실시간 도시철도 승・하차량 예측을 위한 STL 기법 적용 연구
Application of STL for Real-time Prediction of Boarding/Alighting Passenger Flow in Seoul Urban Railway Network
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
도시에서의 효율적이고 안전한 도시철도 운영을 위해 이용 승객에 대한 실시간 혼잡도 예측은 필수적이다. 특히, 인파가 집중될 가능성이 높은 역일수록 혼잡 상황의 사전 예측은 더욱 중요하다. 이에 따라, 본 연구는 기준 시점으로부터 과거 4주(28일)의 승·하차량 데이터를 이용하여 특정 역의 승·하차량을 예측하는 모형을 제시하였다. 승·하차량을 15분 단위로 집계한 스마트카드 데이터에 STL 기법을 적용하여 계절, 추세, 잔차 성분으로 나누고 Holt-Winters 계절성 방법을 통해 각 요소별 예측값을 산출 후 합산하여, 승·하차량 예측 모형을 구축하였다. 승·하차량의 변동이 적은 월곡역과 변동이 큰 종합운동장역, 이태원역에 대해 사례분석을 진행하였으며, 모형 성능평가 결과 해당 모형은 승·하차량 변동이 적은 역에서 상대적으로 더 우수한 예측 성능을 나타내는 것을 확인하였다.
Abstract
Real-time prediction of congestion is crucial to an efficient and safe operation of urban railway systems. This is especially so at stations where there is potential of large crowd gatherings, such as Sports Complex and Itaewon stations. As such this study aims to present a real-time short-term passenger flow prediction model based on ridership data from four previous weeks (28 days). STL method was applied to smart card data aggregated at fifteen minute intervals to count boarding and alighting passengers to divide it into Seasonal, Trend, and Residual terms, to each of which a separate Holt-Winters' prediction process was executed. The resulting values were then summed to yield a prediction value. Case study conducted on Wolgok, Itaewon and Sports Complex stations for performance evaluation of the proposed model, showed outstanding performance for subway stations with low fluctuations.
참고문헌 (11)
- Paul de Nailly, "Analyse et prédiction des flux piétons dans un pôle de transport multimodal àpartir de données multi-sources", PhD, pp.2023
- E. Chen, "Subway Passenger Flow Prediction for Special Events Using Smart Card Data", IEEE Transactions on Intelligent Transportation Systems, 21(3), pp.1109~1120 (2019)
- P. Noursalehi, "Real time transit demand prediction capturing station interactions and impact of special events", Transportation Research Part C : Emerging Technologies, 97, pp.277~300 (2018)
- C. Jia, "Analysis of Crowded Propagation on the Metro Network", Sustainability, 14(16), pp.9829~ (2022)
- Y. Li, "Forecasting short-term subway passenger flow under special events scenarios using multiscale radial basis function networks", Transportation Research Part C : Emerging Technologies, 77, pp.306~328 (2017)
- L. Chen, "Short-term traffic flow prediction : From the perspective of traffic flow decomposition", Neurocomputing, 413, pp.444~456 (2020)
- J. Tang, "Seasonal decomposition and combination model for short-term forecasting of subway ridership", International Journal of Machine Learning and Cybernetics, 13(1), pp.145~162 (2022)
- Y. Zhao, "Short-Term Passenger Flow Prediction With Decomposition in Urban Railway Systems", IEEE Access, 8, pp.107876~107886 (2020)
- C. Robert, "STL : A seasonal-trend decomposition procedure based on loess", Journal of Official Statistics, 6(1), pp.3~73 (1990)
- Charles C. Holt, "Forecasting seasonals and trends by exponentially weighted averages", International Journal of Forecasting, 20(1), pp.5~10 (2004)
- Peter R. Winters, "Forecasting sales by exponentially weighted moving averages", Management Science, 6(3), pp.324~342 (1960)