하천 범람 및 차량 침수 가능성 예측을 통한 딥러닝 기반 차수막 자동화 시스템
Deep-Learning-Based Water Shield Automation System by Predicting River Overflow and Vehicle Flooding Possibility
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
This paper proposes a two-stage Water Shield Automation System (WSAS) to predict the possibility of river overflow and vehicle flooding due to sudden rainfall. The WSAS uses a two-stage Deep Neural Network (DNN) model. First, a river overflow prediction module is designed with LSTM to decide whether the river is flooded by predicting the river's water level rise. Second, a vehicle flooding prediction module predicts flooding of underground parking lots by detecting flooded tires with YOLOv5 from CCTV images. Finally, the WSAS automatically installs the water barrier whenever the river overflow and vehicle flooding events happen in the underground parking lots. The only constraint to implementing is that collecting training data for flooded vehicle tires is challenging. This paper exploits the Image C&S data augmentation technique to synthesize flooded tire images. Experimental results validate the superiority of WSAS by showing that the river overflow prediction module can reduce RMSE by three times compared with the previous method, and the vehicle flooding detection module can increase mAP by 20% compared with the naive detection method, respectively.
Abstract
This paper proposes a two-stage Water Shield Automation System (WSAS) to predict the possibility of river overflow and vehicle flooding due to sudden rainfall. The WSAS uses a two-stage Deep Neural Network (DNN) model. First, a river overflow prediction module is designed with LSTM to decide whether the river is flooded by predicting the river's water level rise. Second, a vehicle flooding prediction module predicts flooding of underground parking lots by detecting flooded tires with YOLOv5 from CCTV images. Finally, the WSAS automatically installs the water barrier whenever the river overflow and vehicle flooding events happen in the underground parking lots. The only constraint to implementing is that collecting training data for flooded vehicle tires is challenging. This paper exploits the Image C&S data augmentation technique to synthesize flooded tire images. Experimental results validate the superiority of WSAS by showing that the river overflow prediction module can reduce RMSE by three times compared with the previous method, and the vehicle flooding detection module can increase mAP by 20% compared with the naive detection method, respectively.
참고문헌 (14)
- 최병성, "포항 참사 진짜 원인, 처참한 현장에 남은 결정적증거", 오마이뉴스, ~ (2022)
- 정재원, "LSTM 기반 딥러닝 기법을 이용한 섬진강 구례교 지점의 홍수위 예측", 한국방재학회논문집, 21(3), pp.193~201 (2021), 한국방재학회[DOI]
- 정성호, "딥러닝 기반 LSTM 모형을 이용한 감조하천 수위 예측", 한국수자원학회 논문집, 51(12), pp.1207~1216 (2018), 한국수자원학회[DOI]
- Y. Bengio, "The Problem of Learning Long-term Dependencies in Recurrent Networks", IEEE International Conference on Neural Networks, 3, pp.1183~1188 (1993)
- S. Hochreiter, "Long Short-Term Memory", Neural Computation, 9(8), pp.1735~1780 (1997)
- 정종진, "LSTM을 이용한 주가예측 모델의 학습방법에 따른 성능분석", 디지털융복합연구, 18(11), pp.259~266 (2020), 한국디지털정책학회[DOI]
- 전병기, "건물 예측 제어용 LSTM 기반 일사 예측 모델", 한국태양에너지학회 논문집, 39(5), pp.41~52 (2019), 한국태양에너지학회[DOI]
- 김은희, "LSTM 모델 기반 주행 모드 인식을 통한 자율 주행에 관한 연구", 한국ITS학회 논문지, 16(4), pp.153~163 (2017), 한국ITS학회[DOI]
- 최지호, "비디오 영상에서 2차원 자세 추정과 LSTM 기반의 행동 패턴 예측 알고리즘", 대한임베디드공학회논문지, 17(4), pp.191~197 (2022), 대한임베디드공학회[DOI]
- A. Dai, "Effects of Clouds Soil Moisture, Precipitation, and Water Vapor on Diurnal Temperature Range", Journal of Climate, 12(8), pp.2451~2473 (1999)
- 민승기, "ERRATUM Erratum to: Changes in Weather and Climate Extremes over Korea and Possible Causes: A Review", Asia-Pacific Journal of Atmospheric Sciences, 51(3), pp.291~291 (2015), 한국기상학회
- ‘카 히스토리’ 통해침수차인지 반드시 확인해야, "소비자경제", 2022
- "YOLOv5"[URL]
- "한국수자원공사_우량수위 관측정보", 공공데이터포털[URL]