scholarly journals Low-Cost Monitoring and Diagnosis System for Rolling Bearing Faults of the Induction Motor Based on Neural Network Approach

Electronics ◽  
2020 ◽  
Vol 9 (9) ◽  
pp. 1334
Author(s):  
Pawel Ewert ◽  
Czeslaw T. Kowalski ◽  
Teresa Orlowska-Kowalska
Keyword(s):  
Neural Network ◽  
Induction Motor ◽  
Computer System ◽  
Low Cost ◽  
Measurement Data ◽  
Training Data ◽  
Vibration Measurement ◽  
Diagnosis System ◽  
Envelope Analysis ◽  
Monitoring And Diagnosis

In this article, a low-cost computer system for the monitoring and diagnosis of the condition of the induction motor (IM) rolling bearings is demonstrated and tested. The system allows the on-line monitoring of the IM bearings and subsequent fault diagnostics based on analysis of the vibration measurement data. The evaluation of the bearing condition is made by a suitably trained neural network (NN), on the basis of the spectral and envelope analysis of the mechanical vibrations. The system was developed in the LabVIEW environment in such a way that it could be run on any PC. The functionality of the application has been tested on a real object. The study was conducted on a low-power IM equipped with a set of specially prepared bearings so as to model the different damages. In the designed computer system, a selected NN for detecting and identifying the defects of individual components of the induction motor’s bearings was implemented. The training data for NNs were obtained from real experiments. The magnitudes of the characteristic harmonics, obtained from the spectral analysis and the envelope analysis, were used for training and testing the developed neural detectors based on Matlab toolbox. The experimental test results of the developed monitoring and diagnosis system are presented in the article. The evaluation of the system’s ability to detect and identify the defects of individual components of bearings, such as the rolling element, and outer race and inner race, was made. It was also shown that the developed NN-based detectors are insensitive to other motor faults, such as short-circuits of the stator winding, broken rotor bars or motor misalignment.

Non-invasive intelligent monitoring system for fault detection in induction motor based on bio piezoelectric sensor using ANN

2022 ◽  
Author(s):  
Massine GANA ◽  
Hakim ACHOUR ◽  
Kamel BELAID ◽  
Zakia CHELLI ◽  
Mourad LAGHROUCHE ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Induction Motor ◽  
Vibration Analysis ◽  
Low Cost ◽  
Piezoelectric Sensor ◽  
Integrated System ◽  
Easy Access ◽  
Non Invasive ◽  
Artificial Neural

Abstract This paper presents a design of a low-cost integrated system for the preventive detection of unbalance faults in an induction motor. In this regard, two non-invasive measurements have been collected then monitored in real time and transmitted via an ESP32 board. A new bio-flexible piezoelectric sensor developed previously in our laboratory, was used for vibration analysis. Moreover an infrared thermopile was used for non-contact temperature measurement. The data is transmitted via Wi-Fi to a monitoring station that intervenes to detect an anomaly. The diagnosis of the motor condition is realized using an artificial neural network algorithm implemented on the microcontroller. Besides, a Kalman filter is employed to predict the vibrations while eliminating the noise. The combination of vibration analysis, thermal signature analysis and artificial neural network provides a better diagnosis. It ensures efficiency, accuracy, easy access to data and remote control, which significantly reduces human intervention.

Multi-Fidelity Physics-Constrained Neural Network and Its Application in Materials Modeling

2019 ◽  
Vol 141 (12) ◽  
Author(s):  
Dehao Liu ◽  
Yan Wang
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Numerical Solutions ◽  
Low Cost ◽  
Training Data ◽  
Physical Constraints ◽  
Physical Knowledge ◽  
Materials Modeling ◽  
Order Of Magnitude ◽  
The Difference

Abstract Training machine learning tools such as neural networks require the availability of sizable data, which can be difficult for engineering and scientific applications where experiments or simulations are expensive. In this work, a novel multi-fidelity physics-constrained neural network is proposed to reduce the required amount of training data, where physical knowledge is applied to constrain neural networks, and multi-fidelity networks are constructed to improve training efficiency. A low-cost low-fidelity physics-constrained neural network is used as the baseline model, whereas a limited amount of data from a high-fidelity physics-constrained neural network is used to train a second neural network to predict the difference between the two models. The proposed framework is demonstrated with two-dimensional heat transfer, phase transition, and dendritic growth problems, which are fundamental in materials modeling. Physics is described by partial differential equations. With the same set of training data, the prediction error of physics-constrained neural network can be one order of magnitude lower than that of the classical artificial neural network without physical constraints. The accuracy of the prediction is comparable to those from direct numerical solutions of equations.

