# Data Driven Approaches for Sustainable Development of E-Mobility in Urban Areas

^{1}

^{2}

^{*}

_{2}Emissions from Vehicles)

## Abstract

**:**

## 1. Introduction

## 2. Literature Outcomes and Research Contribution

- Reducing traffic congestion and related concentration of charging behavior through the modal shift toward public transport by means of Park and Ride sites. It is assumed that by promoting the Park and Ride choice, EVs can also contribute to vehicle-to-grid strategies in the parking sites;
- Supporting urban planners, providing instruments able to evaluate the energy demand as a function of the travel pattern, and, consequently, as a function of the land use of the city.

## 3. Methodology

#### 3.1. Available Dataset

- Park and Ride facilities for public transport inside the Rome metropolitan area;
- GTFS data published by the Mobility Agency of the city of Rome that contains schedules, fares, and geographic transit information;
- Census data about population and activities.

- trip destination locations of each FCD vehicle are extracted and mapped;
- all the destinations are clusterized, adopting a DBScan clustering technique with parameter ε (i.e., the proximity radius between points in the same cluster) equal to 120 meters and the minimum number of points for each cluster equal to one;
- overnight stops of each FCD vehicle based on arrival time and departure time at/from destination are extracted.
- the residence is the cluster of destinations with higher values of overnight stops.

^{2}equal to 0.7465, Figure 1).

#### 3.2. Multimodality

#### 3.2.1. Formulation of the Park and Ride Models

_{O}

^{Park}/G

_{O})|

_{Dt}= f(x

_{1}, x

_{2}, …, x

_{n})

_{O}

^{Park}/G

_{O}is the share of generated trips from the traffic zone O that would make P&R in the time interval Dt. The explanatory variables x

_{1}, x

_{2}, …, x

_{n}are mainly related to: (1) the accessibility of the Park and Ride sites, (2) impedances due to the travel times on the public transport, (3) the location and occupancy of the P&R sites, and (4) characteristics related to the activity system in the starting point of the trip. The proposed explanatory variables were firstly proposed in [28] and then updated according to [14]. Specifically, the following ten explanatory variables have been adopted:

- Variables depending on the starting point of the trip (origin/starting traffic zone O) and the starting time interval (Dt) of the trip:
- Impedance on private transport x
_{1}(2), i.e., the average access time from the traffic zone O to the P&R sites belonging to the catchment area of O, weighted for the O-P&R flows; - Impedances on public transport: x
_{2}, x_{3}:x_{2}(3) is an average travel time from the P&R sites belonging to the catchment area of O to all the destinations, weighted for the O-P&R flows and for the attractiveness of each destination;In x_{3}(4) the travel times on public transport are weighted for the origin-destination (OD) flow; - Travel time differences between P&R and car-only mode x
_{4}(5); x_{4}represents the travel time advantage the user may have following the adoption of the P&R mode. The advantage is positive if the total travel time, the sum of the parking access plus the travel time by public transport, is less than the time required to go from O to D by car; - The average occupancy of the P&R sites belonging to the catchment area of the origin/starting traffic zone O, x
_{5}(6).

- Variables depending only on the characteristics of the starting point of the trip origin/starting traffic zone O) and not by the time interval, specifically:
- The variable x6 that represents the geographical extent of the origin/starting traffic zone O [km
^{2}]; - The variable x7 that represents the population density of O (between 18 and 70 years old as possible users of the P&R service);
- The variable x8, i.e., a binary variable equal to 1 if O is within the Main Ring Road of Rome (where the metro lines are located), 0 otherwise;
- The variable x9 that represents the number of urban railway stations and metro stations in the starting zone O.

- Variables depending only on the time interval, specifically:
- The dummy variable x10 that is equal to 1 if the starting time interval of the trip is inside the morning peak (between 6:00 a.m. and 11:00 a.m. for homework trips by public transport), 0 otherwise.

