Development of a network traffic breakdown prediction method

2023 Fiscal Year Final Research Report

• Project/Area Number: 21H01457
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 22050:Civil engineering plan and transportation engineering-related
• Research Institution: Ehime University
• Principal Investigator: YOSHII TOSHIO 愛媛大学, 理工学研究科(工学系), 教授 (90262120)
• Co-Investigator(Kenkyū-buntansha): 坪田 隆宏 愛媛大学, 理工学研究科(工学系), 准教授 (00780066) ; 塩見 康博 立命館大学, 理工学部, 教授 (40422993) ; 西内 裕晶 高知工科大学, システム工学群, 教授 (40548096) ; 川崎 洋 九州大学, システム情報科学研究院, 教授 (80361393) ; 小野 晋太郎 福岡大学, 工学部, 准教授 (80526799)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: 交通流 / ブレイクダウン / ネットワークネットワーク / 画像処理 / 予測手法 / 事故リスク

Outline of Final Research Achievements

In order to prevent severe and extraordinary traffic congestion (hereafter referred to as "breakdown") caused by sudden events such as traffic accidents or natural disasters, the following research was conducted with the aim of developing a breakdown occurrence prediction method.
(1)We defined the Area Traffic State as an indicator to macroscopically capture the state of network traffic flow and established a method for estimating the MFD (Macroscopic Fundamental Diagram), which shows the relationship between traffic density and traffic volume. Changes in the shape of the MFD due to variations in OD traffic volume and the occurrence of traffic accidents were identified.(2)To develop a method for predicting network breakdowns, we devised a method for determining the occurrence of breakdowns.(3)We collected DR images from probe vehicles and performed image analysis to develop a model that predicts sudden braking occurrences and detects blind spots.

Free Research Field

交通工学

Academic Significance and Societal Importance of the Research Achievements

交通事故や自然災害などの突発事象に起因する非日常の激しい交通渋滞（以下“ブレイクダウン”）を未然に防ぐことを目的とし，空間モニタリング技術，交通流解析手法，ならびにAIによる情報解析技術を組み合わせることで，道路ネットワーク交通流におけるBreakdown発生予測手法の構築を目指す研究で，同予測手法が確立されれば，東日本大震災後に東京で発生した大渋滞（Gridlock現象）などの交通ネットワークブレイクダウン現象を未然に防ぐことが可能となる．