Time-frequency analysis and support vector machine in automatic detection of defect from vibration signal of centrifugal pump

Kumar, Anil; Kumar, Rajesh

doi:10.1016/j.measurement.2017.04.041

Cited by 112 publications

(37 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To enlarge the application cases of training samples collected from fault simulators, an intelligent fault diagnosis approach was proposed using transfer learning to transfer fault samples from laboratory bearings to locomotive bearings [5]. By using limited fault samples as training and testing samples, the performance of machine learning methods have been verified to detect faults in bearings, such as extreme learning machine (ELM), support vector machine (SVM), neural networks (NNs) [6][7][8][9], etc. However, in the real world, it is difficult to obtain sufficient suitable training samples to represent various kinds of bearing faults that may occur in actual mechanical systems.…”

Section: Introductionmentioning

confidence: 99%

A Personalized Diagnosis Method to Detect Faults in a Bearing Based on Acceleration Sensors and an FEM Simulation Driving Support Vector Machine

Liu

Huang

Xiang

2020

Sensors

View full text Add to dashboard Cite

Classification of faults in mechanical components using machine learning is a hot topic in the field of science and engineering. Generally, every real-world running mechanical system exhibits personalized vibration behaviors that can be measured with acceleration sensors. However, faulty samples of such systems are difficult to obtain. Therefore, machine learning methods, such as support vector machine (SVM), neural network (NNs), etc., fail to obtain agreeable fault detection results through smart sensors. A personalized diagnosis fault method is proposed to activate the smart sensor networks using finite element method (FEM) simulations. The method includes three steps. Firstly, the cosine similarity updated FEM models with faults are constructed to obtain simulation signals (fault samples). Secondly, every simulation signal is separated into sub-signals to solve the time-domain indexes to generate the faulty training samples. Finally, the measured signals of unknown samples (testing samples) are inserted into the trained SVM to classify faults. The personalized diagnosis method is applied to detect bearing faults of a public bearing dataset. The classification accuracy ratios of six types of faults are 90% and 92.5%, 87.5% and 87.5%, 85%, and 82.5%, respectively. It confirms that the present personalized diagnosis method is effectiveness to detect faults in the absence of fault samples.

show abstract

Section: Introductionmentioning

confidence: 99%

A Personalized Diagnosis Method to Detect Faults in a Bearing Based on Acceleration Sensors and an FEM Simulation Driving Support Vector Machine

Liu

Huang

Xiang

2020

Sensors

View full text Add to dashboard Cite

show abstract

“…In fact, the flow-induced vibration behaviors of the shaft system contain a lot of useful information, and they can be easily measured in a cost-effective manner [21]. Therefore, the monitoring and analyzing of shaft vibration data has become the mainstream approach for the safety assessment and fault diagnosis of fluid machinery [22,23].…”

Section: Introductionmentioning

confidence: 99%

CFD-Rotordynamics Sequential Coupling Simulation Approach for the Flow-Induced Vibration of Rotor System in Centrifugal Pump

Zhang

You

et al. 2020

Applied Sciences

View full text Add to dashboard Cite

Featured Application: The numerical approach in this article can provide abundant information about the flow induced vibration of the pump rotor and may offer theoretical guidance to the design of reducing vibration and safety monitoring of centrifugal pumps.Abstract: Vibration of the rotor system is closely related with the operation stability of centrifugal pump, and it is inevitably induced by the unsteady inner flow. An unsteady computational fluid dynamics model coupling with a rotordynamics model was presented, and the corresponding numerical calculation program, including a self-designed rotordynamics code, was developed on the commercial package ANSYS Workbench. The validity of the numerical calculation model was verified by a hydraulic performance and vibration test based on an industrial centrifugal pump. The hydraulic radial forces on impeller, pressure pulsation, deformation, and vibration of the main shaft under nine different flow rates were systematically investigated and explained preliminarily from the view of inner unstable flow. Results show that the blade passing frequency is the dominant frequency in the fluctuation of the above dynamical behaviors, which is closely related to the rotor-stator interaction between the impeller and volute casing. This study built the connection between internal fluid flow in the centrifugal pump and the vibration of its external rotor structure, and may provide theoretical references for the design of vibration-reducing and safety monitoring strategies design of centrifugal pumps.3 of 33 former works, the first and foremost improvement of this work is the application of rotordynamics in the CSD, where the bearing damping matrix and gyroscopic matrix have been added to the general dynamic equation. With this advantage, the above new approach is able to provide a wealth shaft oscillation information including vibration velocity and shaft centerline orbit, with both the unsteady hydraulic force on the impeller and the rotor dynamic properties in consideration.In this article, an industrial split-case double suction centrifugal pump was chosen for study, and the unsteady flow as well as rotor assembly vibration behaviors under different operation conditions were investigated. The remaining of this article was organized as follows. In Section 2, the mathematical model and its overall frame are presented. Section 3 describes the computation details and experiment setup for a research case. Sections 4 and 5 are the analysis and discussion for the numerical and experimental results, followed by the conclusions in Section 6. Sequential Coupling Model Hypothesis and SimplificationThis study constructs the CFD-rotordynamics sequential coupling numerical model of centrifugal pump mainly based on following hypotheses and simplifications:(1) The coupling between flow field and rotordynamics belongs to one-way coupling. It means that only influences of flow induced exciting force on the vibration behavior of the rotor system were considered, but interferences of the deformati...

show abstract

“…Samanipour et al 5 detected cavitations of centrifugal pump using pressure timedomain features. Kumar and Kumar 6 presented an automatic detection method of centrifugal pump using time-frequency features. Muralidharan and Sugumaran 7,8 proposed fault diagnosis methods of monoblock centrifugal pump using wavelet features.…”

Section: Introductionmentioning

confidence: 99%

Experimental investigation and multi-conditions identification method of centrifugal pump using Fisher discriminant ratio and support vector machine

Qiu

Huang

2019

Advances in Mechanical Engineering

View full text Add to dashboard Cite

For identifying the operation situations of centrifugal pumps by artificial intelligence, we performed an experiment on multi-flow conditions. The multi-flow conditions were simulated by adjusting an automatic flow-regulating valve installed on outlet pipe, and the vertical vibration signals of 20 flow points at the bearing house were collected by the test system. By time-domain analysis, frequency-domain analysis, information entropy, empirical modal decomposition, and wavelet packet decomposition methods, a comprehensive feature space was constructed. In addition, the optimal features were selected by Fisher discriminant ratio, and the dimensionality of the selected optimal features was reduced with principal component analysis. Finally, support vector machine algorithm was employed to identify the real-time flow condition, and the hyper-parameters of support vector machine classifier model were optimized by a grid search technique. Results show that the vibration test can effectively simulate the operation situation of centrifugal pumps under multi-flow conditions, and the proposed multi-flow conditions identification method has achieved a good identification performance.

show abstract

Time-frequency analysis and support vector machine in automatic detection of defect from vibration signal of centrifugal pump

Cited by 112 publications

References 29 publications

A Personalized Diagnosis Method to Detect Faults in a Bearing Based on Acceleration Sensors and an FEM Simulation Driving Support Vector Machine

A Personalized Diagnosis Method to Detect Faults in a Bearing Based on Acceleration Sensors and an FEM Simulation Driving Support Vector Machine

CFD-Rotordynamics Sequential Coupling Simulation Approach for the Flow-Induced Vibration of Rotor System in Centrifugal Pump

Experimental investigation and multi-conditions identification method of centrifugal pump using Fisher discriminant ratio and support vector machine

Contact Info

Product

Resources

About