버스정차시간과 예측구간 추정 방법- 휴스톤 노선버스 사례연구 -
A Methodology for Estimation of Bus Dwell Time and Prediction Intervals: A Case Study of Houston Metro
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
대중교통우선신호체계의 성패는 얼마나 버스도착시간을 정확하게 예측하는 가에 달려있다. 버스도착시간은 버스정차시간에 의해 크게 영향을 받지만 버스정차시간의 불확실성으로 인해 그 예측이 매우 어렵다. 본 연구에서는 확률모형과 가중회귀분석모형을 사용하여 버스정차시간을 예측하였다. 확률모형은 버스정차시간에 영향을 미치는 요인들을 확률적으로 설명할 수 있는 장점이 있지만 요인간의 선형상관관계를 바탕으로 하기 때문에 이 가정이 만족되지 않는 상황에서는 적용이 불가능하다. 이러한 한계를 극복하기 위해 가중회귀분석법을 활용하여 정차시간을 예측하는 방법과 정차시간의 예측구간을 추정하는 방법론을 제시하였다.
Abstract
The variability in bus dwell time is the key issues in advanced traveler information system(ATIS) and transit signal priority(TSP) systems because the successful estimation of the bus arrival time is greatly dependent on the accurate prediction of the bus dwell time at stops. In this paper, a probabilistic model and the weighted least squares (WLS) regression model were developed to predict the bus dwell time. Although the probabilistic model showed the strength in representing the determinants of dwell time as random variables, the observed data did not support the model assumption of the linear correlation between the number of alighting passengers and the passenger loads. It was concluded from an analysis of the dwell time determinants that the bus headway is most significant determinant and that the regression model with the bus headway as an independent variable gave the best results. Because of the non-constant variance over bus headways, the weighted least squares method was employed in order to consider the non-constant variance. Another advantage of the WLS method is that the prediction interval may be readily calculated. The prediction interval is more informative than mean dwell time in supporting ATIS and TSP systems.
참고문헌 (16)
- "Probabilistic Model of Passenger Service Processes at Bus Stops" (1992)
- "Transformation and Weighting in Regression" (1998), Chapman and Hall
- "EEnhancements of CORSIM Model in Simulation Transit Operations", 126(5), pp.396~404 (2000)
- "Dynamic Bus Arrival Time Prediction with Artificial Neural Networks", 128(5), pp.429~438 (2002)
- "Journal of Public Transportation", 21(1), pp.~40 (2004)
- "Modeling Bus Delays due to Passenger Boardings and Alightings" (1983)
- "An Improved Bus Signal Priority System for Networks with Nearside Bus Stops, Ph.D. dissertation" (2004), Texas A&M University: College Station
- "Improved Transit Signal Priority System for Networks with Nearside Bus Stops", 1925, pp.205~214 (2005)
- "Passenger Counter Evaluation: Implications for National Transit Database Reporting", 1835, pp.109~119 (2003)
- "Evaluation of Passenger Service Times for Street Transit Systems" (1974)
- "Dwell Time Relationships for Light Rail Systems" (1992)
- "IIntroduction to Probability and Statistics: Principles and Applications for Engineering and the Computing Sciences" (1995), McGraw-Hill International Editions
- "Applied Linear Statistical Models" (1996), McGraw-Hill[DOI]
- "A Probabilistic Model of Bus Route Performance" (1983)
- "Estimation of Bus Dwell Times with Automatic Passenger Counter Information" (2003)
- "Highway Capacity Manual" (2000), National Research Council