检索结果-内蒙古大学图书馆

Method of Classifying Motor Failure Conditions Using an Auto-Encoder-Based Ensemble Technique

Transactions of the Korean Institute of Electrical Engineers 2025年第2期74卷 273-279页

作者： Choi, Eui-Jin Han, Ji-Hoon Song, Seung-Min Park, Jong-Hoon Hong, Sun-Ki Dept. of System and Control Engineering Hoseo University Korea Republic of Dept. of Information Control Engineering Hoseo University Korea Republic of

In the field, unsupervised learning-based fault diagnosis is necessary due to the lack of fault data. However, conventional auto-encoder models face challenges in fault classification. To address this issue, this study proposes an unsupervised learning-based auto-encoder model enhanced with an ensemble approach. The proposed model performs fault classification using gear, bearing, and eccentricity fault data, and employs the Layer-wise Relevance Propagation (LRP) technique to improve fault classification. Through the LRP technique, the key frequency bands relevant to each fault are identified and incorporated into the learning process, allowing for frequency band-based training. As a result, the proposed auto-encoder model improves classification performance and effectively distinguishes between different types of faults. © The Korean Institute of Electrical Engineers.

关键词： Unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

Study on Improving Hysteresis Motor Efficiency by Over-Excitation Based on Finite Element Analysis

引用

Transactions of the Korean Institute of Electrical Engineers 2025年第1期74卷 49-55页

作者： Park, Jong-Hoon Han, Ji-Hoon Choi, Eui-Jin Hong, Sun-Ki Dept. of Information Control Engineering Hoseo University Korea Republic of Dept. of System and Control Engineering Hoseo University Korea Republic of

A hysteresis motor is a motor characterized by an increase in the efficiency of the output to input ratio during an over-excitation in which a higher voltage is applied at startup than when the rated voltage is applied, and the voltage is reduced over a period of time after entering synchronous speed. However, due to the specificity of the rotor material, finite element analysis of hysteresis motors is difficult, so the design was conducted by analyzing only the rated voltage application condition through an equivalent circuit. As a result, the improved efficiency and torque that can be achieved by over-excitation were not verified, which was confirmed by experiments after the design was completed. However, recently, finite element analysis simulation that can express the hysteresis loop behavior has become possible, so finite element analysis of hysteresis motors has become possible, and the improved efficiency and torque that can be obtained through over-excitation can be confirmed. Therefore, this paper compares the output and efficiency of hysteresis motors under over-excitation and rated voltage application, and confirms the necessity and feasibility of designing hysteresis motors considering over-excitation. Copyright © The Korean Institute of Electrical Engineers.

关键词： Equivalent circuits

来源：评论

学校读者我要写书评

暂无评论

Performance Improvement of Transfer Learning-based Motor Fault Diagnosis Using Mix-up Data Considering Data Imbalance

引用

Transactions of the Korean Institute of Electrical Engineers 2023年第2期72卷 255-262页

作者： Choi, Eui-Jin Han, Ji-Hoon Hong, Sun-Ki Dept. of System and Control Engineering Hoseo University Korea Republic of Dept. of Information Control Engineering Hoseo University Korea Republic of

Most of the data obtained for motor fault diagnosis is the data of normal motors. Therefore, the model is trained using normal data collected in the field and fault data collected in experiment. In addtion, transfer learning is used for fast learning convergence. Fault diagnosis performance may be reduced by the number of unbalanced data in each state during transfer learning. In this study, a transfer learning method using mix-up data is proposed to solve this data imbalance. A mix-up is a technique of adding insufficient data to the data with clear decision boundaries, and can be used to balance a large number of normal state data with a small number of fault-state data. The proposed method is compared with GAN, another technique of data augmentation. As a result of the experiment, it was confirmed that the proposed mix-up technique is a method to improve the performance of fault diagnosis. Copyright © The Korean Institute of Electrical Engineers.

关键词： Failure analysis

来源：评论

学校读者我要写书评

暂无评论

Performance Analysis of Transfer-Learning Based Physics-Informed Neural Network for Effective Shape Variation Adaptation with Varying Hyper-parameters

引用

Transactions of the Korean Institute of Electrical Engineers 2023年第10期72卷 1149-1158页

作者： Han, Ji-Hoon Choi, Eui-Jin Hong, Sun-Ki Dept. of System and Control Engineering Hoseo University Korea Republic of Dept. of Information Control Engineering Hoseo University Korea Republic of

One of the most time-consuming parts of designing an electrical device is optimizing the geometry. Optimization involves fine-tuning dimensions such as slot widths and airgap lengths from a base geometry to meet user needs. FEA(Finite Element Analysis) is typically used to verify this performance. However, conventional FEA requires a separate analysis if the analysis geometry is varied even slightly. This causes a very long time consumption, and a transfer learning-based PINN(Physics-Informed Neural Network) is proposed to solve this problem. Since PINNs are at the basic level worldwide, their performance is analyzed according to hyper-parameters such as model parameters (e.g. weights or biases), preprocessing methods, activation functions, and the number of training data to increase their performance. Based on the analyzed hyper-parameters, the performance of the transfer learning-based PINN is verified under the condition of varying the airgap length of the E-I core. © 2023 Korean Institute of Electrical Engineers. All rights reserved.

