All Issue

2024 Vol.42, Issue 3 Preview Page

Article

Study on Black Ice Forecasting Techniques Using Atmospheric Data for Efficient Nighttime Highway Patrol (Focusing on Boosting Algorithms)

일반국도 야간 도로 순찰 효율화를 위한 기상 자료 기반 블랙아이스 예측 기법 연구 (부스팅 머신러닝 알고리즘을 중심으로)

Jinhwan JANG1*
장 진환1*
1Research Fellow, Department of Highway and Transportation, Korea Institute of Civil Engineering and Building Technology, Gyeonggi 10223, Korea
1한국건설기술연구원 도로교통연구본부 연구위원
*Corresponding Author
30 June 2024. pp. 297-312
PDF XML Citation
Abstract

Faced with persistent accidents caused by black ice, nighttime National Highway (NH) patrols during winter become mandatory. However, considering the substantial length of the highway (approximately 14,000km) and limited resources (only 17 patrol cars), an efficient strategy for patrolling is urgently needed. Patrolling regions with a high potential for black ice formation can be a reasonable solution. The problem lies in determining which regions have high potential. To address this gap, this study investigates three boosting algorithms—AdaBoost, XGBoost, and CatBoost—to identify regions likely to develop black ice. The models use atmospheric data as inputs, including relative humidity, air temperature, dew point, precipitation probability, and wind speed. Analysis on the atmospheric data reveals that nighttime icing typically happens when the air temperature is below 4°C and the relative humidity exceeds 75%. Moreover, black ice is more likely to form during temperature increases in the absence of precipitation. To assess the three models, baseline data were generated based on a physical principle that ice forms when pavement temperature is below freezing and lower than the dew point temperature. Consequently, all models demonstrated similar performance, achieving an accuracy rate of approximately 90%. When the findings of this study applied to nighttime NH patrolling in winter, winter road maintenance activities can be much more effective and efficient as well.

Keywords
black ice
boosting algorithm
highway patrol
national highway
pavement temperature

최근 해마다 반복되는 블랙아이스 교통사고 예방을 위해 겨울철 야간 도로 순찰이 의무화 되었다. 일반국도의 경우 연장(약 14,000km)이 길고 순찰 자원(차량 17대)이 제한되어 있기 때문에 전체 구간을 매일 순찰하는 것은 사실상 불가능하다. 따라서 블랙아이스가 발생할 것으로 예상되는 위험한 구간을 파악해 해당 구간을 집중적으로 순찰하는 전략이 필요하다. 하지만 블랙아이스 발생이 우려되는 구간 정보가 없어 이에 대한 해결책 마련이 시급하다. 이에 본 연구는 부스팅 머신러닝 알고리즘을 이용해 블랙아이스 발생 가능성이 높은 구간을 추정하는 모형을 개발했다. 분석에 사용된 데이터는 기상청에서 예보하는 상대습도, 기온, 이슬점 온도, 강수 확률, 풍속이고, 사용된 알고리즘은 AdaBoost, XGBoost, CatBoot이다. 개발 모형을 평가하기 위해 일반국도 순찰 차량을 이용해 수집한 노면 온도 데이터를 이용했다. 노면 온도 데이터 분석 결과, 블랙아이스는 기온(氣溫)이 4℃ 이하이고 상대습도가 75% 이상일 때만 발생하는 것으로 나타났다. 또한, 눈, 비가 내리지 않는 경우 블랙아이스는 기온이 전날에 비해 낮아질 때보다는 올라갈 때 발생할 가능성이 높은 것으로 분석되었다. 부스팅 알고리즘을 이용하여 구축한 블랙아이스 추정 모형 평가 결과, 정확도가 약 90%로 나타나 만족할 만한 성능을 보였다. 본 연구 결과를 일반국도 순찰 구간 선정에 활용할 경우, 블랙아이스 발생 위험이 높은 구간을 집중적으로 순찰하고 순찰 중 블랙아이스를 발견할 경우 제설제를 살포할 수 있어, 겨울철 도로관리를 더욱 효율적이고 효과적으로 수행할 수 있을 것으로 기대된다.

