The Journal of The Korea Institute of Intelligent Transport Systems (한국ITS학회 논문지)

The Korea Institute of Intelligent Transport Systems (한국ITS학회)
권호 표지
DOI QR코드

DOI QR Code

Methodology for Developing a Predictive Model for Highway Traffic Information Using LSTM

LSTM을 활용한 고속도로 교통정보 예측 모델 개발 방법론

  • Yoseph Lee (Dept. of Transportation Eng., Ajou Univ.) ;
  • Hyoung-suk Jin (Overseas Consulting Team, EZENSYS Co.,Ltd) ;
  • Yejin Kim (Dept. of Transportation Eng., Ajou Univ.) ;
  • Sung-ho Park (Dept. of Center for Convergence and Open Sharing, Ajou Univ.) ;
  • Ilsoo Yun (Dept. of Transportation Eng., Ajou Univ.)
  • 이요셉 (아주대학교 교통공학과) ;
  • 진형석 ((주) 이젠시스 해외컨설팅팀) ;
  • 김예진 (아주대학교 교통공학과) ;
  • 박성호 (아주대학교 혁신융합단) ;
  • 윤일수 (아주대학교 교통시스템공학과)
  • Received : 2023.07.20
  • Accepted : 2023.08.21
  • Published : 2023.10.31
Download PDF
⟨ Previous Next ⟩

Abstract

With the recent developments in big data and deep learning, a variety of traffic information is collected widely and used for traffic operations. In particular, long short-term memory (LSTM) is used in the field of traffic information prediction with time series characteristics. Since trends, seasons, and cycles differ due to the nature of time series data input for an LSTM, a trial-and-error method based on characteristics of the data is essential for prediction models based on time series data in order to find hyperparameters. If a methodology is established to find suitable hyperparameters, it is possible to reduce the time spent in constructing high-accuracy models. Therefore, in this study, a traffic information prediction model is developed based on highway vehicle detection system (VDS) data and LSTM, and an impact assessment is conducted through changes in the LSTM evaluation indicators for each hyperparameter. In addition, a methodology for finding hyperparameters suitable for predicting highway traffic information in the transportation field is presented.

최근 빅데이터 및 딥러닝 기술의 발전으로 다양한 교통정보가 널리 수집 및 활용되고 있다. 특히 시계열 특성을 갖는 교통정보 예측 분야에서는 장단기 메모리(long short term memory, LSTM)가 널리 사용되고 있다. LSTM에 입력되는 시계열 데이터의 추세, 계절성, 주기 등이 상이하기 때문에 시계열 데이터를 기반으로 한 예측 모델에서도 데이터의 특성에 따라 하이퍼 파라미터의 적합한 값을 찾는 시행착오법이 필수적이다. 이에 적합한 하이퍼 파라미터를 찾는 방법론이 정립된다면, 정확도가 높은 모델 구성에 소요되는 시간을 줄일 수 있다. 따라서, 본 연구에서는 국내 고속도로 차량검지기 데이터와 LSTM을 기반으로 교통정보 예측 모델을 개발하였으며, LSTM의 하이퍼 파라미터별 평가지표 변화를 통해 예측 결과에 미치는 영향평가를 수행하였다. 또한, 이를 기반으로 교통분야에서 고속도로 교통정보 예측에 적합한 하이퍼 파라미터를 찾는 방법론을 제시하였다.

Keywords

Acknowledgement

본 논문은 국토교통부 자율주행 기술개발 혁신사업 '주행 및 충돌상황 대응 안전성 평가기술개발(RS-2021-KA160637)' 과제 지원에 의해 수행되었습니다.

