Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Turbine pump small sample fault judgment method based on data expansion and deep transfer learning

A technology of transfer learning and fault determination, applied in neural learning methods, pattern recognition in signals, biological neural network models, etc., can solve the problems of lack of reliable small-sample data expansion methods, large numbers, etc., to enrich the original data set, Effects of Noise Elimination and Accurate Diagnosis

Pending Publication Date: 2021-12-21
BEIJING AEROSPACE PROPULSION INST
View PDF3 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The technical problem solved by the present invention is: Aiming at the problem that in the current prior art, the traditional fault sample training and learning method requires a large number of labeled samples and lacks a reliable small sample data expansion method, a method based on data expansion and deep transfer learning is proposed. Fault Judgment Method for Small Sample of Turbo Pump

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
  • Turbine pump small sample fault judgment method based on data expansion and deep transfer learning
  • Turbine pump small sample fault judgment method based on data expansion and deep transfer learning
  • Turbine pump small sample fault judgment method based on data expansion and deep transfer learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0046] A small-sample fault determination method for turbo pumps based on data augmentation and deep transfer learning, which includes preprocessing operations such as filtering, sampling extraction, detrending items, and smoothing, which can effectively eliminate noise and make vibration signals more robust On the other hand, make full use of the generative confrontation network, and the high-quality time-spectrum samples generated based on the WGAN-GP model can further enrich the original data set and improve the robustness of the diagnostic model; finally, make full use of the feature extraction in the mature ImageNet model Partial model migration is performed to enable accurate diagnosis of turbo pumps. It solves the problem that it is difficult to carry out fault diagnosis in the case of complex equipment such as turbo pumps with less effective data, insufficient labeled data, variable working conditions, and large differences between experimental data and working conditio...

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

A turbine pump small sample fault judgment method based on data expansion and deep transfer learning comprises preprocessing operations such as filtering, sampling, trend term removal and smoothing, noise can be effectively eliminated, and vibration signals are made to have better robustness; on the other hand, the generative adversarial network is fully utilized, a high-quality time-frequency spectrum sample generated based on the WGAN-GP model can further enrich an original data set, and the robustness of the diagnosis model is improved; finally, a feature extraction part in the mature ImageNet model is fully utilized for model migration, and then accurate diagnosis of the turbine pump is achieved.

Description

technical field [0001] The invention relates to a turbo pump small-sample fault judgment method based on data expansion and deep transfer learning, which belongs to the field of mechanical equipment fault diagnosis. Background technique [0002] The turbo pump is the "heart" of the hydrogen-oxygen engine, and it is a key component for transporting liquid hydrogen and liquid oxygen propellants. Burns and produces enormous thrust, thus providing the enormous power needed for rocket flight. As the most core part of the hydrogen-oxygen engine system, if the operation status of the turbo pump cannot be judged accurately in time, it will lead to sudden failure, which will affect the normal operation and service life of the unit. Casualties, resulting in huge loss of life and property and bad international impact. Therefore, it is of great significance to carry out fault diagnosis on turbopumps to ensure the smooth progress of space missions and help the space industry. [0003]...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06N3/08F02K9/46F02K9/96
CPCG06N3/08F02K9/96F02K9/46G06F2218/04G06F2218/12G06F2218/08
Inventor 窦唯金志磊石珊珊孙铁群李伟张迪石光远
Owner BEIJING AEROSPACE PROPULSION INST
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com