were 98.8% similar.…”

Section: Parameter Euro 6dmentioning

confidence: 99%

Exhaust Emissions from Euro 6 Vehicles in WLTC and RDE—Part 1: Methodology and Similarity Conditions Studies

Pielecha,

Kurtyka

2023

Energies

View full text Add to dashboard Cite

The article is an attempt to perform an ecological assessment of passenger cars with various types of engines in road emission tests. The main research problem and, at the same time, the goal was to develop a method for determining the exhaust emissions from motor vehicles in real traffic conditions based on results obtained in homologation tests. The tests were carried out on vehicles equipped with gasoline, diesel, and hybrid engines, and the obtained results were analyzed. All of the selected vehicles were of the same class—passenger cars, with a similar curb weight, similar maximum engine power, and in the same emission class (Euro 6d). The authors compared the dynamic parameters of vehicle motion in established emission tests: Worldwide harmonized Light vehicles Test Cycles and Real Driving Emissions. Four procedures were used to analyze and compare the operating conditions of the vehicles in the WLTC and RDE tests, differing in how the phases in the tests were divided as well as having a different methodology for determining the road emissions in the tests. The procedures were as follows: WLTC (where the test was divided and the determination of the road emission of exhaust gases was carried out according to the standard WLTP procedure), RDE (the road test was divided into sections and the exhaust emission was determined according to the standard RDE procedure), WLTC1+2 (the test was divided into phases: 1 + 2, 3, and 4; a combination of phases 1 and 2 corresponding to the urban section of the RDE test), WLTCRDE (where drive phases were divided and emissions determined in the same way as in the RDE procedure, which assumes the division of the test into sections based on vehicle speed). The implementation of the research task in the form of an algorithm procedure when comparing the dynamic parameters of the movement in the WLTC and RDE tests is the leading goal presented in this article. The division of the WLTC test into sections (urban, rural, and motorway) according to the RDE procedure and also the calculation of the total emissions in the test according to this procedure resulted in obtaining similar road emission values in the test.

show abstract

Section: Parameter Euro 6dmentioning

confidence: 99%

Exhaust Emissions from Euro 6 Vehicles in WLTC and RDE—Part 1: Methodology and Similarity Conditions Studies

Pielecha,

Kurtyka

2023

Energies

View full text Add to dashboard Cite

show abstract

“…Figures 22 and 23 show the trained digital win versus the measurement. The case is discussed on more details in [11] including use of shorter durations and fewer input parameters.…”

Section: Use Of Ecu Obd Datamentioning

confidence: 99%

“…Several authors [3][4][5][6][7][8][9] have explored the use of AI to analyze and predict fuel consumption and other parameters. On previous author works [10][11] data from the Downloadable Dynamometer Database, Argonne National Laboratory [12,13] were used to develop tribological and supervised machine learning models to predict instantaneous fuel consumption on emission transient cycles. The developed Random Forest and Neural Network models used dynamometer and/or vehicle Engine Control Unit (ECU) or OBD data from a given test and then were used to predict the output for different tests.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Uso de digital twins treinados em machine learning aplicados em teste veiculares de emissões.

TOMANIK,

JIMENEZ-REYES

et al. 2023

Blucher Engineering Proceedings

View full text Add to dashboard Cite

Emission tests generate a huge amount of measurement data but are usually evaluated only regarding the accumulated value according to the emission homologation limits. In this work, a Random Forest machine leaning code was used to create a vehicle digital twin able to predict an output of interest, e.g., instantaneous fuel consumption. The digital twin, after being trained on measurements from the FTP-75 cold start, was able to predict with good accuracy, not only the accumulated values but also the instantaneous values of fuel consumption and CO2 emissions. A first exercise on a hybrid vehicle showed some potential, but more work is needed to reproduce some of the parameters of interest.

show abstract

“…Figure 7.The artificial neural network model scheme for fuel consumption prediction (adopted from Tomanik et al[52]). …”

mentioning

confidence: 99%

Recent Progress of Machine Learning Algorithms for the Oil and Lubricant Industry

2023

View full text Add to dashboard Cite

Machine learning (ML) algorithms have brought about a revolution in many industries where otherwise operation time, cost, and safety would have been compromised. Likewise, in lubrication research, ML has been utilized on many occasions. This review provides an in-depth understanding of seven ML algorithms from a tribological perspective. More specifically, it presents a comprehensive overview of recent advancements in ML applied to lubrication research, organized into four distinct categories. The first category, experimental parameter prediction, highlights the significant contributions of artificial neural networks (ANNs) in accurately forecasting operating conditions related to friction and wear. These predictions offer valuable insights that aid in forensic preparation. Discriminant analysis, Bayesian modeling, and transfer learning approaches have also been used to predict experimental parameters. Second, to predict the lubrication film thickness and identify the lubrication regime, algorithms such as logistic regression and ANN were useful. Such predictions provide up to 99.25% accuracy. Third, to predict the friction and wear for a given experimental condition, support vector machine (SVM), polynomial regression, and ANN offered an accuracy above 93%. Finally, for condition monitoring for bearings, gearboxes, gear trains, and similar critical situations where regular in-person inspection is difficult, Naïve Bayes, SVM, decision trees, and ANN were utilized to predict the safe life of lubricants. This review highlighted these four aspects with state-of-the-art examples and discussed the current situation and projected future possibilities of lubricant design facilitated by ML techniques.

show abstract

Machine-Learning-Based Digital Twins for Transient Vehicle Cycles and Their Potential for Predicting Fuel Consumption

Cited by 6 publications

References 38 publications

Exhaust Emissions from Euro 6 Vehicles in WLTC and RDE—Part 1: Methodology and Similarity Conditions Studies

Exhaust Emissions from Euro 6 Vehicles in WLTC and RDE—Part 1: Methodology and Similarity Conditions Studies

Uso de digital twins treinados em machine learning aplicados em teste veiculares de emissões.

Recent Progress of Machine Learning Algorithms for the Oil and Lubricant Industry

Contact Info

Product

Resources

About