関連する概念動画
Convolution Properties II
583
The important convolution properties include width, area, differentiation, and integration properties.
The width property indicates that if the durations of input signals are T1 and T2, then the width of the output response equals the sum of both durations, irrespective of the shapes of the two functions. For instance, convolving two rectangular pulses with durations of 2 seconds and 1 second results in a function with a width of 3 seconds.
The area property asserts that the area under the...
The width property indicates that if the durations of input signals are T1 and T2, then the width of the output response equals the sum of both durations, irrespective of the shapes of the two functions. For instance, convolving two rectangular pulses with durations of 2 seconds and 1 second results in a function with a width of 3 seconds.
The area property asserts that the area under the...
583
Fatigue
831
Fatigue occurs when materials rupture under repeated or fluctuating loads, even at stress levels far below their static breaking strength. It typically results in brittle failure, even for ductile materials. It is a critical consideration in designing machines and structural components subjected to repetitive or varying loads. The nature of these loadings can range from fluctuating loads like unbalanced pump impellers causing vibrations to repeatedly bending a thin steel rod wire back and forth...
831
Emission Spectra
76.2K
When solids, liquids, or condensed gases are heated sufficiently, they radiate some of the excess energy as light. Photons produced in this manner have a range of energies, and thereby produce a continuous spectrum in which an unbroken series of wavelengths is present.
76.2K
Convolution Properties I
584
Convolution computations can be simplified by utilizing their inherent properties.
The commutative property reveals that the input and the impulse response of an LTI (Linear Time-Invariant) system can be interchanged without affecting the output:
The commutative property reveals that the input and the impulse response of an LTI (Linear Time-Invariant) system can be interchanged without affecting the output:
584
Protein Networks
4.5K
An organism can have thousands of different proteins, and these proteins must cooperate to ensure the health of an organism. Proteins bind to other proteins and form complexes to carry out their functions. Many proteins interact with multiple other proteins creating a complex network of protein interactions.
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
4.5K
Types of Non-structural Cracks in Concrete
498
Non-structural cracks are primarily of three types: plastic, early-age thermal, and drying shrinkage cracks. Plastic cracks are further classified into plastic shrinkage cracks and plastic settlement cracks.
Plastic shrinkage cracks typically form within hours after the concrete is poured. The concrete's surface dries faster than the bottom, creating tensile stress that the still-plastic concrete cannot withstand, leading to diagonal or randomly patterned cracks on the concrete surface.
Plastic shrinkage cracks typically form within hours after the concrete is poured. The concrete's surface dries faster than the bottom, creating tensile stress that the still-plastic concrete cannot withstand, leading to diagonal or randomly patterned cracks on the concrete surface.
498
こちらも読む
関連記事
共著者、ジャーナル、引用グラフによってこの研究に関連する記事。
並び替え
Same author
New Method of Impact Localization on Plate-like Structures Using Deep Learning and Wavelet Transform.
Sensors (Basel, Switzerland)·2025
Same author
Exploring Piezoelectric Actuation towards Its Applications in Laser Powder Bed Fusion Additive Manufacturing.
Sensors (Basel, Switzerland)·2024
Same author
Piezoelectric Transducer-Based Structural Health Monitoring.
Sensors (Basel, Switzerland)·2024
音響放出技術に基づく畳み込みニューラルネットワークを用いた疲労き裂長さ推定
Asaad Migot1,2, Ahmed Saaudi3,2, Roshan Joseph4
1Department of Petroleum and Gas Engineering, College of Engineering, University of Thi-Qar, Nasiriyah 64001, Iraq.
Sensors (Basel, Switzerland)
|January 28, 2026
まとめ
この研究では、深層学習と音響放出(AE)信号を使用して、金属板の疲労き裂長さを推定します。CNNの転移学習は、構造的健全性監視を強化する99%の精度を達成しました。
科学分野:
- 材料科学
- 機械工学
- 人工知能
背景:
- 疲労き裂の伝播は、工学構造における重要な破壊メカニズムです。
- 効果的な監視は、タイムリーなメンテナンスと壊滅的な故障の防止に不可欠です。
- 音響放出(AE)信号は、き裂の成長を検出し分析するための有望な非破壊検査方法を提供します。
研究 の 目的:
- AE信号を使用した金属板の疲労破壊長さ推定のための深層学習フレームワークを開発すること。
- AEデータの分析における畳み込みニューラルネットワーク(CNN)および転移学習の効果を調査すること。
- データ駆動型アプローチによる構造的健全性監視能力の強化。
主な方法:
- AE波形は、Choi-Williams分布を使用して時間周波数画像に変換されました。
- CNNベースのモデルが特徴抽出に使用され、K平均法クラスタリングが疲労長さを分類しました。
- 転移学習モデル(ResNet50V2、VGG16)は、破壊長さ分類のためのカスタムCNNと比較されました。
主要な成果:
- AEデータセットは正常にクラスタリングされ、PCAを使用してデータポイントの近接性が視覚化されました。
- CNNモデルは、破壊長さを3つの異なる範囲に正確に分類しました。
- 転移学習モデルは、カスタムCNN(約93%)と比較して、大幅に高い精度(約99%)を達成しました。
結論:
- 深層学習、特に転移学習は、AEデータの分析に非常に効果的です。
- CNNは、疲労き裂監視のためのAE信号の理解において強力な能力を示しています。
- 提案されたフレームワークは、金属部品のデータ駆動型構造的健全性監視を進歩させます。


