XAI 기법을 활용한 트립 체인 데이터 기반 화물운전자 통행 식별 모델 개발
Development of a Freight Driver Travel Identification Model Using XAI Techniques Based on Trip Chain Data
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
국내 화물의 약 90%는 도로를 통해 운송되는 만큼, 화물운송은 도로의 유지보수 및 교통안전의 측면에서 체계적인 관리가 필요하다. 다만, 기존 화물차량 기종점(origin-destination, O-D) 은 대면조사 및 가구통행실태조사를 기반으로 하고 있기 때문에, 지속가능성, 신뢰성, 시공간적 해상도 등 여러 측면에서 한계를 가진다. 이러한 한계점을 극복하기 위해 모바일 이동통신데이터를 활용하는 방법이 주목받고 있으나, 물류 통행의 분석 및 화물 운송에 이를 활용한연구는 수행되지 않고 있다. 모바일 이동통신 데이터는 높은 표본 수를 가진다는 장점이 있으나, 모바일 통신기기를 소지한 모든 이용자를 대상으로 수집되기 때문에 화물운전자와 비화물운전자의 통행 식별이 선행되어야 한다. 따라서, 본 연구는 모바일 이동통신 데이터를 통해 구축된 트립 체인 데이터의 물류 네트워크 및 O-D 구축에 활용 가능성을 판단하기 위해 트립체인 데이터 기반 화물운전자 통행 식별 모델을 개발하였으며, 설명가능한 AI를 적용하여 화물운전자와 비화물운전자의 통행 특성 차이를 확인하였다. 본 연구를 통해 모바일 이동통신데이터를 활용하여 화물운전자를 분류할 수 있음을 확인하였으며, 향후 다양한 모델과 추가데이터 수집을 통해 성능을 향상시키는 경우, 화물차량 기종점 구축 및 국내 물류 기본계획등에서도 활용될 수 있을 것으로 기대한다.
Abstract
Approximately 90% of domestic cargo is transported by road, necessitating systematic management for road maintenance and traffic safety. However, existing origin-destination (O-D) surveys rely on face-to-face and traffic data collection methods, which have limitations in sustainability, reliability, and spatio-temporal resolution. To address these issues, mobile communication data has gained attention due to its large sample size, but distinguishing cargo drivers from non-cargo drivers is essential. This study developed a cargo driver traffic identification model based on trip chain data from mobile communication data, evaluating its potential for logistics network and O-D construction. Using explainable AI, the study identified traffic characteristic differences between cargo and non-cargo drivers. The findings confirm that mobile communication data can classify cargo drivers, and with enhanced models and additional data, it could support cargo vehicle modeling and national logistics planning.
참고문헌 (29)
- Chen, X., "Trip mode inference from mobile phone signaling data using Logarithm Gaussian Mixture Model", The Journal of Transport and Land Use, 13(1), pp.429~445 (2019)
- Chinu, "Explainable AI : To reveal the logic of black-box models", New Generation Computing, 42(1), pp.53~87 (2024)
- Cho, B. C., "A study on the visiting areas classification of cargo vehicles using dynamic clusting method", The Journal of The Korea Institute of Intelligent Transport Systems, 22(6), pp.141~156 (2023)
- Cho, B. C., "Algorithm eevelopment for extract O", D of air passenger via mobile telecommunication bigdata, 2023, pp.8(2), 1~13
- Cho, B. C., "A trip mobility analysis using big data", The Korea Journal of BigData, 5(2), pp.85~95 (2020)
- Choo, S. H., "Exploring characteristics on trip chaining : The case of Seoul", Journal of Korean Society of Transportation, 26(4), pp.87~97 (2008)
- Chun, Y. E., "Study on credit ration model using explainable AI", Journal of the Korean Data & Information Science Society, 32(2), pp.283~295 (2021)
- Elman, J. L., "Finding Structure in Time", Cognitive Science, 14, pp.179~211 (1990)
- Essadeq, I., "Use of Mobile Telecommunication Data in Transport Modelling", International Transport Forum (2021)
- Goodfellow, I., "Deep Learning" (2016), MIT Press
- Han, Y. H., "A study of measuring traffic congestion for urban network using average link travel time based on DTG Big Data", The Journal of The Korea Institute of Intelligent Transport Systems, 16(5), pp.72~84 (2017)
- Hochreiter, S., "Long short-term memory", Neural Computation, 9, pp.1735~1780 (1997)
- Karim, M. M., "Toward explainable artificial intelligence for early anticipation of traffic accidents", Transportation Research Record, 2676(6), pp.743~755 (2022)
- Kim, D. H., "Estimation of the Origin-Destination Trip Matrices using the Mobile Data", Doctor, pp.2021
- Kim, H. J., "Prediction of Vessel Fuel Consumption Using Explainable Artificial Intelligence (XAI)", Master, pp.2023
- Kim, J. Y., "Upgrading Reliability of Estimating Big Data-driven Transport Demand", Korea Transport Institute, ~ (2019)
- Analysis of Socio-Economy Activities using Mobility Big Data, General Research, "Kim, J.", Korea Transport Institute, ~ (2020)
- "Freight in KOREA" (2017)[URL]
- "Logistics Base Freight Survey Final Report(2019 Edition)" (2019)[URL]
- Lee, J. Y., "XAI analysis of DNN using PM2.5 component input data and improvement of PM2.5 prediction performance", Journal of Korean Society for Atmospheric Environment, 39(4), pp.411~426 (2023)
- Lee, Y. J., "Development of deep learning based deterioration prediction model for the maintenance planning of highway pavement", Korean Journal of Construction Engineering and Management, 20(6), pp.34~43 (2019)
- Lundberg, S. M., "A unified approach to interpreting model predictions", Advances in Neural Information Processing Systems, pp.4765~4774 (2017)
- "Performance of Transport Sector" (2021)[URL]
- Issues in Preliminary Feasibility Study for Transportation Sector(II), "Park, S. J.", Korea Development Institute, ~ (2020)
- Ribeiro, M. T., "“Why should I trust you?” Explaining the predictions of any classifier", Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp.1135~1144 (2016)
- Rosenblatt, F., "The perceptron: A probabilistic model for information storage and organization in the brain", Psychological Review, 65(6), pp.386~408 (1958)
- Wojtas, M., "Feature importance ranking for deep learning", Advances in Neural Information Processing Systems, 33, pp.5105~5114 (2020)
- Won, M. S., "Development of an algorithm to analyze home-based work trips using mobile phone data", Journal of Korean Society of Transportation, 39(3), pp.383~398 (2021)
- A Study on Usability of Big Data to Enhance Reliability of Regional Travel Demand Forecasting, "Yun, S. Y.", Korea Research Institute for Human Settlements, ~ (2015)