References

  1. Brownlee, J.(2018), "Better Deep Learning: Train Faster, Reduce Overfitting, and Make better Predictions", Machine Learning Mastery, p.540.
  2. Bushaev, V.(2018), Understanding RMSprop-faster neural network learning, June, 13, 2023. https://towardsdatascience.com/understanding-rmsprop-faster-neural-network-learning-62e116fcf29a, 2023.07.19.
  3. Goodfellow, I., Bengio, Y. and Courville, A.(2016), Deep learning, MIT Press.
  4. Hossain, M. D., Ochiai, H., Fall, D. and Kadobayashi, Y.(2020), "LSTM-based network attack detection: Performance comparison by hyper-parameter values tuning", 2020 7th IEEE International Conference on Cyber Security and Cloud Computing (CSCloud)/2020 6th IEEE International Conference on Edge Computing and Scalable Cloud (EdgeCom), pp.62-69.
  5. Hyndman, R. K. and Athanasopoulos, G.(2018), Forecasting: Principles and Practice (3rd ed.), OTexts: Melbourne, Australia. OTexts.com/fpp3, 2023.07.19.
  6. Kim, H. J., Park, S. H. and Jang, K. T.(2016), "Short-term Traffic States Prediction Using k-Nearest Neighbor Algorithm: Focused on Urban Expressway in Seoul", Journal of Korean Society of Transportation, vol. 34, no. 2, pp.158-167. https://doi.org/10.7470/jkst.2016.34.2.158
  7. Lee, J. and Han, J.(2021), "Layer-wise Relevance Propagation (LRP) Based Technical and Macroeconomic Indicator Impact Analysis for an Explainable Deep Learning Model to Predict an Increase and Decrease in KOSPI", Journal of Korean Institute of Information Scientists and Engineers, vol. 48, no. 12, pp.1289-1297. https://doi.org/10.5626/JOK.2021.48.12.1289
  8. Malhotra, P., Ramakrishnan, A., Anand, G., Vig, L., Agarwal, P. and Shroff, G.(2016), LSTM-based Encoder-Decoder for Multi-sensor Anomaly Detection, arXiv preprint arXiv:1607.00148.
  9. Moon, J., Park, J., Han, S. and Hwang, E.(2017), "Power Consumption Forecasting Scheme for Educational Institutions Based on Analysis of Similar Time Series Data", Journal o fKorean Institute of Information Scientists and Engineers, vol. 44, no. 9, pp.954-965.
  10. Muzaffar, S. and Afshari, A.(2019), "Short-Term Load Forecasts Using LSTM Networks", Energy Procedia, vol. 158, pp.2922-2927. https://doi.org/10.1016/j.egypro.2019.01.952
  11. Olah, C.(2015), Understanding LSTM Networks(2015), Oct, 10, 2022, https://colah.github.io/posts/2015-08-Understanding-LSTMs, 2023.07.19.
  12. Park, B., Bae, S. and Jung, B.(2021), "Speed Prediction of Urban Freeway Using LSTM and CNN-LSTM Neural Network", The Journal of the Korea Institute of Intelligent Transport System, vol. 20, no. 1, pp.86-99. https://doi.org/10.12815/kits.2021.20.1.86
  13. Park, S., Choi, D., Bok, K. and Yoo, J.(2020), "Road Speed Prediction Scheme Considering Traffic Incidents", The Journal of the Korea Contents Association, vol. 20, no. 4, pp.25-37. https://doi.org/10.5392/JKCA.2020.20.04.025
  14. Ruder, S.(2016), An overview of gradient descent optimization algorithms, arXiv preprint arXiv:1609.04747.
  15. Ryu, K. D. and Kim, W. J.(2018), "Comparative Analysis of Time Series Method for Forecasting the Call Arrival of Call Center", The Journal of Korean Institute of Information Technology, vol. 16, no. 8, pp.83-96. https://doi.org/10.14801/jkiit.2018.16.8.83
  16. Shi, J., Jain, M. and Narasimhan, G.(2022), Time series forecasting (tsf) using various deep learning models, arXiv preprint arXiv:2204.11115.
  17. Sohn, E. S. and Kim, J. K.(2021), "FlappyBird Competition System: A Competition-Based Assessment System for AI Course", Journal of Korea Multimedia Society, vol. 24, no. 4, pp.593-600. https://doi.org/10.9717/KMMS.2020.24.4.593
  18. Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I. and Salakhutdinov, R.(2014), "Dropout: A simple way to prevent neural networks from overfitting", The Journal of Machine Learning Research, vol. 15, no. 1, pp.1929-1958.
  19. Yadav, A., Jha, C. K. and Sharan, A.(2020), "Optimizing LSTM for time series prediction in India stock market", Procedia Computer Science, vol. 167, pp.2901-2100.
  20. Yu, J. H. and Kim, J. H.(2010), "Development of an incident impact analysis system using short-term traffic forecasts", Journal of the Korean Society of Road Engineers, vol. 12, no. 4, pp.1-9.