Unlock instant, AI-driven research and patent intelligence for your innovation.

Fault diagnosis method for large-scale water-turbine generator set

A technology for hydroelectric generators and generator sets, which is applied to hydroelectric power generation, engine components, machines/engines, etc. It can solve problems such as no calibration, deep learning model overfitting, and less data, and achieve high fault diagnosis accuracy.

Active Publication Date: 2021-06-01
浙江富春江水电设备有限公司
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Hydropower units have the characteristics of less abnormal sample data and no calibration of faults. Therefore, the deep learning model is prone to overfitting problems

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Fault diagnosis method for large-scale water-turbine generator set
  • Fault diagnosis method for large-scale water-turbine generator set
  • Fault diagnosis method for large-scale water-turbine generator set

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0033] A method for diagnosing faults of a large hydroelectric generating set, comprising the following steps:

[0034] a. Collect n types of fault vibration signals of hydro-generators and vibration signals of normal operation of hydro-generators. The signal data is divided into training set and test set. Each (n+1) vibration condition needs to be collected for training 100 data points for and 50 data points for testing;

[0035] b. Use the sampling frequency information to convert the hydro-generator vibration time-domain signal into a frequency-domain signal using fast spectral kurtosis,

[0036] K x (f)=S 4 (f) / (S 2 (f)) 2 -2 (Formula 1)

[0037] S n (f)=En > (Formula 2)

[0038] where, f≠0, S n (f) is the nth-order spectral moment of the signal, E is the mean value, |·| is the modulus, L(f,t) is the complex envelope of the signal x(t) at f;

[0039] c. Classify the frequency domain signal with a stacked sparse autoencoder;

[0040] d. Sparse autoencoders impose ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a fault diagnosis method for a large-scale water-turbine generator set. The fault diagnosis method comprises the following steps: firstly, collecting n types of fault vibration signals of a water-turbine generator, and normal operation vibration signals of the water-turbine generator, and dividing signal data into a training set and a test set, wherein 100 data points for training and 50 data points for testing need to be collected for each (n + 1) vibration condition; secondly, converting vibration time-domain signals of the water-turbine generator into frequency-domain signals by utilizing sampling frequency information and using fast spectral kurtosis; classifying the frequency-domain signals by using a stacked sparse auto-encoder; performing hyper-parameter selection on the stacked sparse auto-encoder by using a particle swarm optimization algorithm so as to select an optimal hyper-parameter suitable for fault diagnosis of the water-turbine generator set; and testing a test sample by adopting the stacked auto-encoding network which is trained to be qualified so as to identify the fault type of the vibration signals of the water-turbine generator set. The fault of the water-turbine generator can be diagnosed without a large amount of fault sample data, and the fault diagnosis precision is high.

Description

technical field [0001] The invention relates to the technical field of hydraulic generators, in particular to a fault diagnosis method for large hydraulic generator sets. Background technique [0002] With the enhancement of people's awareness of energy saving and environmental protection, hydropower as a green energy is developing vigorously. Large-scale hydropower stations in my country have the characteristics of high water head, high altitude, poor cavitation performance, strong mechanical vibration, complex layout, multi-unit hydraulic unit, long water diversion pipeline, huge water flow inertia, and close hydraulic and electric coupling. The environment is getting worse and worse, and there are more and more incentive factors that cause the failure of hydroelectric generating units, which brings a series of international academic frontier problems and engineering technical problems that need to be solved urgently for the safe and stable operation of hydroelectric genera...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): F03B11/00
CPCF03B11/008Y02E10/20
Inventor 马建峰潘骏周叶陈文华缪熙熙曹登峰刘胜柱成德明张炜邵保安
Owner 浙江富春江水电设备有限公司