상세 보기
실시간 소통정보 및 LSTM을 이용한 주변 차량 모델 개발
Development of Surrounding Vehicle Model Using LSTM and Real-Time Traffic Information
- 이승연;
- 지용혁;
- 이형철
초록
In this paper, we develop a surrounding vehicle velocity profile generation model based on deep learning method. An LSTM network is developed for generating human-like velocity profile. Real driver’s driving data for training LSTM network is obtained from DHIL simulation environment. DHIL environment created with HIL simulator and driving wheel. Traffic data can be applied by using Real-time traffic information provided by Seoul TOPIS(Transport Operation & Information Service). For verifying generated velocity profile, LSTM-based velocity profile was compared with PI controller-based profile and DHIL data.
키워드
Traffic Information(소통정보); AI(인공지능); LSTM(장단기 메모리); Time-series Prediction(시계열 예측)
- 제목
- 실시간 소통정보 및 LSTM을 이용한 주변 차량 모델 개발
- 제목 (타언어)
- Development of Surrounding Vehicle Model Using LSTM and Real-Time Traffic Information
- 저자
- 이승연; 지용혁; 이형철
- 발행일
- 2022-11
- 유형
- Proceeding
- 저널명
- 2022 한국자동차공학회 추계학술대회
- 페이지
- 2109 ~ 2113