关键词： Geometry

来源：评论

学校读者我要写书评

暂无评论

Deep Learning-Based Analysis of Neuronal Signal Streams: Estimating the Activity Range of Cultured Neuronal Networks in Consecutive Stimulus Inputs Using CNN 13

Deep Learning-Based Analysis of Neuronal Signal Streams: Est...

引用

Joint 13th International Conference on Soft Computing and Intelligent systems and 25th International Symposium on Advanced Intelligent systems, SCIS and ISIS 2024

作者： Asada, Hiroki Kudoh, Suguru N. Dept. of Human System Interaction Sanda Japan Artificial Intelligence and Mechanical Engineering Course Dept. of Engineering Sanda Japan

ISBN: (纸本)9798350373332

We utilized a Convolutional Neural Network (CNN) -that incorporates a structure designed to integrate local relationship features. The CNN-based deep learning approach enabled us to identify response patterns following stimulus input, using images of neural activity that depict the activity of individual neurons. The feature of CNN-based neural network, integrating local features, considered to contribute to decipher the relationship within time-series data, which are interpreted as streams of signals passing through a single neuron. Utilizing this feature, we attempted to determine the scope of activity that preserves the history of ongoing stimulation within a cultured living neuronal network. We repeatedly subjected the network to the same stimulus, and from the post-stimulus response activity, we extracted spatial combinations of neurons that fired simultaneously within brief 1 ms time windows. We calculated the firing rates for each of these 1 ms intervals in response to 10 different stimuli. By arranging these firing rates into a distinctive configuration that sustains temporal continuity, we generated images that illustrate neuronal activity. We then attempted to discriminate response images obtained at various stimulus intervals, and compared the accuracy of the discrimination to gauge the scope and degree of history retention in sequential activation. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

The Method of for Analyzing EMI Conducted Noise Analysis in Power Conversion Devices using Partial Element Equivalent Circuit Modeling

引用

Transactions of the Korean Institute of Electrical Engineers 2023年第7期72卷 809-819页

作者： Seo, Hyun-Woo Kim, Rae-Young Dept. of Electrical Engineering Hanyang University Korea Republic of Dept. of Energy Power Electronics Control System Lab Hanyang University Korea Republic of

EMI conducted noise is becoming increasingly important in power conversion systems, and various analysis methods are being researched. The proposed method enables accurate EMI conducted noise analysis by extracting partial element equivalent circuits from the 3D structure of power cables and PCBs, and the extracted model can be interpreted in SPICE simulations. The implemented inverter operation mode of Totem-Pole bridge-less PFC in the EMI conducted noise analysis simulation model shows a good correlation between simulation results and experimental results in all CE bands. Furthermore, this method identifies the main causes of EMI noise through CM (Common Mode) and DM (Differential Mode) noise analysis, and proposes a method to effectively attenuate EMI noise through CM coil modeling. Therefore, the EMI conduction noise analysis method using simulation models allows for efficient prediction of EMI conduction noise in advance. Copyright The Korean Institute of Electrical Engineers.

关键词： Poles

来源：评论

学校读者我要写书评

暂无评论

Performance Evaluation of the Continuos Wavelt Transformation Data in Motor Fault Diagnosis through XAI Algorithm

引用

Transactions of the Korean Institute of Electrical Engineers 2022年第1期71卷 225-232页

作者： Han, Ji-Hoon Park, Sang-Uk Hong, Sun-Ki Dept. of Information Control Engineering Hoseo University Korea Republic of Dept. of System and Control Engineering Hoseo University Korea Republic of

The data mainly used for motor fault diagnosis is FFT. However, preprocessing such as continuous discrete wavelet transform is used. When using deep learning algorithms, the performance of the data is evaluated by the model output. However, in order to improve the learning possibilities and preprocessing performance of data, performance evaluation from a model perspective is required. For this purpose, data performance evaluation using the LRP algorithm, one of the XAI techniques, is studied. Initial fault state diagnosis using current data, which is difficult to solve with FFT data, is performed with STFT and CWT data, and performance is evaluated through LRP. Experimental Results STFT and CWT are preprocessing techniques that enable the use of current signals for early fault diagnosis. Among the two preprocessing methods, the use of CWT is more preferable because the flexibility of the preprocessing is increased. © 2022 Korean Institute of Electrical Engineers. All rights reserved.

