Human Driving Data based Yield Intention Inference Algorithm for Lane Change of Autonomous Vehicles in Congested Unrban Traffic
- 提供方法
- 本サイト上にてダウンロード・閲覧可
- 形態
- 価格
- 一般価格(税込):¥1,100 会員価格(税込):¥880
- 文献番号
- 20219007
- 文献・情報種別
- その他の国際会議
- 掲載ページ
- 1-5(Total 5 p)
- 発行年月
- 2021年 9月
- 出版社
- (公社)自動車技術会
- 言語
- 英語
書誌事項
著者(英) | 1) Dabin Seo, 2) Heungseok Chae, 3) Kyongsu Yi |
---|---|
勤務先(英) | 1) Seoul National University, 2) Seoul National University, 3) Seoul National University |
抄録(英) | This paper describes LSTM-based RNN yield intention ininference algorithm for the interactive lane change in congested urban traffic situations, which is regarded as one of the significant challenges on urban autonomous driving. The performance of each LSTM-based RNN network was evaluated through MSE (Mean Squared Error) and calculation time for real vehicle implementation. The optimal prediction horizon, input variables, network structure, and optimized method were determined based on MSE and calculation time. The finally selected, through ablative study, network considers the longitudinal and lateral interactions with the surrounding vehicles over time. The proposed yield intention inference algorithm has been evaluated through a real vehicle test. 翻訳 |