키워드
블랙아이스
부스팅 알고리즘
도로순찰
일반국도
노면온도
References
1

Abdullah G. M. S., Ahmad M., Babur M., Badshah M. U., Al-Mansob R. A., Gamil Y., Fawad M. (2024), Boosting- based Ensemble Machine Learning Models for Predicting Unconfined Compressive Strength of Geopolymer Stabilized Clayey Soil, Scientific Reports, 14, 2323, Nature Portfolio.

10.1038/s41598-024-52825-738282061PMC10822860
2

ASTM International (2008), Standard Practice for Dealing with Outlying Observations, E 178- 08.

3

Bogren J., Caran P. E. (2010), SRIS-Slippery Road Information System, Intelligent Vehicle Safety Systems.

4

Chen T., Guestrin C. (2016), XGBoost: A Scalable Tree Boosting System, Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, August 13-17, 785-794.

10.1145/2939672.2939785
5

Claudia D. N., Roberto C., Gianluca A., Guido B. (2016), Thermal Mapping as a Valuable Tool for Road Weather Forecast and Winter Road Maintenance: An Example from the Italian Alps, Proceedings of the Fourth International Conference on Remote Sensing and Geo-information of the Environment, Cyprus Remote Sensing Society.

6

Dorogush A., Ershov V., Gulin A. (2018), CatBoost: Gradient Boosting with Categorical Features Support, ArXiv:1810.11363.

7

Gustavsson T. (1990), Variation in Road Surface Temperature due to Topography and Wind, Theor. Appl. Climatol., 41, 227-236.

10.1007/BF00866454
8

Hancock J., Khoshgoftaar T. (2020), CatBoost for Big Data: An Interdisciplinary Review, Journal of Big Data 7, Springer Open.

10.1186/s40537-020-00369-833169094PMC7610170
9

Khan A., Ziad A., Subaii A. (2021), Boosting Algorithm Choice in Predictive Machine Learning Models for Fracturing Applications, SPE/IATMI Asia Pacific Oil & Gas Conference and Exhibition (Virtual).

10.2118/205642-MS
10

Kim Y., Lee M., Yun Y., Jun Y., Kim K. (2022), Development of a Forecasting Model for Traffic Accident Probability on Icy Roads using Deep Learning, J. Korean Soc. Transp., 40(1), Korean Society of Transportation, 111-127.

10.7470/jkst.2022.40.1.111
11

KoROAD (2023), Road Traffic Accident Statistics, Korea Road Safety Authority.

12

Lake A. (2023), Rapid Road Weather Hazard Forecasting using Machine Learning, Weather, 78(6), Royal Meteorological Society.

10.1002/wea.4382
13

Lawrence M. G. (2005), The Relationship between Relative Humidity and the Dew point Temperature in Moist Air: A Simple Conversion and Applications, Bulletin of the American Meteorological Society, 86(2), American Meteorological Society.

10.1175/BAMS-86-2-225
14

Lee C., John E. T., Andrew V. B. (2001), Modeling of Road Surface Temperature from a Geographical Parameter Database (Part 2: Numerical), Meteorol. Appl., 8, 421-436.

10.1017/S1350482701004042
15

Lee H., Kang M., Song J., Hwang K. (2022), Pix2Pix-Based Data Augmentation Method for Building an Image Dataset of Black Ice, J. Korean Soc. Transp., 40(4), Korean Society of Transportation, 539-554.

10.7470/jkst.2022.40.4.539
16

Lee M., Kim Y., Jun Y., Shin Y. (2019), Random Forest Based Prediction of Road Surface Condition using Spatio-Temporal Features, J. Korean Soc. Transp., 37(4), Korean Society of Transportation, 338-349.

