DSRC와 TCS 정보를 이용한 고속도로 경로통행시간 예측 = Forecasting of Motorway Path Travel Time by Using DSRC and TCS Information

국문 초록 (Abstract)

출발지 기준 고속도로 경로 통행시간(PTTDP)은 첨단여행자정보시스템(ATIS)의 핵심 정보이다. 이러한 필요성에도 불구하고, 지능형교통체계(ITS)의 예측분야에서 PTTDP에 대한 연구는 성공적으로 극복해야할 핵심 도전과제중 하나로 남아있는 상태이다. 이러한 문제점을 효과적으로 극복하기 위하여, 본 연구에서는 고속도로 IC간 경로통행시간을 동적으로 예측하는 방법론을 제시하고자 한다. 제안된 모형은 고속도로망에서 TG의 교통수요와 TG간 출발지기준 경로통행시간간의 관계를 기반으로 개발되었다. 모형의 입력 자료로(TCS로 수집되는) 통행수요와 (DSRC로 수집되는) 경로통행시간 자료가 이용되었다. 개발 모형은 고속도로 정보시스템에 탑재/운영하기 위하여 Data Ming 기법중 연산속도가 빠른 k-최근린 이웃을 이용하였다. 실제 자료를 이용한 적용 실험에서, 제안된 모형은 예측의 신뢰성과 연산수행속도 측면에서 ATIS에 적용이 가능한 수준의 성능을 보였다.

다국어 초록 (Multilingual Abstract)

Path travel time based on departure time (PTTDP) is key information in advanced traveler information systems (ATIS). Despite the necessity, forecasting PTTDP is still one of challenges which should be successfully conquered in the forecasting area of intelligent transportation systems (ITS). To address this problem effectively, a methodology to dynamically predict PTTDP between motorwayinterchanges is proposed in this paper. The method was developed based on the relationships between traffic demands at motorway tollgates and PTTDPs between TGs in the motorway network. Two different data were used as the input of the model: traffic demand data and path travel time data are collected by toll collection system (TCS) and dedicated short range communication (DSRC), respectively. The proposed model was developed based on k-nearest neighbor, one of data mining techniques, in order for the real applications of motorway information systems. In a feasible test with real-world data, the proposed method performed effectively by means of prediction reliability and computational running time to the level of real application of current ATIS.

참고문헌 (Reference)

1 심상우, "하이패스 DSRC 자료를 활용한 고속도로 오프라인 경로통행시간 추정기법 개발" 대한교통학회 31 (31): 45-54, 2013

2 Oswald, R. K., "Traffic Flow Forecasting using Approximate Nearest Neighbor Nonparametric Regression" U.S. DOT University transportation center 2000

3 Karlsson, M., "Rainfall-Runoff Forecasting Methods, Old and New" 1 (1): 303-318, 1987

4 Namkoong, S., "Progressive Interative Forward and Backward (PIFAB) Search Method to Estimate Path-Travel Time on Freeways Unsing Toll Collection System Data" 20 (20): 147-155, 2005

5 Yoon, B. J., "Potentialities of Data-Driven Nonparametric Regression in Urban Signalized Traffic Flow Forecasting" 2014

6 Davis, G., "Nonparametric Regression and Short-Term Freeway Traffic Forecasting" 117 (117): 178-188, 1991

7 Robinson, P., "Nonparametric Estimators for Time Series" 4 (4): 185-207, 1983

8 Chang, H. H., "Multiple Time Period Imputation Technique for Multiple Missing Traffic Variables: Nonparametric Regression Approach" 39 : 448-459, 2012

9 Qi, Y., "Identifying Nearest-Neighbors in a Large-Scale Incident Data Archive" 1879 : 89-98, 2004

10 Chang, H. H., "Dynamic Near-Term Traffic Flow Prediction: System-Oriented Approach Based on Past Experiences" 6 (6): 292-305, 2012

11 Chang, H. H., "Dynamic Multi-Interval Bus Travel Time Prediction using Bus Transit Data" 6 (6): 19-38, 2010

12 Namkoong, S., "Development of Optimal-Path Finding System (X-PATH) using Search Space Reduction Technique Based on Expert System" 14 (14): 51-67, 1996

13 Smith, B. L., "Comparison of Parametric and Nonparametric Models for Traffic Flow Forecasting" 10 (10): 303-321, 2002

14 Mulhern, F. J., "A Nearest Neighbor Model for Forecasting Market Response" 10 (10): 191-207, 1994