An Enhanced Framework for Assessing Pluvial Flooding Risk with Integrated Dynamic Population Vulnerability at Urban Scale

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(4), P. 654 - 654

Published: Feb. 14, 2025

Under the combined influence of climate change, accelerated urbanization, and inadequate urban flood defense standards, pluvial flooding has become an increasingly severe issue. This not only poses significant challenges to social stability economic development but also makes accurate risk assessment crucial for improving control drainage capabilities. study uses Jinan, a typical foothill plain city in Shandong Province, as case compare performance differential evolution (DE), genetic algorithm (GA), particle swarm optimization (PSO) calibrating SWMM. By constructing hydrological–hydrodynamic coupled model using SWMM LISFLOOD-FP, this evaluates capacity pipe network surface inundation characteristics under both historical design rainfall scenarios. An agent-based (ABM) is developed analyze dynamic risks vulnerabilities population building agents different scenarios, capturing macroscopic emergent patterns from individual behavior rules analyzing them time space dimensions. Additionally, multi-source remote sensing data, vulnerability, hazard processes, quantitative analysis conducted based on cloud models. The results demonstrated following: (1) PSO performed best area, with Nash–Sutcliffe efficiency (NSE) values ranging 0.93 0.69. (2) Drainage system was low, over 90% exceeding scenarios return periods 1 100 years. (3) vulnerability people buildings increased higher intensity duration. Most affected individuals were located roads. In Event 6, 11.41% at after 1440 min; 20-year event, 26.69% 180 min. (4) Key features influencing included DEM, PND, NDVI, slope. High-risk areas area expanded 36.54% 30 min 38.05%

Language: Английский

---

Convolutional Neural Networks for forecasting flood process in Internet-of-Things enabled smart city

Computer Networks, Journal Year: 2020, Volume and Issue: 186, P. 107744 - 107744

Published: Dec. 17, 2020

Language: Английский

---

Hybrid Models Incorporating Bivariate Statistics and Machine Learning Methods for Flash Flood Susceptibility Assessment Based on Remote Sensing Datasets

Remote Sensing, Journal Year: 2021, Volume and Issue: 13(23), P. 4945 - 4945

Published: Dec. 5, 2021

Flash floods are considered to be one of the most destructive natural hazards, and they difficult accurately model predict. In this study, three hybrid models were proposed, evaluated, used for flood susceptibility prediction in Dadu River Basin. These integrate a bivariate statistical method fuzzy membership value (FMV) machine learning methods support vector (SVM), classification regression trees (CART), convolutional neural network (CNN). Firstly, geospatial database was prepared comprising nine conditioning factors, 485 locations, non-flood locations. Then, train test models. Subsequently, receiver operating characteristic (ROC) curve, seed cell area index (SCAI), accuracy evaluate performances The results reveal following: (1) ROC curve highlights fact that CNN-FMV had best fitting performance, under (AUC) values success rate 0.935 0.912, respectively. (2) Based on performance evaluation methods, all better capabilities than their respective single Compared with models, AUC SVM-FMV, CART-FMV, 0.032, 0.005, 0.055 higher; SCAI 0.05, 0.03, 0.02 lower; accuracies 4.48%, 1.38%, 5.86% higher, (3) indices, between 13.21% 22.03% study characterized by high very susceptibilities. proposed especially CNN-FMV, have potential application assessment specific areas future studies.

Language: Английский

---

Modulation Recognition of Underwater Acoustic Signals Using Deep Hybrid Neural Networks

IEEE Transactions on Wireless Communications, Journal Year: 2022, Volume and Issue: 21(8), P. 5977 - 5988

Published: Jan. 27, 2022

It is a huge challenge for the receiver to correctly identify modulation types due complex underwater channel environment and severe noise interference. Additionally, real-time communications have strict requirements in terms of time. In order solve this well-known issue, work, we combine automatic feature extraction learning ability recurrent neural network (RNN) convolutional (CNN) designing recognition model acoustic signals. The proposed based on deep hybrid networks called (R&CNN). As compared with traditional techniques, method achieves higher accuracy without manual extraction. experimental results show that validation R&CNN's Trestle data set 98.21%. Similarly, South China Sea 99.38%. average time 7.164ms. conventional methods, R&CNN not only has accuracy, but also greatly reduces

Language: Английский

---

Assessing the urban road waterlogging risk to propose relative mitigation measures

The Science of The Total Environment, Journal Year: 2022, Volume and Issue: 849, P. 157691 - 157691

Published: July 27, 2022

Language: Английский

---