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

Speed Prediction of Urban Freeway Using LSTM and CNN-LSTM Neural Network

LSTM 및 CNN-LSTM 신경망을 활용한 도시부 간선도로 속도 예측

  • Park, Boogi (Dept. of Spatial Information Eng., Pukyong National University) ;
  • Bae, Sang hoon (Dept. of Spatial Information Eng., Pukyong National University) ;
  • Jung, Bokyung (Dept. of Spatial Information Eng., Pukyong National University)
  • 박부기 (부경대학교 지구환경시스템과학부 공간정보시스템공학전공) ;
  • 배상훈 (부경대학교 공간정보시스템공학과) ;
  • 정보경 (부경대학교 공간정보시스템공학과)
  • Received : 2020.12.28
  • Accepted : 2021.01.25
  • Published : 2021.02.28
Download PDF
⟨ Previous Next ⟩

Abstract

One of the methods to alleviate traffic congestion is to increase the efficiency of the roads by providing traffic condition information on road user and distributing the traffic. For this, reliability must be guaranteed, and quantitative real-time traffic speed prediction is essential. In this study, and based on analysis of traffic speed related to traffic conditions, historical data correlated with traffic flow were used as input. We developed an LSTM model that predicts speed in response to normal traffic conditions, along with a CNN-LSTM model that predicts speed in response to incidents. Through these models, we try to predict traffic speeds during the hour in five-minute intervals. As a result, predictions had an average error rate of 7.43km/h for normal traffic flows, and an error rate of 7.66km/h for traffic incident flows when there was an incident.

교통혼잡을 완화하기 위한 방안 중 하나로 도로 이용자에게 교통상황 예측정보를 제공함으로써 교통량을 분산 시켜 도로 이용 효율을 증대시키는 방법이 있다. 이를 위해서는 신뢰성이 보장되고 정량적인 실시간 교통 속도 예측이 필수적이다. 본 연구에서는 상황별 교통속도 분석을 기반으로 이력 속도 데이터와 이력 속도 외의 교통류에 상관관계가 있는 데이터를 LSTM 입력 데이터로 활용하였다. 정상 교통류 상황에 대응하여 속도를 예측하는 LSTM 모델과 유고상황에 대응하여 속도를 예측하는 CNN-LSTM 모델을 개발하여 유고발생 후 1시간까지 5분 단위로 교통속도 예측을 시도하였다. 모델의 검증은 테스트 데이터를 통하여 교통상황별 예측성능을 분석하였다. 그 결과 정상 교통류에서는 평균 7.43km/h, 유고상황에서는 7.66km/h의 오차율로 각각 예측되었다.

Keywords

References

  1. ALIO(2013), Korea Road Traffic Authority Traffic Science Institute, pp.86-87.
  2. FHWA(2006), Traffic Detector Handbook, pp.3-17.
  3. Jung H. J., Yoon J. S. and Bae S. H.(2017), "Traffic Congestion Estimation by Adopting Recurrent Neural Network," The Journal of the Korea Institute of Intelligent Transport Systems, vol. 16, no. 6, pp.67-78. https://doi.org/10.12815/kits.2017.16.6.67
  4. Kim D. H., Hwang K. Y. and Yoon Y.(2019), "Prediction of Traffic Congestion in Seoul by Deep Neural Network," The Journal of the Korea Institute of Intelligent Transport Systems, vol. 18, no. 4, pp.44-57.
  5. Korea Meteorological Administration Data Center ASOS Data, https://data.kma.go.kr, 2020.07.27.
  6. Lee H. S. and Bui K. H. N.(2019), "Deep Learning LSTM for Long-Short Term Traffic Flow Prediction," The Korean Institute of Information Scientists and Engineers 2019, pp.724-726.
  7. Park S. H., Choi D. J., Bok K. S. and Yoo J. S.(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
  8. Sepp H. and Jurgen S.(1997), "Long Short-Term Memory," Neural Computation, vol. 9, no. 8, pp.1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735
  9. Simard P. Y., Steinkraus D. and Platt J. C.(2003), "Best Practices for Convolutional Neural Networks Applied to Visual Document Analysis," Proceedings of the 7th International Conference on Document Analysis and Recognition, no. 3, pp.958-963.
  10. The Korea's Transport Institute Press Release, https://www.koroad.or.kr, 2020.10.13.
  11. Traffic Accident Analysis System, http://taas.koroad.or.kr, 2020.10.13.
  12. Zheng H., Lin F., Feng X. and Chen Y.(2020), "A Hybrid Deep Learning Model With Attention-Based Conv-LSTM Networks for Short-Term Traffic Flow Prediction," IEEE Transactions on Intelligent Transportation Systems, pp.1-11.