What Geotechnical Engineers Want to Know about Reliability

ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems Part A Civil Engineering, Journal Year: 2023, Volume and Issue: 9(2)

Published: April 5, 2023

The purpose of this paper is to address the "what," "why," and "how" questions posed by engineers who are not familiar with geotechnical reliability have kept abreast recent rapid developments in field. Geotechnical can be broadly defined as a methodology that enhances decision making at different life-cycle stages covering design, construction, operation maintenance, retrofit, decommission/reuse exploiting richer characterization data using probabilistic models. Besides engineered systems, it also covers risk assessment management geohazards such earthquakes landslides. Various application areas related design construction systems put context form an uncertainty-informed Burland Triangle. Among these areas, estimation characteristic value, load resistance factor (LRFD), calibration for simplified reliability-based (RBD), first-order second-moment (FOSM) analysis do need in-depth knowledge/expertise significant amount information exists support applications practice. This argues their adoption because they will nudge mindset shift more responsive data. Data infrastructure now considered important physical infrastructure. concern there insufficient has been largely comprehensively resolved advances Bayesian machine learning methods deal MUSIC-3X (Multivariate, Uncertain Unique, Sparse, Incomplete, potentially Corrupted "3X" denoting 3D spatial variability) site directly. Sparsity (insufficient data) only one out six attributes real-world set. pictured step toward digital transformation engaging complex new challenges climate change resilience engineering. urges engineering profession lay aside its on quantity, quality, and/or other "ugly" offer our opportunity speak itself. There prima facie evidence warrant thorough exploration data-centric geotechnics.

Language: Английский

---

Machine Learning: Models, Challenges, and Research Directions

Future Internet, Journal Year: 2023, Volume and Issue: 15(10), P. 332 - 332

Published: Oct. 9, 2023

Machine learning techniques have emerged as a transformative force, revolutionizing various application domains, particularly cybersecurity. The development of optimal machine applications requires the integration multiple processes, such data pre-processing, model selection, and parameter optimization. While existing surveys shed light on these techniques, they mainly focused specific domains. A notable gap that exists in current studies is lack comprehensive overview architecture its essential phases cybersecurity field. To address this gap, survey provides holistic review learning, covering applicable to any domain. Models are classified into four categories: supervised, semi-supervised, unsupervised, reinforcement learning. Each categories their models described. In addition, discusses progress related pre-processing hyperparameter tuning techniques. Moreover, identifies reviews research gaps key challenges field faces. By analyzing gaps, we propose some promising directions for future. Ultimately, aims serve valuable resource researchers interested about providing them with insights foster innovation across diverse

Language: Английский

---

Investigation of disc cutter wear during shield tunnelling in weathered granite: A case study

Tunnelling and Underground Space Technology, Journal Year: 2023, Volume and Issue: 140, P. 105323 - 105323

Published: July 20, 2023

Language: Английский

---

A spatiotemporal deep learning method for excavation-induced wall deflections

Journal of Rock Mechanics and Geotechnical Engineering, Journal Year: 2024, Volume and Issue: 16(8), P. 3327 - 3338

Published: Jan. 29, 2024

Data-driven approaches such as neural networks are increasingly used for deep excavations due to the growing amount of available monitoring data in practical projects. However, most network models only use from a single point and neglect spatial relationships between multiple points. Besides, lack flexibility providing predictions days after activity. This study proposes sequence-to-sequence (seq2seq) two-dimensional (2D) convolutional long short-term memory (S2SCL2D) predicting spatiotemporal wall deflections induced by excavations. The model utilizes all points on entire extracts features combining 2D layers (LSTM) layers. S2SCL2D achieves long-term prediction through recursive seq2seq structure. excavation depth, which has significant impact deflections, is also considered using feature fusion method. An project Hangzhou, China, illustrate proposed model. results demonstrate that superior accuracy robustness than LSTM S2SCL1D (one-dimensional) models. demonstrates strong generalizability when applied an adjacent excavation. Based results, practitioners can plan allocate resources advance address potential engineering issues.

