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

Method for diagnosing faults of nonlinear analog circuit based on Wiener kernels and neural network

A technology for simulating circuit faults and neural networks, which is applied in the field of pattern recognition of nonlinear analog circuits, and can solve the problems that Volterra series are not mutually orthogonal and cannot be expanded by Volterra series, so as to improve diagnostic efficiency, wide adaptability, and accuracy. high degree of effect

Inactive Publication Date: 2010-08-25
哈尔滨海恒博涵科技发展有限公司
View PDF0 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, many nonlinear circuits are described by Volterra functional series. However, for some non-analytic nonlinear systems, Volterra series expansion cannot be used, and the terms of the Volterra series are not mutually orthogonal.

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
  • Method for diagnosing faults of nonlinear analog circuit based on Wiener kernels and neural network
  • Method for diagnosing faults of nonlinear analog circuit based on Wiener kernels and neural network
  • Method for diagnosing faults of nonlinear analog circuit based on Wiener kernels and neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0035] The steps of the non-linear analog circuit fault diagnosis method of Wiener kernel and neural network of the present invention:

[0036] (1) First determine the various fault states of the nonlinear analog circuit under test, assuming that there are m kinds of states in total, and establish a fault state set;

[0037] (2) Then sequentially apply Gaussian white noise to the non-linear analog circuit under test in the above fault states as an input signal, and measure the input and output signals at the same time to obtain a sampled data sequence, and obtain each part of the circuit under test through data processing The corresponding first n-order Wiener kernel in the fault state;

[0038] (3) Take the order n of the Wiener kernel obtained in the previous step as the number of input neurons of the neural network, take the number of fault states as the number of output neurons of the neural network, establish a BP neural network, and use each state of the diagnosed system...

Embodiment 2

[0041] The non-linear analog circuit fault diagnosis method of Wiener kernel and neural network described in embodiment 1, in step (1), determine the possible m kinds of fault states of the non-linear analog circuit under test, and carry out numbering, including:

[0042] (a) It is determined that all components of the nonlinear analog circuit under test are in a normal state with nominal parameters;

[0043] (b) Determine the soft fault states such as the actual value of the element in the non-linear analog circuit being tested is too large or too small;

[0044] (c) Determine the hard fault states such as short circuit and open circuit of the components in the tested nonlinear analog circuit;

[0045] (d) Number the various states mentioned above, which are respectively 1, 2, ..., m, where m is a natural number.

Embodiment 3

[0047] The non-linear analog circuit fault diagnosis method of the Wiener kernel and neural network of embodiment 1 or 2, in step (2), the first n order Wiener kernels of each fault state are obtained by following steps:

[0048] (a) Make the non-linear analog circuit under test in fault state 1;

[0049] (b) Apply Gaussian white noise to the above circuit as the input signal, and measure the input and output signals at the same time to obtain the sampling sequence data, and use the method of finding multi-order correlation functions to calculate the Wiener kernel k of each order 10 , k 11 , k 12 , k 13 …k 1n , the specific calculation formula is as follows:

[0050] For nonlinear systems, when the input x(t) is Gaussian white noise, the output y(t) can be expanded into a Wiener series form

[0051] y ( t ) = Σ i = 0 ∞ ...

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 provides a method for diagnosing faults of a nonlinear analog circuit based on Wiener kernels and a neural network. The existing nonlinear systems are difficult to describe mathematically and lack uniform description method. The invention relates to feature extraction, mode recognition and fault diagnosis technology of the nonlinear analog circuit and is characterized by determining a fault state set of the nonlinear analog circuit to be tested; obtaining the first n-order Wiener kernel of each fault state in sequence; establishing a BP neural network and training the neural network with each state code and the corresponding n-order Wiener kernel of the state code; and obtaining the first n-order Wiener kernel of the circuit to be diagnosed and using the kernel as input of the neural network and output of the neural network as the result of diagnosis. By the method, the features of part of nonlinear circuits with Volterra series unable to be described can be extracted, the terms of output and expanded series are orthogonal, feature extraction and data processing are simpler, the diagnostic system has strong generalization capability, and the method is high in accuracy and strong in practicability. The method is used for diagnosing the faults of electronic circuits.

Description

Technical field: [0001] The invention relates to pattern recognition, feature extraction method and fault diagnosis of nonlinear analog circuit, in particular to extraction of Wiener kernel, establishment and training of neural network and fault diagnosis method of nonlinear analog circuit. Background technique: [0002] With the development of digital technology and the improvement of integration technology, the proportion of analog circuits in hybrid circuits is getting smaller and smaller, but analog circuits cannot be replaced, and the links connected with specific processes must use analog circuits. Although the proportion of analog circuits is small, the faults caused by analog circuits are much higher than those caused by digital circuits. However, the diagnostic theory of analog circuits, especially nonlinear analog circuits, is not yet perfect. Therefore, there is an urgent need for a good fault diagnosis method for analog circuits. The method of the present appli...

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): G01R31/316G06N3/02G06N3/08
Inventor 林海军
Owner 哈尔滨海恒博涵科技发展有限公司
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