_{OP}= travel time by car from the origin/starting traffic zone O to the P&R site P;

_{OP}= vehicle flow from the origin/starting traffic zone O to the P&R site P;

_{O}

^{Park}= generated trips from the origin/starting traffic zone O making P&R;

_{OD}= vehicle flow from the origin/starting traffic zone O to the final destination D;

_{PD}= travel time by public transport from the P&R site P to the final destination D;

_{O}= generated trips from the origin/starting traffic zone O

_{OD}= travel time by private transport from the origin/starting traffic zone O to the final destination D;

_{D}= attractiveness of the destination zone D;

_{10}, since the disaggregated model has been calibrated considering only those trips related to the morning peak hours. Regarding the disaggregated ones, six new variables have been added to describe the behavior of each individual user, specifically:

- the traveled distance by Park and Ride x
_{11}; - the total P&R travel time x
_{12}; - the number of transfers by public transport x
_{13}; - the average waiting time by public transport x
_{14}; - the traveled distance by car x
_{15}; - the travel time by car x
_{16}.

- Only the main destinations accounting for the 70% of the demand for each starting zone are considered if the destination is unknown;
- The features related to destinations are weighted for the attractiveness AD;
- For users using a car, the potential Park and Ride site is considered the one closest in time with respect to the starting point of the trip.

#### 3.2.2. Calibration and Validation Methods

#### 3.3. Energy-Oriented Land-Use Model

_{k}), for both home-based and not home-based trips, i and time interval h, the energy demand, are computed as follows:

_{i,h}= number of trips of type i arriving in k during time h;

_{i,h}= parking time of trips of type i arriving in k during time h, where distribution of parking times can be exploited by FCD;

_{i}= Recharge power. It can be assumed equal to 3.7 kW for home-based trips and 7.4 kW for not-home-based trips (thus assuring, respectively, an average recharge time of about 8 and 4 h).

## 4. Results

#### 4.1. Multimodality: Results

^{2}) of about 0.73; in contrast, ANN shows a mean squared error of 1.085 with a correlation index (R

^{2}) of about 0.66.

- for RF, the analyses showed that the variability of the results decreases as the number of trees increases (Figure 3a);
- a clear difference exists between the RF with 10 depth levels and the others (Figure 3b)
- the optimal parameter values for the RF have been set equal to 1000 as number of trees and 20 as the maximum depth (Figure 3);
- about the different algorithms adopted to train the ANN, the Levenberg–Marquardt algorithm resulted to be the most performing one. Indeed, despite the mean squared error seems to decrease more with the Bayesian regularization (Figure 4a,b), the error with respect to the only test set demonstrates the higher accuracy of the Levenberg–Marquardt (Figure 4c);
- the optimal parameter values for the ANN have been set as 2 hidden layers and 8 neurons for each hidden layer (Figure 4c).

_{P&R}= −2.57 − 2.18 x

_{5}+ 0.0155 x

_{9}

V

_{car}= −0.01 x

_{4}+ 0.269 x

_{8}

_{5}), and the number of stations (x

_{9}); instead, the choice of the car is mainly due to the difference in travel time between car and Park and Ride (x

_{4}) and to the location of the starting zone of the trip (inside or outside the Main Ring Road). However, although this is the most reliable random utility model obtained, the resulting accuracy is very low. The model is able to reproduce only the car choices, while it is not able to reproduce any Park and Ride choices. This is probably due to the low number of trips by Park and Ride (about 16,000 with respect to the total sample of 1.6 million records). Additional behavioral models can be calibrated starting with the same data structure, for example, mixed logit able to consider for panel data (several trips related to the same vehicle as in the FCD sample) or a logit model with a penalized likelihood such as the weighted exogenous sample maximum likelihood [47]. However, problems in the model applicability or in the reproduction of the sample can, respectively, occur.

- In the case of the aggregated approach:
- the resulting variable is the share of Park and Ride demand generated by each traffic zone of the city;
- only machine learning techniques have been adopted since we are dealing with a regression model; the best form of machine learning was the random forest. However, the results of RF are also not sufficiently able to reproduce the observed outcomes: this is probably due to the low number of examples adopted as an input.

- In the case of the disaggregated approach:
- the resulting variable is the choice of Park and Ride or car to perform the trip; thus, it is a binary variable, and it is related to each single trip;
- both random utility models and machine learning (specifically neural networks) have been tested; however, random utility models were not satisfactory due to the oversampling of cars with respect to the Park and Ride choices. In contrast, specific architectures of neural networks combined with the right optimization algorithm can reproduce perfectly the sample despite the oversampling and thanks to the high number of examples adopted as an input.

