Multi-agent 주행 시뮬레이션을 이용한 운전자 주행패턴을 반영한 공격운전 검지기법 개발

Development of a Methodology for Detecting Intentional Aggressive Driving Events using Multi-agent Driving Simulations
  • 김윤종; 
  • 오철; 
  • 최병호; 
  • 최새로나; 
  • 김기용

초록

Intentional aggressive driving (IAD) is defined as a hazardous driving event that the aggressive driver intentionally threatens neighbor drivers with abrupt longitudinal and lateral maneuvering. This study developed a methodology for detecting IAD events based on the analysis of interactions between aggressive driver and normal driver. Three major aggressive events including rear-close following, side-close driving, and sudden deceleration were analyzed to develop the algorithm. Then, driving simulation experiments were conducted using a multi-agent driving simulator to obtain data to be used for the development of the detection algorithm. In order to detect the driver’s intention to attack, a relative evaluation index (Erratic Driving Index, EDI) reflecting the driving pattern was derived. The derived IAD event detection algorithm utilizes both the existing absolute detection method and the relative detection method. It is expected that the proposed methodology can be effectively used for detecting IAD events in support of in-vehicle data recorder technology in practice.

키워드

driving pattern; driving simulation; EDI; intentional aggressive driving; multi-agent; 공격운전; 상대적 검지기법; 주행패턴; 주행시뮬레이션; multi-agent
제목
Multi-agent 주행 시뮬레이션을 이용한 운전자 주행패턴을 반영한 공격운전 검지기법 개발
제목 (타언어)
Development of a Methodology for Detecting Intentional Aggressive Driving Events using Multi-agent Driving Simulations
저자
김윤종; 오철; 최병호; 최새로나; 김기용
DOI
10.7470/jkst.2018.36.1.051
발행일
2018-02
저널명
대한교통학회지
권
36
호
1
페이지
51 ~ 65