Multi-Fidelity Physics-Constrained Neural Network and its Application in Materials Modeling

2019 ◽  
Author(s):  
Dehao Liu ◽  
Yan Wang
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Numerical Solutions ◽  
Low Cost ◽  
Training Data ◽  
Physical Constraints ◽  
Physical Knowledge ◽  
Materials Modeling ◽  
Order Of Magnitude ◽  
The Difference

Abstract Training machine learning tools such as neural networks requires the availability of sizable data, which can be difficult for engineering and scientific applications where experiments or simulations are expensive. In this work, a novel multi-fidelity physics-constrained neural network is proposed to reduce the required amount of training data, where physical knowledge is applied to constrain neural networks, and multi-fidelity networks are constructed to improve training efficiency. A low-cost low-fidelity physics-constrained neural network is used as the baseline model, whereas a limited amount of data from a high-fidelity simulation is used to train a second neural network to predict the difference between the two models. The proposed framework is demonstrated with two-dimensional heat transfer and phase transition problems, which are fundamental in materials modeling. Physics is described by partial differential equations. With the same set of training data, the prediction error of physics-constrained neural network can be one order of magnitude lower than that of a classical artificial neural network without physical constraints. The accuracy of the prediction is comparable to those from direct numerical solutions of equations.

scholarly journals Fault Diagnosis System of Induction Motors Based on Neural Network and Genetic Algorithm Using Stator Current Signals

2006 ◽  
Vol 2006 ◽  
pp. 1-13 ◽  
Author(s):  
Tian Han ◽  
Bo-Suk Yang ◽  
Won-Ho Choi ◽  
Jae-Sik Kim
Keyword(s):  
Neural Network ◽  
Genetic Algorithm ◽  
Wavelet Transform ◽  
Fault Diagnosis ◽  
Induction Motors ◽  
Measurement Data ◽  
Discrete Wavelet ◽  
Diagnosis System ◽  
Stator Current ◽  
Fault Diagnosis System

This paper proposes an online fault diagnosis system for induction motors through the combination of discrete wavelet transform (DWT), feature extraction, genetic algorithm (GA), and neural network (ANN) techniques. The wavelet transform improves the signal-to-noise ratio during a preprocessing. Features are extracted from motor stator current, while reducing data transfers and making online application available. GA is used to select the most significant features from the whole feature database and optimize the ANN structure parameter. Optimized ANN is trained and tested by the selected features of the measurement data of stator current. The combination of advanced techniques reduces the learning time and increases the diagnosis accuracy. The efficiency of the proposed system is demonstrated through motor faults of electrical and mechanical origins on the induction motors. The results of the test indicate that the proposed system is promising for the real-time application.

scholarly journals Analisis Sistem Pengukuran Getaran Mems Accelerometer ADXL345

2019 ◽  
Vol 9 (02) ◽  
pp. 63-67
Author(s):  
Indra Feriadi ◽  
Fajar Aswin ◽  
M Iqbal Nugraha
Keyword(s):  
Measurement System ◽  
Low Cost ◽  
Measurement Data ◽  
Micro Electro Mechanical System ◽  
Vibration Measurement ◽  
Measurement Technology ◽  
Mems Accelerometer ◽  
Analysis Technique ◽  
Develop Approach ◽  
Mems Technology

Vibration measurement technology using conventional sensors such as piezoelectric (PZT) Accelerometer is still expensive. Currently, many low-cost vibration measuring devices have been developed by using Micro Electro Mechanical System (MEMS) technology. This study aims to analyze the results of vibration measurement system MEMS Accelerometer ADXL345 with PZT Accelerometer. This research applies design and develop approach with comparative data analysis technique, that is comparing data of result of measurement of MEMS Accelerometer ADXL345 to PZT Accelerometer Vibroport80. The construction comprises the ADXL345 sensor connected to the Arduino Mega 2560 microcontroller operated by Widows operating system and programming language Arduino IDE 1.08. Testing of measurements at Bearing speeds of 500, 1000, and 1500 RPM with length of time measurements at 5, 10, and 20 seconds respectively. The analysis of the test results shows that the MEMS Accelerometer ADXL345 of vibration measurement system can measure, process and display vibration measurement data larger 3% than PZT Accelerometer and can provide the best measurement accuracy at 20 seconds measurement length of time.

scholarly journals Torque characteristic of SI engine in dynamic operating states

2017 ◽  
Vol 171 (4) ◽  
pp. 175-180
Author(s):  
Piotr BERA
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Measurement Data ◽  
Operating Conditions ◽  
Training Data ◽  
Opening Angle ◽  
Si Engine ◽  
Engine Operation ◽  
Dynamic Operating ◽  
Artificial Neural

The article presents torque characteristic of the engine in dynamic operating conditions as a function of engine speed and throttle opening angle. All mentioned parameters are analyzed as independent variables over time. To develop such a characteristic an artificial neural network is used. The training data were obtained from measurements carried out on the test bench on SI engine. The operating states reflect all possible configurations of these parameters, which may occur during use of the vehicle in real traffic conditions. The article shows design of an artificial neural network that allows to designate the required dependences. Moreover, it describes the fit of the model to the measurement data, which clearly indicates its correctness. Then the developed characteristic in dynamic states is compared with the characteristic in static working states. The differences between them for selected cases of engine operation states are presented. It shows the versatility of the presented method.