#### 4.2. Energy-Oriented Land-Use Model: Results

- strictly residential zones correspond to intermediate/high values of population density (about 3500–9000 inhabitants per km
^{2}) and residential buildings are more than 80% of the total buildings; - strictly business/industrial zones have a low population density (mostly below 450 inhabitants per km
^{2}), and residential buildings are less than 70% of the total buildings; - strictly commercial zones showed average population density while quite high values of residential buildings: this is probably due to the typical configuration of shopping activities in Rome, where most of the shops are located on the ground floor of residential buildings facing the street.

## 5. Discussion and Policy Implications

## 6. Conclusions and Further Developments

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## References

- European Commission, Directorate-General Mobility and Transport. Transport in the European Union—Current trends and issues. Available online: https://ec.europa.eu/transport/sites/transport/files/2019-transport-in-the-eu-current-trends-and-issues.pdf (accessed on 17 January 2020).
- Liberto, C.; Valenti, G.; Lelli, M.; Ferrara, M.; Nigro, M. Evaluation of the impact of e-mobility scenarios in large urban areas. In Proceedings of the 2017 5th IEEE International Conference on Models and Technologies for Intelligent Transportation Systems (MT-ITS), Naples, Italy, 26–28 June 2017; pp. 379–384. [Google Scholar]
- IEA–International Energy Agency, Global EV Outlook. Available online: https://www.iea.org/reports/global-ev-outlook-2020 (accessed on 20 April 2021).
- Deloitte “Electric Vehicles—Setting a Course for 2030”. 2020. Available online: https://www2.deloitte.com/content/dam/Deloitte/rs/Documents/about-deloitte/DI_Electric-Vehicles.pdf (accessed on 12 February 2021).
- Canizes, B.; Soares, J.; Costa, A.; Pinto, T.; Lezama, F.; Novais, P.; Vale, Z. Electric Vehicles’ User Charging Behaviour Simulator for a Smart City. Energies
**2019**, 12, 1470. [Google Scholar] [CrossRef] [Green Version] - Krajzewicz, D.; Erdmann, J.; Behrisch, M.; Bieker-Walz, L. Recent Development and Applications of SUMO—Simulation of Urban MObility. Int. J. Adv. Syst. Meas.
**2012**, 5, 128–138. [Google Scholar] - Axhausen, K.W. The Multi-Agent Transport Simulation MATSim; Ubiquity Press: London, UK, 2016. [Google Scholar]
- Kurczveil, T.; López, P.Á.; Schnieder, E. Implementation of an Energy Model and a Charging Infrastructure in SUMO. In Proceedings of the Transactions on Petri Nets and Other Models of Concurrency XV, Berlin, Germany, 7 November 2014; pp. 33–43. [Google Scholar]
- Luin, S.P.; Al-Mansour, F. Microsimulation of Electric Vehicle Energy Consumption; Energy Scenario Toward 2030; Elsevier: New York, NY, USA, 2019; Volume 174, pp. 24–32. [Google Scholar]
- Fiori, C.; Marzano, V.; Punzo, V.; Montanino, M. Energy Consumption Modeling in Presence of Uncertainty. IEEE Trans. Intell. Transp. Syst.
**2020**, 1–12. [Google Scholar] [CrossRef] - Farid, A.M. A Hybrid Dynamic System Model for Multimodal Transportation Electrification. IEEE Trans. Control. Syst. Technol.
**2016**, 25, 940–951. [Google Scholar] [CrossRef] - Kassakian, J.G. The Future of the Electric Grid: An Interdisciplinary MIT Study; MIT Press: Cambridge, MA, USA, 2011; Available online: http://web.mit.edu/mitei/research/studies/documents/electric-grid-2011/Electric_Grid_Full_Report.pdf (accessed on 10 May 2021).
- Al Junaibi, R.; Farid, A.M. A method for the technical feasibility assessment of electrical vehicle penetration. In Proceedings of the 2013 IEEE International Systems Conference (SysCon), Orlando, FL, USA, 15–18 April 2013; pp. 606–611. [Google Scholar]
- Ferrara, M.; Monechi, B.; Valenti, G.; Liberto, C.; Nigro, M.; Biazzo, I. A simulation tool for energy management of e-mobility in urban areas. In Proceedings of the 2019 6th International Conference on Models and Technologies for Intelligent Transportation Systems (MT-ITS), Cracow, Poland, 5–7 June 2019; pp. 1–7. [Google Scholar]
