scholarly journals Gearbox Fault Identification and Classification with Convolutional Neural Networks

2015 ◽  
Vol 2015 ◽  
pp. 1-10 ◽  
Author(s):  
ZhiQiang Chen ◽  
Chuan Li ◽  
René-Vinicio Sanchez
Keyword(s):  
Fault Diagnosis ◽  
Learning Algorithm ◽  
Fault Identification ◽  
Test Case ◽  
Vibration Signals ◽  
Time Domain Signal ◽  
Deep Learning Algorithm ◽  
Statistical Measures ◽  
Basic Fault ◽  
The Time Domain

Vibration signals of gearbox are sensitive to the existence of the fault. Based on vibration signals, this paper presents an implementation of deep learning algorithm convolutional neural network (CNN) used for fault identification and classification in gearboxes. Different combinations of condition patterns based on some basic fault conditions are considered. 20 test cases with different combinations of condition patterns are used, where each test case includes 12 combinations of different basic condition patterns. Vibration signals are preprocessed using statistical measures from the time domain signal such as standard deviation, skewness, and kurtosis. In the frequency domain, the spectrum obtained with FFT is divided into multiple bands, and the root mean square (RMS) value is calculated for each one so the energy maintains its shape at the spectrum peaks. The achieved accuracy indicates that the proposed approach is highly reliable and applicable in fault diagnosis of industrial reciprocating machinery. Comparing with peer algorithms, the present method exhibits the best performance in the gearbox fault diagnosis.

scholarly journals Application of Model-Based Deep Learning Algorithm in Fault Diagnosis of Coal Mills

2020 ◽  
Vol 2020 ◽  
pp. 1-14
Author(s):  
Yifan Jian ◽  
Xianguo Qing ◽  
Yang Zhao ◽  
Liang He ◽  
Xiao Qi
Keyword(s):  
Deep Learning ◽  
Fault Diagnosis ◽  
Learning Algorithm ◽  
Mechanism Model ◽  
Model Based ◽  
Deep Learning Algorithm ◽  
Stacked Autoencoders ◽  
Different Types ◽  
Operation Status ◽  
Types Of Faults

The coal mill is one of the important auxiliary engines in the coal-fired power station. Its operation status is directly related to the safe and steady operation of the units. In this paper, a model-based deep learning algorithm for fault diagnosis is proposed to effectively detect the operation state of coal mills. Based on the system mechanism model of coal mills, massive fault data are obtained by analyzing and simulating the different types of faults. Then, stacked autoencoders (SAEs) are established by combining the said data with the deep learning algorithm. The SAE model is trained by the fault data, which provide it with the learning and identification capability of the characteristics of faults. According to the simulation results, the accuracy of fault diagnosis of coal mills based on SAE is high at 98.97%. Finally, the proposed SAEs can well detect the fault in coal mills and generate the warnings in advance.

scholarly journals An SDP Characteristic Information Fusion-Based CNN Vibration Fault Diagnosis Method

2019 ◽  
Vol 2019 ◽  
pp. 1-14 ◽  
Author(s):  
Xiaoxun Zhu ◽  
Jianhong Zhao ◽  
Dongnan Hou ◽  
Zhonghe Han
Keyword(s):  
Neural Network ◽  
Fault Diagnosis ◽  
Information Fusion ◽  
Fault Identification ◽  
High Complexity ◽  
Test Rig ◽  
Rotor Vibration ◽  
Vibration Signals ◽  
Image Identification ◽  
Diagnosis Method

This study proposes a symmetrized dot pattern (SDP) characteristic information fusion-based convolutional neural network (CNN) fault diagnosis method to resolve issues of high complexity, nonlinearity, and instability in original rotor vibration signals. The method was used to conduct information fusion of real modal components of vibration signals and SDP image identification using CNN in order to achieve vibration fault diagnosis. Compared with other graphic processing methods, the proposed method more fully expressed the characteristics of different vibration signals and thus presented variations between different vibration states in a simpler and more intuitive way. The proposed method was experimentally investigated using simulation signals and rotor test-rig signals, and its validity and advancements were demonstrated using experimental analysis. By using CNN through deep learning to adaptively extract SDP characteristic information, vibration fault identification was ultimately realized.