An Intelligent Nonlinear System Identification Method With Robust State Estimation

2011 ◽  
Author(s):  
Jason R. Kolodziej
Keyword(s):  
Neural Network ◽  
System Identification ◽  
Nonlinear System ◽  
State Estimation ◽  
Nonlinear System Identification ◽  
Measurement Data ◽  
Training Data ◽  
Feed Forward Neural Network ◽  
State Estimate ◽  
Smooth State

An approach is presented for nonlinear system identification by combining a proven state estimation technique, Minimum Model Error (MME) estimation, with a feed-forward neural network. The MME/NN hybrid algorithm first determines from measurement data a smooth state estimate of the system as well as identifying a correction to the assumed system model. Next, the state estimates are then used as training data for the neural network’s representation of the unmodeled system dynamics. The performance is then shown to be an improvement over a stand alone black-box neural network.

scholarly journals Fast Terahertz Coded-Aperture Imaging Based on Convolutional Neural Network

2020 ◽  
Vol 10 (8) ◽  
pp. 2661
Author(s):  
Fengjiao Gan ◽  
Chenggao Luo ◽  
Xingyue Liu ◽  
Hongqiang Wang ◽  
Long Peng
Keyword(s):  
Neural Network ◽  
Signal To Noise Ratio ◽  
Low Cost ◽  
Mapping Function ◽  
Training Data ◽  
Coded Aperture ◽  
Model Errors ◽  
Echo Signal ◽  
Simulation Experiments ◽  
Coded Aperture Imaging

Terahertz coded-aperture imaging (TCAI) has many advantages such as forward-looking imaging, staring imaging and low cost and so forth. However, it is difficult to resolve the target under low signal-to-noise ratio (SNR) and the imaging process is time-consuming. Here, we provide an efficient solution to tackle this problem. A convolution neural network (CNN) is leveraged to develop an off-line end to end imaging network whose structure is highly parallel and free of iterations. And it can just act as a general and powerful mapping function. Once the network is well trained and adopted for TCAI signal processing, the target of interest can be recovered immediately from echo signal. Also, the method to generate training data is shown, and we find that the imaging network trained with simulation data is of good robustness against noise and model errors. The feasibility of the proposed approach is verified by simulation experiments and the results show that it has a competitive performance with the state-of-the-art algorithms.

scholarly journals COVID-19 Lockdown in Belgrade: Impact on Air Pollution and Evaluation of a Neural Network Model for the Correction of Low-Cost Sensors’ Measurements

2021 ◽  
Vol 11 (22) ◽  
pp. 10563
Author(s):  
Ivan Vajs ◽  
Dejan Drajic ◽  
Zoran Cica
Keyword(s):  
Neural Network ◽  
Air Pollution ◽  
Convolutional Neural Network ◽  
Network Model ◽  
Neural Network Model ◽  
Low Cost ◽  
Training Data ◽  
Traffic Density ◽  
The Impact ◽  
Public Monitoring

In this paper, we explore the impact of the COVID-19 lockdown in Serbia on the air pollution levels of CO, NO2 and PM10 alongside the possibility for low-cost sensor usage during this period. In the study, a device with low-cost sensors collocated with a reference public monitoring station in the city of Belgrade is used for the same period of 52 days in 2019 (pre-COVID-19 period), 2020 (COVID-19 lockdown) and 2021 (post-COVID-19 period). Low-cost sensors’ measurements are improved by using a convolutional neural network that applies corrections of the influence of temperature and relative humidity on the low-cost sensors. As a result of this study we have noticed a remarkable decrease in NO2 (primarily related to traffic density), while on the other hand CO and PM10, related to domestic heating sources and heating plants, showed constant or slightly higher levels. The obtained results are in accordance with other published work in this area. The low-cost sensors have shown a satisfactory correlation with the reference CO measurements during the lockdown, while the NO2 and PM10 measurements of 2020 were corrected using a convolutional neural network trained on meteorological and pollutant data from 2019. The results include an improvement of 0.35 for the R2 of NO2 and an improvement of 0.13 for the R2 of PM10, proving that our neural network model trained on data from 2019 can improve the performance of the sensor in the lockdown period in 2020. This means that our neural network model is very robust, as it exhibits good performance even in the case where training data from the prior year (2019) are used in the following year (2020) in very different environment circumstances—a lockdown.

Sign in / Sign up

Export Citation Format

Share Document