- RAND Europe. Modelling Park-and-Ride. Understanding Multi-Modal Travel. 2007. Available online: www.rand.org/randeurope (accessed on 26 February 2021).
- Zhao, X.; Li, Y.; Xia, H. Behavior decision model for park-and-ride facilities utilization. Adv. Mech. Eng.
**2017**, 9, 1–9. [Google Scholar] [CrossRef] [Green Version] - Sharma, B.; Hickman, M.; Nassir, N. Park-and-ride lot choice model using random utility maximization and random regret minimization. Transportation
**2017**, 46, 217–232. [Google Scholar] [CrossRef] - Ma, T.-Y. Bayesian networks for multimodal mode choice behaviour modelling: A case study for the cross border workers of Luxembourg. In Proceedings of the 18th Euro Working Group on Transportation, EWGT 2015, Delft, The Netherlands, 14–16 July 2015. [Google Scholar]
- Thanh, T.T.M.; Ly, H.-B.; Pham, B.T. A Possibility of AI Application on Mode-choice Prediction of Transport Users in Hanoi. In Lecture Notes in Civil Engineering; Springer Science and Business Media LLC: Cham, Switzerland, 2019; Volume 54, pp. 1179–1184. [Google Scholar]
- Guan, Y.; Wang, Y.; Yan, X.; Guo, H.; Zhou, Y. A Big-Data-Driven Framework for Parking Demand Estimation in Urban Central Districts. J. Adv. Transp.
**2020**, 2020, 1–13. [Google Scholar] [CrossRef] - Xiao, J.; Lou, Y.; Frisby, J. How likely am I to find parking?—A practical model-based framework for predicting parking availability. Transp. Res. Part B Methodol.
**2018**, 112, 19–39. [Google Scholar] [CrossRef] - Schuster, T.; Volz, R. Predicting Parking Demand with Open Data. In Proceedings of the Transactions on Petri Nets and Other Models of Concurrency XV.; Springer Science and Business Media LLC: Cham, Switzerland, 2019; pp. 218–229. [Google Scholar]
- Karolemeas, C.; Tsigdinos, S.; Tzouras, P.; Nikitas, A.; Bakogiannis, E. Determining Electric Vehicle Charging Station Location Suitability: A Qualitative Study of Greek Stakeholders Employing Thematic Analysis and Analytical Hierarchy Process. Sustainability
**2021**, 13, 2298. [Google Scholar] [CrossRef] - Wu, Z.; Zhuang, Y.; Zhou, S.; Xu, S.; Yu, P.; Du, J.; Luo, X.; Abbas, G. Bi-Level Planning of Multi-Functional Vehicle Charging Stations Considering Land Use Types. Energies
**2020**, 13, 1283. [Google Scholar] [CrossRef] [Green Version] - Orsi, F. On the sustainability of electric vehicles: What about their impacts on land use? Sustain. Cities Soc.
**2021**, 66, 102680. [Google Scholar] [CrossRef] - Zhang, R.; Zhang, J.; Long, Y.; Wu, W.; Liu, J.; Jiang, Y. Long-term implications of electric vehicle pene-tration in urban decarbonization scenarios: An integrated land use–transport–energy model. Sustain. Cities Soc.
**2021**, 68, 102800. [Google Scholar] [CrossRef] - Tizghadam, A.; Khazaei, H.; Moghaddam, M.H.Y.; Hassan, Y. Machine Learning in Transportation. J. Adv. Transp.
**2019**, 2019, 1–3. [Google Scholar] [CrossRef] [Green Version] - Ferrara, M.; Liberto, C.; Nigro, M.; Trojani, M.; Valenti, G. Multimodal choice model for e-mobility scenarios. Transp. Res. Procedia
**2019**, 37, 409–416. [Google Scholar] [CrossRef] - Hastie, T.; Tibshirani, R.; Friedman, J. The Elements of Statistical Learning: Data Mining, Inference and Prediction. Math. Intell.
**2005**, 27, 83–85. [Google Scholar] - Breiman, L. Random Forests. Mach. Learn.
**2001**, 45, 5–32. [Google Scholar] [CrossRef] [Green Version] - Zhang, Y.; Haghani, A. A gradient boosting method to improve travel time prediction. Transp. Res. Part C Emerg. Technol.
**2015**, 58, 308–324. [Google Scholar] [CrossRef] - Kazic, B.; Mladenić, D.; Košmerlj, A. Traffic Flow Prediction from Loop Counter Sensor Data using Machine Learning Methods. In Proceedings of the International Conference on Vehicle Technology and Intelligent Transport Systems, VEHITS, Lisbon, Portugal, 20–22 May 2015; SCITEPRESS—Science and Technology Publications: Setúbal, Portugal, 2015; pp. 119–127. [Google Scholar]
- Zarei, N.; Ghayour, M.A.; Hashemi, S. Road Traffic Prediction Using Context-Aware Random Forest Based on Volatility Nature of Traffic Flows. In Transactions on Petri Nets and Other Models of Concurrency XV; Springer Science and Business Media LLC: Cham, Switzerland, 2013; Volume PART 1, pp. 196–205. [Google Scholar]
