Sensor Fault Classification Based on Support Vector Machine and Statistical Time-Domain Features

Jan, Sana Ullah; Lee, Young-Doo; Shin, Jungpil; Koo, Insoo

doi:10.1109/access.2017.2705644

Cited by 209 publications

(92 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The machine learning models utilized in this study are KNN [13], Logistic Regression [14], Decision Tree [15], Random Forest [16], Gradient Boosting Regression Tree [16], Support Vector Machine (SVM) [17], and MLP [18]. KNN classifies data based on the decision boundary, and Logistic Regression classifies data using linear function as shown in Equation (4).…”

Section: Machine Learning Models and Performance Evaluationmentioning

confidence: 99%

Improved Practical Vulnerability Analysis of Mouse Data According to Offensive Security based on Machine Learning in Image-Based User Authentication

Lee

2020

Entropy

View full text Add to dashboard Cite

The objective of this study was to verify the feasibility of mouse data exposure by deriving features to improve the accuracy of a mouse data attack technique using machine learning models. To improve the accuracy, the feature appearing between the mouse coordinates input from the user was analyzed, which is defined as a feature for machine learning models to derive a method of improving the accuracy. As a result, we found a feature where the distance between the coordinates is concentrated in a specific range. We verified that the mouse data is apt to being stolen more accurately when the distance is used as a feature. An accuracy of over 99% was achieved, which means that the proposed method almost completely classifies the mouse data input from the user and the mouse data generated by the defender.

show abstract

Section: Machine Learning Models and Performance Evaluationmentioning

confidence: 99%

Improved Practical Vulnerability Analysis of Mouse Data According to Offensive Security based on Machine Learning in Image-Based User Authentication

Lee

2020

Entropy

View full text Add to dashboard Cite

show abstract

“…The unit of each value is µg/m 3 . When the diameter of PM is 2.5µg/m 3 it called P M 2.5 , and when it between 2.5µg/m 3 and 10µg/m 3 , called P M 10 .…”

Section: Introductionmentioning

confidence: 99%

Anomaly Detection in Particulate Matter Sensor Using Hypothesis Pruning Generative Adversarial Network

Park

Kim

2020

Preprint

View full text Add to dashboard Cite

World Health Organization (WHO) provides the guideline for managing the Particulate Matter (PM) level because when the PM level is higher, it threats the human health. For managing PM level, the procedure for measuring PM value is needed firstly. We use Tapered Element Oscillating Microbalance (TEOM)-based PM measuring sensors because it shows higher cost-effectiveness than Beta Attenuation Monitor (BAM)-based sensor. However, TEOM-based sensor has higher probability of malfunctioning than BAM-based sensor. In this paper, we call the overall malfunction as an anomaly, and we aim to detect anomalies for the maintenance of PM measuring sensors. We propose a novel architecture for solving the above aim that named as Hypothesis Pruning Generative Adversarial Network (HP-GAN). We experimentally compare the several anomaly detection architectures to certify ours performing better.

show abstract

“…The dataset will train a support vector machine (SVM) classifier. Recently, SVM as well as other machine learning algorithms has been drawing attention in the field of fault diagnosis and health monitoring . Using MPM in conjunction with SVM not only classifies an induction motor as healthy or faulty, but also identifies the severity of the fault through the identification of the number of broken bars or the number of broken connectors in the rotor.…”

Section: Introductionmentioning

confidence: 99%

“…Recently, SVM as well as other machine learning algorithms has been drawing attention in the field of fault diagnosis and health monitoring. [19][20][21][22][23][24] Using MPM in conjunction with SVM not only classifies an induction motor as healthy or faulty, but also identifies the severity of the fault through the identification of the number of broken bars or the number of broken connectors in the rotor. The interest of the proposed approach hence lies in presenting an effective method for monitoring motor fault severities that would possibly enable predicting the time of complete failure of a motor, thus preventing unexpected interruptions of industrial processes.…”

mentioning

confidence: 99%

Rotor fault diagnosis in induction motors by the matrix pencil method and support vector machine

Chahine

2018

Int Trans Electr Energ Syst

View full text Add to dashboard Cite

Summary This paper proposes a high‐resolution motor‐current‐signature‐analysis method and a supervised learning algorithm for the detection and classification of broken rotor bars and broken end‐ring connectors in 3‐phase induction motors. Signature analysis relies on the matrix pencil method (MPM), a well‐known model‐based parameter estimation technique, to extract representative features or signatures from stator current signals. Extracted feature vectors are subsequently used to train off‐line a support vector machine classifier. Once trained, the classifier is tested on a benchmark dataset of simulated stator current signals representing healthy and faulty rotors with the aim of classifying the underlying motor condition. The obtained results validate the matrix pencil method as a feature extraction method and show that the trained classifier achieves a 100% success rate in identifying the number of broken bars and connectors. Moreover, the advantages of the matrix pencil method over fast Fourier transform are demonstrated using experimental data.

show abstract

Sensor Fault Classification Based on Support Vector Machine and Statistical Time-Domain Features

Cited by 209 publications

References 25 publications

Improved Practical Vulnerability Analysis of Mouse Data According to Offensive Security based on Machine Learning in Image-Based User Authentication

Improved Practical Vulnerability Analysis of Mouse Data According to Offensive Security based on Machine Learning in Image-Based User Authentication

Anomaly Detection in Particulate Matter Sensor Using Hypothesis Pruning Generative Adversarial Network

Rotor fault diagnosis in induction motors by the matrix pencil method and support vector machine

Contact Info

Product

Resources

About