Deep Learning-Based Fatigue Strength Prediction for Ferrous Alloy

Processes, Journal Year: 2024, Volume and Issue: 12(10), P. 2214 - 2214

Published: Oct. 11, 2024

As industrial development drives the increasing demand for steel, accurate estimation of material’s fatigue strength has become crucial. Fatigue strength, a critical mechanical property is primary factor in component failure within engineering applications. Traditional testing both costly and time-consuming, can lead to severe consequences. Therefore, need develop faster more efficient methods predicting evident. In this paper, dataset was established, incorporating data on material element composition, physical properties, performance parameters that influence strength. A machine learning regression model then applied facilitate rapid prediction ferrous alloys. Twenty characteristic parameters, selected their practical relevance applications, were used as input variables, with output. Multiple algorithms trained dataset, deep employed The models evaluated using metrics such MAE, RMSE, R2, MAPE. results demonstrated superiority proposed effectiveness methodologies.

Language: Английский

---

Prediction of Equipment Remaining Useful Life Based on Graph Learning and Spatiotemporal Knowledge Graph

Quality and Reliability Engineering International, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 3, 2025

ABSTRACT As equipment structures and functionalities become more complex, ensuring safety reliability has increasingly critical. Hence, accurately predicting the remaining useful life (RUL) of gained significant importance. Recent advances in graph learning have contributed significantly to RUL prediction by leveraging monitoring signals extract temporal features build predictive models. However, a key challenge persists: structured prior knowledge that describes spatiotemporal correlations between data structure is often lacking, relational priors are not effectively incorporated modeling process. To address these challenges, this paper proposes (STKG) method for equipment, combined with graph‐based feature algorithm prediction. The main contributions work as follows: (1) STKG models hierarchical relationships among sensor signals, state transitions across both spatial dimensions; (2) A attention convolution‐pooling network, incorporating priors, proposed from at different time points, constructing aggregation mappings; (3) informer network employed capture decay patterns, generating cross‐time representations validated on public dataset, demonstrating superior performance compared existing

Language: Английский

---

Evolution of the Fatigue Failure Prediction Process from Experiment to Artificial Intelligence: A Review

Materials, Journal Year: 2025, Volume and Issue: 18(5), P. 1153 - 1153

Published: March 4, 2025

An analysis of the time evolution fatigue break prediction shows increasingly shorter developmental stages. The experimental period was longest; combination more powerful mathematical methods led to a leap in and shortening implementation time. All rupture have proven limitations due multitude influencing factors insufficient number practical considered. Recently, attempts been made increase accuracy by combining based on physical mechanisms failure process with data-driven assisted artificial intelligence. We attempt present this herein. There are several review suitable for analyzing subject: systematic, semi-systematic, integrative. From these, semi-systematic integrative chosen precisely because two complement each other.

Language: Английский

---

Digital Twin Model and Platform Based on a Dual System for Control Rod Drive Mechanism Safety

Reliability Engineering & System Safety, Journal Year: 2025, Volume and Issue: unknown, P. 111075 - 111075

Published: April 1, 2025

Language: Английский

---

Data-Driven Feature Extraction-Transformer: A Hybrid Fault Diagnosis Scheme Utilizing Acoustic Emission Signals

Processes, Journal Year: 2024, Volume and Issue: 12(10), P. 2094 - 2094

Published: Sept. 26, 2024

This paper introduces a novel network, DDFE-Transformer (Data-Driven Feature Extraction-Transformer), for fault diagnosis using acoustic emission signals. The network integrates two primary modules: the DDFE module, focusing on noise reduction and feature enhancement, Transformer module. module employs techniques: Wavelet Kernel Network (WKN) Convolutional Block Attention Module (CBAM) enhancement. wavelet function in WKN reduces noise, while attention mechanism CBAM enhances features. then processes vectors sends results to softmax layer classification. To validate proposed method’s efficacy, experiments were conducted datasets from NASA Ames Research Center University of California, Berkeley. compared four key metrics obtained through confusion matrix analysis. Experimental show that method performs excellently signals, achieving high average accuracy 99.84% outperforming several baseline models, such as CNN, CNN-LSTM, CNN-GRU, VGG19, ZFNet. best-performing model, only achieved an 88.61%. Additionally, findings suggest integrating enhancement single framework significantly improves network’s classification robustness when analyzing

Language: Английский

---

Wear Reliability Assessment of Piston Ring and Cylinder Liner Parts Based on Probability Mapping

Published: Jan. 1, 2024

Language: Английский

---

A random effects Wiener process model with acceleration factors for remaining useful life prediction

Journal of the Brazilian Society of Mechanical Sciences and Engineering, Journal Year: 2024, Volume and Issue: 46(12)

Published: Nov. 27, 2024

Language: Английский

---