GEAR FAULT DIAGNOSIS IN TIME DOMAINS VIA BAYESIAN NETWORKS

2013 ◽  
Vol 37 (3) ◽  
pp. 665-672 ◽  
Author(s):  
Chun-Chieh Wang ◽  
Yuan Kang ◽  
Chin-Chi Liao
Keyword(s):  
Neural Network ◽  
Fault Diagnosis ◽  
Bayesian Networks ◽  
Rolling Bearing ◽  
Vibration Signals ◽  
The Neural Network ◽  
Training Samples ◽  
Time Domains ◽  
Gear Fault ◽  
The Time Domain

In gear or rolling bearing systems, it is difficult to extract symptoms from vibration signals where shock vibration signals are present. However, the neural network method cannot provide satisfactory diagnosis results without adequate training samples. Bayesian networks provide an effective approach for fault diagnosis in cases given uncertain and incomplete information. In this study, the statistical factors of vibration signals in the time-domain were used and the diagnosis results by using Bayesian networks were superior to other neural network methods.

Rolling Bearing Fault Diagnosis Research

2012 ◽  
Vol 155-156 ◽  
pp. 87-91
Author(s):  
Zhong Hu Yuan ◽  
Yang Su ◽  
Xiao Xuan Qi
Keyword(s):  
Fault Diagnosis ◽  
Time Domain ◽  
Recognition Rate ◽  
Principal Component ◽  
Rolling Bearing ◽  
Bearing Fault ◽  
Bearing Fault Diagnosis ◽  
Time Domain Signal ◽  
Feature Information ◽  
The Time Domain

According to the characteristics of the rolling bearing fault, we make the research on fault diagnosis. Time domain signal can not perform the fault feature information well. The power spectrum changes the time domain signals into the frequency signals. It sets up the new data model. It uses the principal component analysis on fault diagnosis. It uses T square statistics and Q statistics methods to make fault diagnosis. Simulation experiment results demonstrate that this method provides a high recognition rate.

A hybrid approach for aircraft fault diagnosis based on fault inference and fault identification

2014 ◽  
Vol 118 (1199) ◽  
pp. 81-97 ◽  
Author(s):  
X. Liu ◽  
Z. Liu
Keyword(s):  
Fault Diagnosis ◽  
Flight Control ◽  
Correct Diagnosis ◽  
Hybrid Approach ◽  
Fault Identification ◽  
Multiple Model ◽  
Basic Fault ◽  
Cause Of Failure ◽  
Fault Mode ◽  
Fault Inference

Abstract A cockpit instrumentation system provides various elements of information for pilots. However, logical inference based on a cockpit instruments fault tree (FT) and reliability sometimes cannot give a correct diagnosis of failures. In addition, in flight control systems (FCS), a fault identification method based on the multiple-model (MM) estimator cannot find the basic fault cause. To deal with these problems, a hybrid approach which is capable of integrating inference and fault identification is proposed. In this approach, the event nodes of the FT which have correlations to the FCS are separated into modules. Each module corresponds to a fault mode of the FCS. To use these correlations, fault inference and the MM estimator can share fault diagnosis information. Simulation results show that the proposed approach is helpful in detecting the root cause of failure and is more correct than single fault diagnosis method.

scholarly journals Fault diagnosis of motor bearing based on deep learning

2019 ◽  
Vol 11 (9) ◽  
pp. 168781401987562 ◽  
Author(s):  
Yifan Jian ◽  
Xianguo Qing ◽  
Liang He ◽  
Yang Zhao ◽  
Xiao Qi ◽  
...  
Keyword(s):  
Deep Learning ◽  
Fault Diagnosis ◽  
Working Conditions ◽  
Input Data ◽  
Learning Algorithm ◽  
Random Noise ◽  
Thermal Power ◽  
Efficient Operation ◽  
Deep Learning Algorithm ◽  
Motor Bearing

