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

A diagnosis method and system based on machine failure

A diagnostic method and technology for machine faults, applied in instruments, computer parts, and pattern recognition in signals, etc., can solve problems such as noise pollution, gear tooth manufacturing errors, affecting the accuracy of fault diagnosis, and achieve the effect of improving accuracy.

Active Publication Date: 2020-09-01
ZHEJIANG NORMAL UNIVERSITY
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, due to structural resonance, gear tooth manufacturing error, transmission ratio mismatch, non-Gaussian pulse interference, noise pollution, and information redundancy of multi-channel sensing observations, it is usually difficult to extract accurate eigenfrequency, and it often depends on experience.
[0004] Direct time-frequency analysis of the machine failure observation signal, although the overall nonlinear and unsteady vibration characteristics of the machine can be obtained, but the non-Gaussian pulse interference and noise pollution caused by some objective abnormal factors (see the underlined part) And the information redundancy of multi-channel sensing observation cannot be solved, especially due to the influence of mechanical structure resonance and multi-channel sensing observation information redundancy, the signals measured by the sensors installed on the machine shell are often multiple Combination of machine component vibrations, fault-related features appearing simultaneously in multiple observed sensory signals
Even relying on manual experience, it is difficult to achieve accurate fault feature extraction, which seriously affects the accuracy of fault diagnosis

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
  • A diagnosis method and system based on machine failure
  • A diagnosis method and system based on machine failure
  • A diagnosis method and system based on machine failure

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0107] Redundancy cancellation and scale-invariant feature change fusion based on independent component analysis:

[0108] Assuming that C time-frequency distribution image redundancy cancels the training set Γ c ={x (k) ,k=1,2,…,K},c=1,2,…,C belong to different pattern classes, here x (k) is the kth pixel component training sample, and c is the training set number, that is, the corresponding pattern class number. K is the training set capacity, that is, the number of training samples contained in each training set. According to the following learning rules, the kth training sample x (k) A linear transformation estimate W can be obtained (k) .

[0109] ΔW (k) ∝[(W (k-1) ) T ] -1 +(1-2y (k) )(x (k) ) T

[0110]

[0111] Finally, C ICA redundancy cancellation networks W can be obtained from a given training set of C different pattern classes c =W (K),c=1,2,...,C. These trained networks jointly constitute an image pixel information redundancy canceller, which i...

Embodiment 2

[0119] Training of Independent Component Analysis Redundancy Cancellation Network and Fault Identification and Diagnosis Based on Optimal Support Vector Machine Classification

[0120] The following global separation-mixing matrix is ​​used as a measure to evaluate the training convergence performance of the ICA redundancy cancellation network:

[0121]

[0122] where b ij 、a ij are the elements in the global mixing matrix B and separation matrix A respectively, and ||·|| is a 2-norm function. From Figure 4 It can be seen that the modular parameter of the global mixing-separation matrix BA corresponding to the convergent ICA redundancy cancellation network converges to 0, while the corresponding part of the non-converging network converges to 1. If the convergence and divergence of the measurement parameters are uncertain, the step size and the total number of iterations of network training need to be adjusted.

[0123] The inner product kernel of support vector machin...

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 diagnosis method and system based on machine faults. The diagnosis method includes: obtaining observation signals of machine faults to be diagnosed; determining time-frequency grayscale images corresponding to each product function according to each observation signal; The time-frequency grayscale image corresponding to each of the product functions determines a time-frequency image pixel matrix; performing redundancy cancellation processing on the time-frequency image pixel matrix to obtain a redundancy cancellation image pixel matrix; through the scale-invariant feature, Determine the Gaussian image difference function, gradient magnitude and gradient direction according to the redundant cancellation image pixel matrix; determine multiple base images according to the Gaussian image difference function, gradient magnitude and gradient direction; determine according to multiple base images Image slices; input image slices into the support vector machine classification model to realize fault classification and diagnosis, and improve the accuracy of fault diagnosis.

Description

technical field [0001] The invention relates to the technical field of fault diagnosis, in particular to a diagnosis method and system based on machine faults. Background technique [0002] Mechanical fault diagnosis is essentially a pattern recognition problem. It extracts features from different patterns (such as normal state, gear defect, bearing fault, rotor fault, etc.) and characterizes the data of each pattern. [0003] At present, in order to carry out effective feature extraction, most of the fault diagnosis technologies often need to adopt some special technical methods, such as spectrum analysis, time-frequency analysis and filtering technology, etc., using these technical methods to achieve the purpose of detecting and obtaining the fault characteristic frequency . Typical fault characteristic frequency components include gear mesh frequency and its higher harmonic components, modulation components and so on. However, due to structural resonance, gear tooth man...

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 Patents(China)
IPC IPC(8): G06K9/00G06K9/62
CPCG06F2218/10G06F2218/12G06F18/2411
Inventor 焦卫东常永萍
Owner ZHEJIANG NORMAL UNIVERSITY
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