A hybrid of FEM simulations and generative adversarial networks to classify faults in rotor-bearing systems

Gao, Yun; Liu, Xiaoyang; Huang, Huimin; Xiang, Jiawei

doi:10.1016/j.isatra.2020.08.012

Cited by 69 publications

(21 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In the process of vibration signals transmitting to the sensors, noise will inevitably be mixed into the signals and reduces the signalnoise ratio [12,13]. erefore, the sampled original signals are nonstationary and nonlinear signals [14,15].…”

Section: Noise Reduction Algorithm Combined Eemd Withmentioning

confidence: 99%

Axial Piston Pump Fault Diagnosis Method Based on Symmetrical Polar Coordinate Image and Fuzzy C-Means Clustering Algorithm

Jiang

Zhang

et al. 2021

Shock and Vibration

View full text Add to dashboard Cite

In this paper, a fault diagnosis method based on symmetric polar coordinate image and Fuzzy C-Means clustering algorithm is proposed to solve the problem that the fault signal of axial piston pump is not intuitive under the time-domain waveform diagram. In this paper, the sampled vibration signals of axial piston pump were denoised firstly by the combination of ensemble empirical mode decomposition and Pearson correlation coefficient. Secondly, the data, after noise reduction, was converted into images, called snowflake images, according to symmetric polar coordinate method. Different fault types of axial piston pump can be identified by observing the snowflake images. After that, in order to evaluate the research results objectively, the obtained images were converted into Gray-Level Cooccurrence Matrixes. Their multiple eigenvalues were extracted, and the eigenvectors consisting of multiple eigenvalues were classified by Fuzzy C-Means clustering algorithm. Finally, according to the accuracy of classification results, the feasibility of applying the symmetric polar coordinate method to axial piston pump fault diagnosis has been validated.

show abstract

Section: Noise Reduction Algorithm Combined Eemd Withmentioning

confidence: 99%

Axial Piston Pump Fault Diagnosis Method Based on Symmetrical Polar Coordinate Image and Fuzzy C-Means Clustering Algorithm

Jiang

Zhang

et al. 2021

Shock and Vibration

View full text Add to dashboard Cite

show abstract

“…The torsional vibration characteristics of the HTGU are analyzed, meanwhile, the influences of the UMP [21], the fluid seal excitation [22], the rotors Rub-impact [23], the hydraulic unbalance [24], the guide bearing loose [25], the gyroscopic effect [26], the shaft misalignment [27], and pressure pulsation of the draft tube [28] effect on shafting system vibrations are investigated. With the further development of the research, the hydraulic-mechanical-electrical coupling factors are considered in the shafting system to explore the nonlinear vibration dynamic behavior.…”

Section: Symbolmentioning

confidence: 99%

Stability and Sensitivity Analysis and Optimization Control of the Hydro-turbine Generator Unit

Shi

Zhou

2021

Preprint

View full text Add to dashboard Cite

The hydro-turbine governing system (HTGS) and shafting system are mutually coupled. However, the interaction between them has always been neglected. This paper aims to explore the stability and sensitivity of the governor control parameters to the HTGS and shafting system and make the optimal control of the stable operation for the hydro-turbine generator unit(HTGU). First, a novel HTGU motion equation is proposed, which can make connections between the HTGS and the shafting system of the HTGU. And on this basis, the nonlinear coupling mathematical model of the HTGS and the shafting system is established. According to the nonlinear mathematical model, the sensitivity of the governor control parameters on the operating stability of the HTGU is obtained. Then, a multi-objective governor control parameters optimization strategy is proposed. Furthermore, the chaotic-dominated sorting genetic algorithm II(NSGA-II) and multi-objective evolutionary algorithm based on decomposition(MOEAD) were introduced to obtain the optimal control parameter and mutually verify the effectiveness of the optimization effect. Finally, the nonlinear dynamic characteristics of HTGU under optimal control were revealed. The simulation results show that the rotation speed deviation and shafting system vibrations are sensitive on the PID parameters in some ranges and the stable region will be decreased when considering the shafting system vibrations. The multi-objective PID parameter optimization strategy shows good control performance on the nonlinear dynamic characteristics of the HTGU. The shafting system vibrations excited by the coupled vibration sources are quasi-period in 3D space. In addition to this, these results and the optimization strategy can provide some bases for the design and stable operation of the HTGU.

show abstract

“…In Figure 2c, an algorithm-based method is the optimization strategy which uses prior knowledge to search through H in order to find the best hypothesis h3 in H. Data augmentation technology has been used extensively in tasks such as computer vision and natural language processing in the past. In the field of bearing fault diagnosis, Zhang [24] performed data augmentation by manually copying and intercepting the original signal, Li [25] used Generative Adversarial Networks (GAN) to solve the problem of category imbalance, Gao [26] used a combination of finite element (FEM) and GAN, not only to supplement the number of missing labeled data, but also to supplement the missing attributes, and Cubuk [27] described a simple procedure called AutoAugment, which automatically learns the augmentation policy for deep network training. The core idea of the above method is based on the existing labeled data, that is, prior knowledge, to create similar labeled data or copy directly according to the extracted features, so as to train the neural network on a large amount of labeled data to obtain a good performance.…”

Section: Few-shot Learningmentioning

confidence: 99%

Few-Shot Rolling Bearing Fault Diagnosis with Metric-Based Meta Learning

Wang

Kong

et al. 2020

Sensors

View full text Add to dashboard Cite

Fault diagnosis methods based on deep learning and big data have achieved good results on rotating machinery. However, the conventional deep learning method of bearing fault diagnosis is mostly based on laboratory artificial simulation data, and there is an error with actual fault data, which will reduce the generalization performance of the deep learning method. In addition, labeled data are very precious in real industrial environment. Due to expensive equipment and personnel safety issues, it is difficult to obtain a large amount of high-quality fault labeling data. Therefore, in this paper, we propose a metric-based meta-learning method named Reinforce Relation Network (RRN) for diagnosing bearing faults with few-shot samples. In the proposed method, a 1D convolution neural network is used to extract fault features, and a metric learner is used to predict the similarity between samples under different transfer conditions. Label smoothing and the Adabound algorithm are utilized to further improve the performance of network classification. The performance of the proposed method is verified on a dataset which contains artificial damage and natural damage data. The comparison studies with other methods demonstrate the superiority of the proposed method in the few-shot scenario.

show abstract

A hybrid of FEM simulations and generative adversarial networks to classify faults in rotor-bearing systems

Cited by 69 publications

References 37 publications

Axial Piston Pump Fault Diagnosis Method Based on Symmetrical Polar Coordinate Image and Fuzzy C-Means Clustering Algorithm

Axial Piston Pump Fault Diagnosis Method Based on Symmetrical Polar Coordinate Image and Fuzzy C-Means Clustering Algorithm

Stability and Sensitivity Analysis and Optimization Control of the Hydro-turbine Generator Unit

Few-Shot Rolling Bearing Fault Diagnosis with Metric-Based Meta Learning

Contact Info

Product

Resources

About