The effective fault diagnosis of the motor bearings not only can ensure the smooth and efficient operation of equipment but also can detect and eliminate the running fault in time to prevent major accidents. Based on deep learning algorithm, this article constructs a stacked auto-encoder network. The input data are compressed and reduced by introducing sparsity constraint, so that the network can accurately extract the fault characteristics of the input data, and the fault recognition ability of the network can be improved by introducing random noise. The simulation result shows that the stacked auto-encoder network can not only overcome the shortcomings of traditional fault diagnosis method that requires to distinguish fault samples manually and needs a large number of prior knowledge but also realize the self-learning of fault signal feature. The accuracy rate of fault identification reaches 98%, 94%, 96%, and 95.5% in four different working conditions. What’s more, the network can exhibit strong robustness under different working conditions. Finally, the new research ideas of fault diagnosis in thermal power plant are put forward by copying the idea of fault diagnosis of motor bearing.

Gearbox Fault Diagnosis Using Vibration Signal with Wavelet De-Noising

2011 ◽  
Vol 86 ◽  
pp. 735-738
Author(s):  
Zhi Feng Dong ◽  
Hui Cheng ◽  
Hui Jia Yang ◽  
Wei Fu ◽  
Ji Wei Chen ◽  
...  
Keyword(s):  
Fault Diagnosis ◽  
Vibration Signal ◽  
Frequency Domain Analysis ◽  
Sensor Data ◽  
Spectrum Estimation ◽  
Vibration Signals ◽  
Fault Features ◽  
Vibration Signal Analysis ◽  
Simulation Signal ◽  
The Time Domain

This paper dealt with the gearbox fault diagnosis with vibration signal analysis. The vibration signals from experiment contained a lot of noises which result from motor, gears, bears and box, and were collected through accelerate sensor, data collector and computer. The wavelet de-noising stratification was used to de-noise the vibration signals before the frequency-domain analysis was done. The effects of the simulation signal de-noising was contrasted, and the noise cancellation the power spectrum estimation was carried out. The experimental and analytical results show that the different features are indicated with vibration signal of the normal gearbox and the signal with bolts loosened of the gearbox. The gearbox fault with bolts loosened can be diagnosed by extracting the time-domain fault features of vibration signals.

scholarly journals Adaptive Transfer Learning Based on a Two-Stream Densely Connected Residual Shrinkage Network for Transformer Fault Diagnosis over Vibration Signals

Electronics ◽  
2021 ◽  
Vol 10 (17) ◽  
pp. 2130
Author(s):  
Xiaoyan Liu ◽  
Yigang He ◽  
Lei Wang
Keyword(s):  
Deep Learning ◽  
Fault Diagnosis ◽  
Transfer Learning ◽  
Learning Algorithm ◽  
Operating Conditions ◽  
Data Set ◽  
Vibration Signals ◽  
Time Frequency ◽  
Diagnosis Method ◽  
Transformer Fault

Vibration signal analysis is an efficient online transformer fault diagnosis method for improving the stability and safety of power systems. Operation in harsh interference environments and the lack of fault samples are the most challenging aspects of transformer fault diagnosis. High-precision performance is difficult to achieve when using conventional fault diagnosis methods. Thus, this study proposes a transformer fault diagnosis method based on the adaptive transfer learning of a two-stream densely connected residual shrinkage network over vibration signals. First, novel time-frequency analysis methods (i.e., Synchrosqueezed Wavelet Transform and Synchrosqueezed Generalized S-transform) are proposed to convert vibration signals into different images, effectively expanding the samples and extracting effective features of signals. Second, a Two-stream Densely Connected Residual Shrinkage (TSDen2NetRS) network is presented to achieve a high accuracy fault diagnosis under different working conditions. Furthermore, the Residual Shrinkage layer (RS layer) is applied as a nonlinear transformation layer to the deep learning framework to remove unimportant features and enhance anti-interference performance. Lastly, an adaptive transfer learning algorithm that can automatically select the source data set by using the domain measurement method is proposed. This algorithm accelerates the training of the deep learning network and improves accuracy when the number of samples is small. Vibration experiments of transformers are conducted under different operating conditions, and their results show the effectiveness and robustness of the proposed method.

Sign in / Sign up

Export Citation Format

Share Document