Analysis of Uneven Settlement of Long-Span Bridge Foundations Based on SBAS-InSAR

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(2), P. 248 - 248

Published: Jan. 11, 2025

Bridge foundation settlement monitoring is crucial for infrastructure safety management, as uneven can lead to stress redistribution, structural damage, and potentially catastrophic collapse. While traditional contact sensors provide reliable measurements, their deployment labor-intensive costly, especially long-span bridges. Current remote sensing methods have not been thoroughly evaluated capability detect analyze complex patterns in challenging environments with multiple influencing factors. Here, we applied Small Baseline Subsets Synthetic Aperture Radar Interferometry (SBAS-InSAR) technology monitor of a bridge. Our analysis revealed distinct deformation patterns: uplift the north bank approach bridge left-side main (maximum rate: 36.97 mm/year), concurrent subsidence right-side south 35.59 mm/year). We then investigated relationship between these various environmental factors, including geological conditions, Sediment Transport Index (STI), Topographic Wetness (TWI), precipitation, temperature. The observed were attributed combined effects stratigraphic heterogeneity, dynamic hydrological seasonal climate variations. These findings demonstrate that SBAS-InSAR effectively capture processes, offering cost-effective alternative methods. This advancement could enable more widespread frequent assessment stability, ultimately improving management.

Language: Английский

---

Enhancing VBAC Prediction with AI-Powered Temporal Dynamics: Integrating Decision Support into a Shared Decision-Making Platform for Intrapartum Care

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: March 18, 2025

Background: Taiwan has a high caesarean section (CS) rate, ranging from 37% to 38%. Vaginal Birth After Cesarean (VBAC) offers potential solution reduce these rates. However, the prevalence of VBAC remains below 0.5%, primarily due concerns about risks adverse maternal and perinatal outcomes. Objectives: This study aims evaluate predictive performance various machine learning (ML) models using pregnancy, labor, intervention-related features predict success support real-time clinical decision-making during labor. Study Design: This retrospective exploratory analyzed data collected hospital in northern between January 2019 May 2023. Statistical methods included demographic comparisons, feature evaluations, model metrics such as accuracy, precision, recall, F1-score, area under curve (AUC). SHapley Additive exPlanations (SHAP) analysis was used interpret importance labor progression. Results: A comparison Failure group (n=22) Success (n=33), totaling 55 records 36 pregnant women, revealed significant differences parity, spontaneous rupture membranes, cervical dilation (at both 0 cm 10 cm), progression slope. Models incorporating high-impact demonstrated superior compared those utilizing only pregnancy-related data. The Random Forest achieved an accuracy 94% AUC 0.96 predicting SHAP further identified key predictors across different stages including (body mass index, prior vaginal birth, age), static (spontaneous time since rupture), dynamic (cervical slope). Conclusion: integrative approach, which combines expertise with analytics, provides clinicians valuable tool for evaluation decision-making. By offering more accurate predictions progression, particularly context VBAC, this approach significantly improve neonatal outcomes

Language: Английский

---

A high performance hybrid LSTM CNN secure architecture for IoT environments using deep learning

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 20, 2025

Abstract The growing use of IoT has brought enormous safety issues that constantly demand stronger hide from increasing risks intrusions. This paper proposes an Advanced LSTM-CNN Secure Framework to optimize real-time intrusion detection in the context. It adds LSTM layers, which allow for temporal dependencies be learned, and CNN layers decompose spatial features makes this model efficient identifying threats. is important note used BoT-IoT dataset involves various cyber attack typologies like DDoS, botnet, reconnaissance, data exfiltration. These outcomes present proposed 99.87% accuracy, 99.89% precision, 99.85% recall with a low false positive rate 0.13% exceeds CNN, RNN, Standard LSTM, BiLSTM, GRU deep learning models. In addition, 90.2% accuracy conditions adversarial proving robust can practical purposes. Based on feature importance analysis using SHAP, work finds packet size, connection duration, protocol type should possible indicators threat detection. suggest Hybrid could useful improving security devices provide increased reliability alarm rates.

