| Title |
A Study on the Analysis of Bicycle Road Service Level by Using Adaptive Neuro-Fuzzy Inference System |
| Keywords |
ANFIS;Bicycle Road Service Level;Fuzzy Inference System;Neuro-Fuzzy;자전거도로 서비스수준;퍼지추론 시스템;뉴로-퍼지 |
| Abstract |
Currently our country has very serious problems of traffic congestion and urban environment due to increasing automobile ownership. Recently, our concern about environmentally sustainable transportation and green transportation is increasing, so the government is pushing ahead the policy of bicycle using activation. So it is needed to develop a model to analyze the service level of bicycle roads more realistically. In this study, a neuro-fuzzy inference model to analyze the service level of bicycle roads was built selecting the width of bicycle roads, the number of conflicts during cycling and pedestrian volume, which have fuzzy characteristics, as input variables. The predictability of the model was evaluated comparing the surveyed and the estimated. The values of the statistics, R², MAE and MSE were 0.987, 0.142, 0.032. Therefore, It may be judged that the explainability of the model is very high. The service levels of bicyle roads estimated by the model are 1~3 steps lower than KHCM assessments. The reason may be explained that the model estimates the service level considering the width of bicycle roads and the number of conflicts simultaneously besides pedestrian volume. |