关键词： Failure analysis

来源：评论

学校读者我要写书评

暂无评论

Research on the Efficient Preparation Method of Metal-enhanced Fluorescent Composite Structure Microarray Chip

Research on the Efficient Preparation Method of Metal-enhanc...

引用

2024 International Conference on Optoelectronic Information and Optical engineering, OIOE 2024

作者： Hu, Hongzhong Kim, Seok-Min Lu, Xun Dept. of Mechanical Engineering Yanbian Univ. Yanji133002 China Dept. of Mechanical System Engineering Chung-Ang Univ. Seoul06974 Korea Republic of

ISBN: (数字)9781510680470

ISBN: (纸本)9781510680463

Microarray chip with micro-nano composite structure consists of an array of micropost structures and nanorods covering the top of the posts. This structure can significantly enhance the signal-to-noise ratio and improve the cross-infection problem to increase the detection sensitivity and accuracy. However, the construction of such structures is still limited to the laboratory research stage, which is mainly realized by UV nanoimprinting and glancing angle deposition methods, and the preparation process is complicated, time-consuming, and inefficient, which greatly restricts the wide application of highly sensitive microarray chips. Therefore, to achieve high accuracy and sensitivity in the large-scale, high-efficiency production of microarray chips, a micropost array structure was fabricated via injection molding. The top of the micropost structure (spotting area) was selectively deposited with Ag nanorods using glancing angle deposition, forming a micro-nano composite structure substrate. Despite a 4 μm discrepancy between the height of the micropost structure on the microarray chip (54 μm) and its designed height (58 μm), the deviation was within acceptable limits. While the uniformity of Cy-3 fluorescence signal across the micro-nano composite structure substrate (with an average signal value of 12519.0 and a coefficient of variation of 4.79%) was slightly lower than that of the UV nanoimprinting chip (with an average signal value of 12495.8 and a coefficient of variation of 1.35%), both exhibited similar fluorescence signal enhancement characteristics. Therefore, this method can serve as a reference for the high-throughput, efficient production of microarray chips. © 2024 SPIE.

关键词： Nanorods

来源：评论

学校读者我要写书评

暂无评论

mechanical Properties and Deformation Behavior of Metal Injection Molded Products with Cellulose Nanofiber

引用

Funtai Oyobi Fummatsu Yakin/Journal of the Japan Society of Powder and Powder Metallurgy 2023年第3期70卷 168-172页

作者： Osada, Toshiko Kobayashi, Satoshi Dept. Mechanical System Engineering Tokyo Metropolitan University 1-1 Minami-Osawa Hachioji192-0397 Japan

Metal Injection Molding (MIM) process is suitable for fabricating small and complex shaped metal parts with high production volume. In this study, in order to improve the MIM process, cellulose nanofibers (CNF) were focused. There are problems such as deformation during debinding and sintering due to the addition of a large amount of binder. On the other hand, the strength of the green compact decreased with decreasing binder contents. In this study, cellulose nanofibers were added with a binder during mixing in order to improve the handling of green compacts, and decrease the deformation during debinding and sintering. The effects of CNF contents on the mechanical properties were investigated. It was clarified that CNF is effective in improving the handleability of the debound compacts and suppressing deformation during sintering. ©2023 Japan Society of Powder and Powder Metallurgy.

关键词： Injection molding

来源：评论

学校读者我要写书评

暂无评论

A Comparison Study of Ball Bearing Fault Diagnosis and Classification Analysis Using XAI Grad-CAM

引用

Transactions of the Korean Institute of Electrical Engineers 2022年第9期71卷 1315-1325页

作者： Kim, Yejin Jeon, Hyeonjick Kim, Young-Keun Dept. of Mechanical and Control Engineering Handong Global University Korea Republic of

Various machine learning and deep learning methods were proposed to monitor and classify the bearing's health state using vibration signals since bearing faults are one of the most causes of failure of rotationary machine. The process of diagnosing bearing faults using machine learning is as follows. First, the features, including the fault characteristic of the vibration signals, are extracted, and these features are selected to reduce the dimension of the features. These features are input into the machine learning classifier to diagnose the system's health. In addition to machine learning methods, CNN, one of the deep learning methods, is widely used. Since the deep learning model extracts features by itself, only the preprocessing process of converting the bearing signals into 2D is needed. The fault classification accuracy of two vibration signal transformation methods as preprocessing methods for the CNN model was compared. This paper compares the bearing fault classification performance of several machine learning commonly used and the CNN model for the lab-made wind turbine machinery testbed. By comparing different feature extraction, feature selection, and classification methods, the most appropriate pipeline is selected for the testbed. Also, grad-cam, an explainable AI(XAI) technique, is applied to interpret the CNN based classification in terms of interested frequency bandwidth. The XAI analysis was verified by designing preprocessing filters based on the grad-cam outputs for enhancing classification performance. © 2022 Korean Institute of Electrical Engineers. All rights reserved.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：