Applying machine learning to construct braking emission model for real-world road driving

Wei, Ning; Men, Zhengyu; Ren, Chunzhe; Jia, Zhenyu; Zhang, Yanjie; Jin, Jiaxin; Chang, Junyu; Lv, Zongyan; Guo, Dongping; Yang, Zhiwen; Guo, Jiliang; Wu, Lin; Peng, Jianfei; Wang, Ting; Du, Zhuofei; Zhang, Qi‐Jun; Mao, Hongjun

doi:10.1016/j.envint.2022.107386

Cited by 9 publications

(6 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To figure out this difference, a specific survey was conducted to investigate the special properties of NEBCs by comparing them with the commonly used BC prepared with agricultural residues (ABC500 and ABC900). First, the Eli5 algorithm was implemented for inspecting black-box models . Hence, it was used to inspect an individual prediction of a model, calculating the importance index of the model input feature.…”

Section: Resultsmentioning

confidence: 99%

“…First, the Eli5 algorithm was implemented for inspecting black-box models. 43 algorithm, we can attempt to ascertain the reasoning behind the model's choice and which feature would work for the prediction. Figure 4d shows that the SSA represented the most contribution, followed by I D /I G , T°C, C%, Gra, O% and Oxi, which have played a positive role in the prediction results of FBC900.…”

Section: Data Collection and Preprocessingmentioning

confidence: 99%

See 1 more Smart Citation

Enhancing Biochar-Based Nonradical Persulfate Activation Using Data-Driven Techniques

Wang

Zhang

Chen

et al. 2023

Environ. Sci. Technol.

View full text Add to dashboard Cite

Converting biomass into biochar (BC) as a functional biocatalyst to accelerate persulfate activation for water remediation has attracted much attention. However, due to the complex structure of BC and the difficulty in identifying the intrinsic active sites, it is essential to understand the link between various properties of BC and the corresponding mechanisms promoting nonradicals. Machine learning (ML) recently demonstrated significant potential for material design and property enhancement to help tackle this problem. Herein, ML techniques were applied to guide the rational design of BC for the targeted acceleration of nonradical pathways. The results showed a high specific surface area, and O% values can significantly enhance nonradical contribution. Furthermore, the two features can be regulated by simultaneously tuning the temperatures and biomass precursors for efficient directed nonradical degradation. Finally, two nonradical-enhanced BCs with different active sites were prepared based on the ML results. This work serves as a proof of concept for applying ML in the synthesis of tailored BC for persulfate activation, thereby revealing the remarkable capability of ML for accelerating bio-based catalyst development.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Data Collection and Preprocessingmentioning

confidence: 99%

Enhancing Biochar-Based Nonradical Persulfate Activation Using Data-Driven Techniques

Wang

Zhang

Chen

et al. 2023

Environ. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…Brake dust is typically produced as the result of mechanical action during breaking events, and a range of factors have been associated with instantaneous emission rates, including composition brake pads, design of braking mechanism, vehicle mass, brake temperature and driving conditions (see, e.g., [12,32,33]). NAEI methods provide initial estimates of EF brake of 53.6, 27.1 and 8.4 mg.km −1 for PM 10 on urban, rural and motorway routes, respectively, and 21.4, 10.8 and 3.4 mg.km −1 for PM 2.5 on urban, rural and motorway routes, respectively [34].…”

Section: Brake Particulate Emission Factors Ef Brakementioning

confidence: 99%

“…Equation ( 6) assumes a strong association between EFs and speed at an aggregated level, i.e., speeds averaged across several minutes or kilometers. Elsewhere, researchers have identified other statistical measures of driving as better proxies for EF brake , e.g., the US EPA used acceleration ≤−2 miles.h −1 .s −1 or vehicle specific power (VSP) ≤−4 kW.tonne −1 in their motor vehicle emission simulator (MOVES) model [44], and Wei et al [32] identified brake energy intensity (BEI) in their more recent machine learning study. Alternative EF tyre parameters are less commonly cited although elevated emissions are associated with a range of driving activities, including both acceleration and braking [45].…”

mentioning

confidence: 99%

Meta-Analysis as Early Evidence on the Particulate Emissions Impact of EURO VI on Battery Electric Bus Fleet Transitions