10.7470/jkst.2019.37.4.338
17

Louis-Philippe C., Yves D. (2001), METRo: A New Model for Road-Condition Forecasting in Canada, Journal of Applied Meteorology, 40, American Meteorological Society, 2026-2037.

10.1175/1520-0450(2001)040<2026:MANMFR>2.0.CO;2
18

Luque P., Wideberg J., Mantars D. (2013), ITS to Improve Safety and Efficiency OBD-II and Smartphone Apps, CreateSpace Independent Publishing Platform.

19

Mats R., Torbj G., Jorgen B., Per-Erik J. (2012), Ice Formation Detection on Road Surfaces using Infrared Thermometry, Cold Regions Science and Technology, Elsevier, 83-84, 71-76.

10.1016/j.coldregions.2012.06.004
20

MOLIT (2023), Road Snow Removal Manual, Ministry of Land, Infrastructure, and Transport.

21

PIARC Technical Committee (2018), Snow and Ice Databook 2018, PIARC.

22

Shao J., Lister P. J. (1995), Data Filtering for Thermal Mapping of Road Surface Temperatures, Meteorol. Appl., 2, 131-135.

10.1002/met.5060020206
23

Son Y., Park S. (2018), A Estimating Model of Road Surface Varying Condition Considering Sunshine Duration Affected by Structure and Topography Near Corresponding Roadway, J. Korean Soc. Transp., 36(4), Korean Society of Transportation, 289-304.

10.7470/jkst.2018.36.4.289
24

Song I., Shin V. (2012), Robust Urban Road Surface Monitoring System using Bayesian Classification with Outlier Rejection Algorithm, 2012 12th International Conference on Control, Automation and Systems, IEEE.

25

Sonntag D. (1990), Vapour Pressure Formulations Based on the IST-90 and Psychrometer Formulae, Z. Meteorol., 70 (5), 340-344.

26

Tina M. G., Eugene S. T. (2006), Bridge Frost Prediction by Heat and Mass Transfer Methods, Journal of Applied Meteorology and Climatology, 45, American Meteorological Society, 517-525.

10.1175/JAM2356.1
27

U.S. FHWA [Online], Available: https://ops.fhwa.dot.gov/weather/weather_events/snow_ice.htm.

28

Veronica B., Adrian R., Tilmann G., Richard S. (2010), Probabilistic Weather Forecasting for Winter Road Maintenance, Journal of American Statistical Association, American Statistical Association, 105(490), 522-537.

10.1198/jasa.2009.ap07184
29

Virve K. (2019), Observing and Forecasting Road Surface Temperatures, Ph.D. Dissertation, University of Helsinki, Finland.

30

Yumei H., Esben A., Torbjorn G., Jrgen B. (2019), Modeling Road Surface Temperature from Air Temperature and Geographical Parameters-Implication for the Application of Floating Car Data in a Road Weather Forecast Model, Jour. of Appli. Meteolo. and Clima., American Meteorological Society, 58, 517-525.

10.1175/JAMC-D-18-0145.1
31

Zhu J., Zou H., Rosset S., Hastie T. (2009), Multi-class AdaBoost, Statistics and Its Interface, 2, International Press of Boston, 349-360.

10.4310/SII.2009.v2.n3.a8
Information
  • Publisher :Korean Society of Transportation
  • Publisher(Ko) :대한교통학회
  • Journal Title :Journal of Korean Society of Transportation
  • Journal Title(Ko) :대한교통학회지
  • Volume : 42
  • No :3
  • Pages :297-312
  • Received Date : 2024-03-22
  • Revised Date : 2024-04-12
  • Accepted Date : 2024-05-29
  • DOI :https://doi.org/10.7470/jkst.2024.42.3.297
Journal Informaiton Journal of Korean Society of Transportation Journal of Korean Society of Transportation
Archives
: 2015 Vol.34~
  • NRF
  • KOFST
  • crossref cited-by
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close