Language: Английский

---

Decoding drinking water flavor: A pioneering and interpretable machine learning approach

Journal of Water Process Engineering, Journal Year: 2025, Volume and Issue: 72, P. 107577 - 107577

Published: March 30, 2025

Language: Английский

---

Prediction of Waste Sludge Production in Municipal Wastewater Treatment Plants by Deep-Learning Algorithms with Antioverfitting Strategies

ACS ES&T Engineering, Journal Year: 2025, Volume and Issue: unknown

Published: April 3, 2025

Language: Английский

---

Development and application of an intelligent nitrogen removal diagnosis and optimization framework for WWTPs: Low-carbon and stable operation

Water Research, Journal Year: 2024, Volume and Issue: 266, P. 122337 - 122337

Published: Aug. 30, 2024

Language: Английский

---

Real-Time Control of A2O Process in Wastewater Treatment Through Fast Deep Reinforcement Learning Based on Data-Driven Simulation Model

Water, Journal Year: 2024, Volume and Issue: 16(24), P. 3710 - 3710

Published: Dec. 22, 2024

Real-time control (RTC) can be applied to optimize the operation of anaerobic–anoxic–oxic (A2O) process in wastewater treatment for energy saving. In recent years, many studies have utilized deep reinforcement learning (DRL) construct a novel AI-based RTC system optimizing A2O process. However, existing DRL methods require use mechanistic models training. Therefore they specified data construction models, which is often difficult achieve plants (WWTPs) where collection facilities are inadequate. Also, training time-consuming because it needs multiple simulations model. To address these issues, this study designs data-driven method. The method first creates simulation model using LSTM and an attention module (LSTM-ATT). This established based on flexible from LSTM-ATT simplified version large language (LLM), has much more powerful ability analyzing time-sequence than usual but with small architecture that avoids overfitting dynamic data. Based this, new framework constructed, leveraging rapid computational capabilities accelerate proposed WWTP Western China. An built used train reduction aeration qualified effluent. For simulation, its mean squared error remains between 0.0039 0.0243, while R-squared values larger 0.996. strategy provided by DQN effectively reduces average DO setpoint 3.956 mg/L 3.884 mg/L, acceptable provides pure WWTPs DRL, effective saving consumption reduction. It also demonstrates purely process, providing decision-support management.

Language: Английский

---

Comparative Analysis of Machine Learning Models and Explainable Artificial Intelligence for Predicting Wastewater Treatment Plant Variables

Advances in Environmental and Engineering Research, Journal Year: 2024, Volume and Issue: 05(04), P. 1 - 23

Published: Oct. 17, 2024

Increasing urban wastewater and rigorous discharge regulations pose significant challenges for treatment plants (WWTP) to meet regulatory compliance while minimizing operational costs. This study explores the application of several machine learning (ML) models specifically, Artificial Neural Networks (ANN), Gradient Boosting Machines (GBM), Random Forests (RF), eXtreme (XGBoost), hybrid RF-GBM in predicting important WWTP variables such as Biochemical Oxygen Demand (BOD), Total Suspended Solids (TSS), Ammonia (NH₃), Phosphorus (P). Several feature selection (FS) methods were employed identify most influential variables. To enhance ML models’ interpretability understand impact on prediction, two widely used explainable artificial intelligence (XAI) methods-Local Interpretable Model-Agnostic Explanations (LIME) SHapley Additive exPlanations (SHAP) investigated study. Results derived from FS XAI compared explore their reliability. The model performance results revealed that ANN, GBM, XGBoost, have great potential variable prediction with low error rates strong correlation coefficients R² value 1 training set 0.98 test set. also common each model’s prediction. is a novel attempt get an overview both LIME SHAP explanations

Language: Английский

---

Exploring the Impact of Climate Variables and Scenario Simulation on Ecosystem Service Value Profits and Losses in China

Land, Journal Year: 2024, Volume and Issue: 13(11), P. 1924 - 1924

Published: Nov. 15, 2024

Ecosystem restoration can yield multiple benefits, and the quantitative accounting of ecosystem service value (ESV) profits losses is significant importance to economic benefits restoration. This study reveals dynamic impacts climate change on ESVs by analyzing effects variables ESV across different periods scenarios. The research findings are as follows: (1) From 1990 2020, extending simulated projections for 2030, China’s exhibits a high distribution pattern in southern regions. In under natural development scenario (NDS), southwestern region shows coexistence low ESVs. Under ecological protection (EPS), increases, whereas urban (UDS), southwest decreases. (2) both NDS UDS, trends continue from 2010 2020. EPS, there increase region. largest contributors loss conversion grassland unused land forest farmland. most spatial differences losses, with an northeastern contrast, other regions show no losses. (3) 2000, Bio13 (the precipitation wettest month) Bio12 (annual precipitation) had positive impact indicating that increased promotes functioning indicates fluctuations temperature factors influencing ESV. Due change, patterns swings now key determinants changes. By carefully studying their driving factors, this serve scientific basis management strategies.

Language: Английский

---