- Ghasri, M.; Rashidi, T.H.; Waller, T. Developing a disaggregate travel demand system of models using data mining techniques. Transp. Res. Part A Policy Pr.
**2017**, 105, 138–153. [Google Scholar] [CrossRef] - Ashqar, H.I.; Elhenawy, M.; Almannaa, M.; Ghanem, A.; Rakha, H.A.; House, L. Modeling bike availability in a bike-sharing system using machine learning. In Proceedings of the 2017 5th IEEE International Conference on Models and Technologies for Intelligent Transportation Systems (MT-ITS), Naples, Italy, 26–28 June 2017; pp. 374–378. [Google Scholar] [CrossRef]
- Liu, X.; Liu, K.; Li, M.; Lu, F. A ST-CRF Map-Matching Method for Low-Frequency Floating Car Data. IEEE Trans. Intell. Transp. Syst.
**2016**, 18, 1241–1254. [Google Scholar] [CrossRef] - Hensher, D.A.; Ton, T.T. A comparison of the predictive potential of artificial neural networks and nested logit models for commuter mode choice. In Transportation Research Part E: Logistics and Transportation Review; Elsevier: New York, NY, USA, 1999; Volume 36, pp. 155–172. [Google Scholar]
- Cantarella, G.E.; de Luca, S. Multilayer feedforward networks for transportation mode choice analysis: An analysis and a comparison with random utility models. Transp. Res. Part C
**2005**, 13, 121–155. [Google Scholar] [CrossRef] - Nijkamp, P.; Reggiani, A.; Tsang, W.F. Comparative modelling of interregional transport flows: Applications to multimodal European freight transport. Eur. J. Oper. Res.
**2004**, 155, 584–602. [Google Scholar] [CrossRef] [Green Version] - Gazder, U.; Ratrout, N.T. A new logit-artificial neural network ensemble for mode choice modeling: A case study for border transport. J. Adv. Transp. J. Adv. Transp.
**2015**, 49, 855–866. [Google Scholar] [CrossRef] - Dabiri, S.; Heaslip, K. Inferring transportation modes from GPS trajectories using convolutional neural network. Transp. Res. Part C Emerg. Technol.
**2018**, 86, 360–371. [Google Scholar] [CrossRef] [Green Version] - Bierlaire, M. A short introduction to PandasBiogeme. Tech. Rep. Transp. Mobil. Lab. ENAC EPFL 2020. Available online: https://transp-or.epfl.ch/documents/technicalReports/Bier20.pdf (accessed on 13 January 2021).
- Caceres, N.; Benítez, F.G. Supervised Land Use Inference from Mobility Patterns. J. Adv. Transp.
**2018**, 2018, 1–12. [Google Scholar] [CrossRef] - MacQueen, J. Some methods for classification and analysis of multivariate observations. In Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, Berkeley, CA, USA, 21 June 1965; Volume 1, pp. 281–297. [Google Scholar]
- Nigro, M.; Cipriani, E.; Del Giudice, A. Exploiting floating car data for time-dependent Origin–Destination matrices estimation. J. Intell. Transp. Syst.
**2017**, 22, 159–174. [Google Scholar] [CrossRef] - Kayri, M. Predictive Abilities of Bayesian Regularization and Levenberg–Marquardt Algorithms in Artificial Neural Networks: A Comparative Empirical Study on Social Data. Math. Comput. Appl.
**2016**, 21, 20. [Google Scholar] [CrossRef] - Train, K. Discrete Choice Methods with Simulation; Cambridge University Press; Available online: https://eml.berkeley.edu/books/train1201.pdf (accessed on 13 January 2021).
- European Commission (EC). The European Green Deal: Striving to Be the First Climate Neutral Continent; European Commission: Brussels, Belgium, 2019; Available online: https://eur-lex.europa.eu/legal-content/EN/TXT/?qid=1596443911913&uri=CELEX:52019DC0640#document2 (accessed on 26 August 2020).

