非衝突条件を用いた追従運転データ特性の考察

概要

論文の詳細を見る
In the development of a simulator for the microscopic road traffic (MITRAM), we proposed a fuzzy model for vehicles (FMV) using a neural network and fuzzy approach. In FMV, we can only study the actual data so as to recreate the real succeeding driving operation. In the analysis and evaluation of the traffic simulator used FMV in consecutive time series, the data of vehicular gap is very close to the actual data. But by using the neural network, some parts of the model will become black box. So we couldn't explain which characteristics of succeeding driving operation recreated. In order to recreate the characteristics of the succeeding driving delicately on the simulator, we think it is necessary to analyze the characteristics of the actual data by studying to construct the model. In order to examine the mechanism for succeeding driving operation, we firstly focused on the modeling of the succeeding driving operation on the condition of the non-collision of a preceding vehicle, based on the assumption that the succeeding driving operation is understood as the operation that does not collide with a preceding vehicle. The feature of this model is to suppose that the preceding vehicle decelerate by the minimum acceleration of the actual data, and a succeeding driving operation is such the one that a succeeding vehicle does not collide with a preceding vehicle. Under this supposition in this study, some features of the succeeding driving operation were found from the model of succeeding driving considering non-collision condition.
日本シミュレーション学会の論文
2006-12-15

非衝突条件を用いた追従運転データ特性の考察

スポンサーリンク

概要

著者

関連論文

スポンサーリンク