Tivey¹,

Davies

Levine

et al. 2023

Sustainability

View full text Add to dashboard Cite

The current generation of Zero Emission Vehicle (ZEV) policies are designed to accelerate the transition away from conventional internal combustion engine (ICE) petrol and diesel vehicle fleets. However, the current focus on zero exhaust emissions and the lack of more detailed guidance regarding Non-Exhaust Emissions (NEEs) may mean that some of the trade-offs in transitioning to, e.g., Battery Electric Vehicle (BEV) fleets may be missed by many in the commercial sector. Here, as part of early work on the scoping of the First Bus EURO VI Diesel Vehicle (E6DV) to BEV fleet upgrades, we estimate E6DV total particulate emissions to be ca. 62–85 and 164–213 mg.veh−1.km−1 for PM2.5 and PM10, respectively, and that the majority, typically 93–97%, are NEEs. We also discuss the complex interaction between E6DV/BEV properties and estimate potential changes resulting from the transition to BEVs as ranging from a decrease of ca. 2–12% to an increase of ca. 12–50% depending on a combination of weight difference, regenerative brake performance and journey type. Finally, we propose metrics that would allow fleet operators more insight into a wider range of emission outcomes at the scoping stage of a fleet upgrade.

show abstract

“…Compared to complex deep learning models, using four ensemble models made it easier to capture the variation in parameters and variable interpretation within each model. For the ensemble learning component, the RF was a typical bagging algorithm that accomplished a classification task by voting and a regression task by averaging [43]. Specifically, the RF was a set of decision trees, and each tree was constructed using the best split for each node among a subset of predictors randomly chosen at that node.…”

Section: Model Building and Performance Evaluationmentioning

confidence: 99%

An Explainable Prediction Model for Aerodynamic Noise of an Engine Turbocharger Compressor Using an Ensemble Learning and Shapley Additive Explanations Approach

Huang,

Ni,

Qiao

et al. 2023

Sustainability

View full text Add to dashboard Cite

In the fields of environment and transportation, the aerodynamic noise emissions emitted from heavy-duty diesel engine turbocharger compressors are of great harm to the environment and human health, which needs to be addressed urgently. However, for the study of compressor aerodynamic noise, particularly at the full operating range, experimental or numerical simulation methods are costly or long-period, which do not match engineering requirements. To fill this gap, a method based on ensemble learning is proposed to predict aerodynamic noise. In this study, 10,773 datasets were collected to establish and normalize an aerodynamic noise dataset. Four ensemble learning algorithms (random forest, extreme gradient boosting, categorical boosting (CatBoost) and light gradient boosting machine) were applied to establish the mapping functions between the total sound pressure level (SPL) of the aerodynamic noise and the speed, mass flow rate, pressure ratio and frequency of the compressor. The results showed that, among the four models, the CatBoost model had the best prediction performance with a correlation coefficient and root mean square error of 0.984798 and 0.000628, respectively. In addition, the error between the predicted total SPL and the observed value was the smallest, at only 0.37%. Therefore, the method based on the CatBoost algorithm to predict aerodynamic noise is proposed. For different operating points of the compressor, the CatBoost model had high prediction accuracy. The noise contour cloud in the predicted MAP from the CatBoost model was better at characterizing the variation in the total SPL. The maximum and minimum total SPLs were 122.53 dB and 115.42 dB, respectively. To further interpret the model, an analysis conducted by applying the Shapley Additive Explanation algorithm showed that frequency significantly affected the SPL, while the speed, mass flow rate and pressure ratio had little effect on the SPL. Therefore, the proposed method based on the CatBoost algorithm could well predict aerodynamic noise emissions from a turbocharger compressor.

show abstract

Applying machine learning to construct braking emission model for real-world road driving

Cited by 9 publications

References 28 publications

Enhancing Biochar-Based Nonradical Persulfate Activation Using Data-Driven Techniques

Enhancing Biochar-Based Nonradical Persulfate Activation Using Data-Driven Techniques

Meta-Analysis as Early Evidence on the Particulate Emissions Impact of EURO VI on Battery Electric Bus Fleet Transitions

An Explainable Prediction Model for Aerodynamic Noise of an Engine Turbocharger Compressor Using an Ensemble Learning and Shapley Additive Explanations Approach

Contact Info

Product

Resources

About