A Study on an ETCS Demand Forecasting Model of Toll Roads in Changwon City

유료도로 ETCS 이용수요 예측모형에 관한 연구 (창원시를 중심으로)

  • 김경환 (경상대학교 건설공학부) ;
  • 하만복 (경상대학교 환경 및 지역발전연구소) ;
  • 전연후 (교통안전공단 부산경남지사) ;
  • 이익수 (경상대학교 대학원 도시공학과)
  • Published : 2007.03.15

Abstract

Since early 1990s, several developed countries have applied the Electronic Toll Collection System (ETCS) to toll roads in order to solve traffic congestion and delay problems at toll plazas. For the successful operation of the ETCS, it is important to correctly forecast the ETCS using rate. In this study, it was conceived to develop a sophisticated demand forecasting model of the ETCS for toll roads in Changwon City The Binary Logit and neural network models were tested for the model considering 11 explaining variables. The best results in prediction accuracy and goodness-of-fit were obtained on the neural network model. However, because of the difficulty in predicting the 11 variables and its fitness in wide range, the Binary Logit model which considers three policy variables only is recommended as the model to forecast the ETCS using rate.