**Figure 1.**Map of the residences after the application of the residence location procedure and comparison with census data.

**Figure 4.**Mean square error (MSE) as a function of number of neurons for one hidden layer (

**a**) and two hidden layers (

**b**). R2 on only the test set comparing Levenberg–Marquardt and Bayesian regularization with different numbers of neurons for two hidden layers (

**c**).

**Figure 5.**Comparison between features’ importance of RF and ANN (with one or two hidden layers) in the aggregated approach.

**Figure 6.**Results of different calibrated ANNs (disaggregated approach) for different activation function (Relu and Sigmoid, relu), numbers of hidden layers, and number of neurons in the hidden layers.

**Figure 7.**K-means clustering applied to (

**1**) attracted trips and (

**2**) generated trips on the weekdays. Resulting clusters: mixed-use (

**a1**,

**b2**), business/industrial (

**b1**,

**a2**), residential (

**c1**,

**c2**), and commercial (

**d1**,

**d2**).

**Figure 8.**K-means clustering applied to (

**1**) attracted trips and (

**2**) generated trips on the weekends. Resulting clusters: residential (

**a1**,

**c2**), mixed-use (

**c1**,

**a2**), and nightlife and leisure (

**b1**,

**b2**).

**Figure 9.**Parking time distribution during the daytime as a function of the land use of the arrival zone for home-based and not-home-based trips.

**Figure 10.**Examples of energy demand computation for two different land use zones in the city of Rome.

ML Technique | Best Architecture | R2 Train | R2 Test | MSE (All) |
---|---|---|---|---|

Random Forest | 1000 number of trees | 0.891 | 0.734 | 0.560 |

20 max depth | ||||

Neural Network | 2 hidden layers 8 neurons | 0.704 | 0.657 | 1.085 |

Parameters | Value | Std Error | t-Test | p-Value |
---|---|---|---|---|

ASC_P&R | −2.570 | 0.0225 | −114.43 | 0.00 |

β4 | −0.010 | 0.00055 | −18.26 | 0.00 |

β5 | −2.180 | 0.0261 | −83.41 | 0.00 |

β8 | 0.269 | 0.0163 | 16.47 | 0.00 |

β9 | 0.0155 | 0.00486 | 3.18 | 0.00 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Nigro, M.; Ferrara, M.; De Vincentis, R.; Liberto, C.; Valenti, G.
Data Driven Approaches for Sustainable Development of E-Mobility in Urban Areas. *Energies* **2021**, *14*, 3949.
https://doi.org/10.3390/en14133949

**AMA Style**

Nigro M, Ferrara M, De Vincentis R, Liberto C, Valenti G.
Data Driven Approaches for Sustainable Development of E-Mobility in Urban Areas. *Energies*. 2021; 14(13):3949.
https://doi.org/10.3390/en14133949

**Chicago/Turabian Style**

Nigro, Marialisa, Marina Ferrara, Rosita De Vincentis, Carlo Liberto, and Gaetano Valenti.
2021. "Data Driven Approaches for Sustainable Development of E-Mobility in Urban Areas" *Energies* 14, no. 13: 3949.
https://doi.org/10.3390/en14133949