Language: Английский

---

Real-time classification model for tunnel surrounding rocks based on high-resolution neural network and structure–optimizer hyperparameter optimization

Computers and Geotechnics, Journal Year: 2024, Volume and Issue: 168, P. 106155 - 106155

Published: Feb. 20, 2024

Language: Английский

---

Future-proofing geotechnics workflows: accelerating problem-solving with large language models

Georisk Assessment and Management of Risk for Engineered Systems and Geohazards, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 18

Published: July 25, 2024

The integration of Large Language Models (LLMs), such as ChatGPT, into the workflows geotechnical engineering has a high potential to transform how discipline approaches problem-solving and decision-making. This paper investigates practical uses LLMs in addressing challenges based on opinions from diverse group, including students, researchers, professionals academia, industry, government sectors gathered workshop dedicated this study. After introducing key concepts LLMs, we present preliminary LLM solutions for four distinct problems illustrative examples. In addition basic text generation ability, each problem is designed cover different extended functionalities that cannot be achieved by conventional machine learning tools, multimodal modelling under unified framework, programming knowledge extraction, embedding. We also address potentials implementing particularly achieving precision accuracy specialised tasks, underscore need expert oversight. findings demonstrate effectiveness enhancing efficiency, data processing, decision-making engineering, suggesting paradigm shift towards more integrated, data-driven field.

Language: Английский

---

Applying Optimized Machine Learning Models for Predicting Unconfined Compressive Strength in Fine-Grained Soil

Transportation Infrastructure Geotechnology, Journal Year: 2024, Volume and Issue: 11(4), P. 2235 - 2269

Published: Feb. 8, 2024

Language: Английский

---

Future of Machine Learning in Geotechnics (FOMLIG), 5–6 Dec 2023, Okayama, Japan

Georisk Assessment and Management of Risk for Engineered Systems and Geohazards, Journal Year: 2024, Volume and Issue: 18(1), P. 288 - 303

Published: Jan. 2, 2024

This report presents the key talking points in First Workshop on Future of Machine Learning Geotechnics (FOMLIG), that include data infrastructure, geotechnical context, computational cost, and human judgment. On first point, it was argued further growth sharing needs stronger demonstration value to practice protection privacy. second significant progress has been made addressing site specificity (site recognition challenge). third is costly interpret monitoring context machine learning guided observational method (MLOM) because 3D domain influencing structure large, real-time dataset very large its attributes are complicated, fusion remains challenging, computation speed must support decision making. Real-time learning-based clearly not useful if providing engineer with sufficient lead time adjust construction process. fourth capability generative AIs such as ChatGPT act an intelligent companion making exciting. The role judgment human-machine teaming unclear, but for be effective, a deliberate approach needed build trust between AI/robot partner.

Language: Английский

---

A novel tool for probabilistic modeling of liquefaction behavior in alluvial soil

Georisk Assessment and Management of Risk for Engineered Systems and Geohazards, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 24

Published: Aug. 27, 2024

This research introduces and validates advanced machine learning models designed to predict the probability of liquefaction failure (pf) in alluvial soil deposits. Three optimisation algorithms namely Northern Goshawk Optimization (NGO), Jellyfish Search Optimizer (JSO), Horse Herd Algorithm (HHO) coupled with Adaptive Neuro Fuzzy inference system (AFS) has been employed present research. Among tested, AFS-HHO model exhibited superior predictive ability, R2 = 0.93 RMSE 0.06 during stage, 0.89 0.07 testing stage. highlights model's efficiency accurately predicting pf using only corrected SPT-N value i.e. (N1)60 cyclic stress ratio (CSR). The study also emphasises importance influencing probabilistic assessment failure, proposes a novel chart as reliable tool for estimating Considering overall analysis, proposed offer geotechnical engineers estimate thereby holding substantial implications field evaluation studies.

Language: Английский

---

Assessing the strength of deep-sea surface ultrasoft sediments with T-bar penetration: A machine learning approach

Engineering Geology, Journal Year: 2024, Volume and Issue: 338, P. 107632 - 107632

Published: July 7, 2024

Language: Английский

---