Supercharge Your Innovation With Domain-Expert AI Agents!

On-line abnormal detection method of direct current power supply based on statistical characteristic evaluation and Gaussian process regression

A Gaussian process regression, DC power technology, applied in power supply testing, calculation, electrical digital data processing and other directions, can solve the problem of large amount of calculation that DC power cannot be detected online abnormally, and achieve the effect of enhancing the difference

Inactive Publication Date: 2018-11-16
HARBIN INST OF TECH
View PDF6 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the shortcomings of the existing method that the DC power supply with an unknown circuit structure cannot be effectively detected online and the amount of calculation is large, and propose an online abnormal detection method for DC power supplies based on statistical feature evaluation and Gaussian process regression

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
  • On-line abnormal detection method of direct current power supply based on statistical characteristic evaluation and Gaussian process regression
  • On-line abnormal detection method of direct current power supply based on statistical characteristic evaluation and Gaussian process regression
  • On-line abnormal detection method of direct current power supply based on statistical characteristic evaluation and Gaussian process regression

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0022] Specific embodiment 1: The online anomaly detection method of DC power supply based on statistical feature evaluation and Gaussian process regression includes the following steps:

[0023] Step 1: Obtain M normal samples of the time-domain output of the normal state of the DC power supply;

[0024] Step 2: Use the M normal samples obtained in Step 1 as the training set of the Gaussian regression process, establish a Gaussian process regression model, obtain N samples of the online output of the DC power supply, and use the index values ​​of the N samples as the Gaussian process established in Step 2 The input of the test set of the regression model uses the Gaussian process regression model to predict the normal output interval of the DC power supply, and the length of the normal output interval is N; wherein the covariance function of the Gaussian process regression model is the joint covariance function k C Including the periodic covariance function k PER sum square ...

specific Embodiment approach 2

[0033] Specific embodiment two: the difference between this embodiment and specific embodiment one is that in the step two, the M normal samples obtained in step one are used as the training set of the Gaussian regression process, and the Gaussian process regression model is established to obtain the online data of the DC power supply. Output N samples, use the index values ​​of N samples as the input of the test set of the Gaussian process regression model established in step 2, and use the Gaussian process regression model to predict the normal output range of the DC power supply. The specific process is:

[0034] Step 21: Take the index values ​​of the M normal samples obtained in step 1 as the input of the training set, which are recorded as 1~M; use the M normal samples as the output of the training set; output the index values ​​of the N samples online as the test set input;

[0035] Step 22: Establish a joint covariance function k C , according to the waveform output b...

specific Embodiment approach 3

[0040] Specific implementation mode three: the difference between this implementation mode and specific implementation mode one or two is: the joint covariance function k is established in the step two or two C The specific process is:

[0041] The present invention first uses the Gaussian regression process to effectively and accurately predict the normal output range of the DC power supply. The Gaussian regression process function f(t) can be completely described by the mean function m(t) and the covariance function k(t, t'), denoted f(t)~GP[m(t),k(t,t′))], t is the input training set, that is, the normal output of the DC power supply. Among them, in order to improve the accuracy of the prediction interval, the covariance function of the Gaussian regression process adopts the joint covariance function k C .

[0042] According to the circuit structure of the DC power supply, the output time domain signal is determined to be a regular periodic signal, so the preferred functi...

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 relates to an on-line abnormal detection method of a direct current power supply based on statistical characteristic evaluation and Gaussian process regression. In order to solve the problem that a prior method cannot effectively carry out on-line abnormal detection on a direct current power supply with an unknown circuit structure and has large calculation amount, the method is provided. The method comprises the following steps of: firstly, acquiring a normal output of a direct current power supply as a training set of a Gaussian regression process, and predicting a normal output interval by utilizing the Gaussian regression process; secondly, acquiring the outputs of the upper limit and lower limit of the predicted normal interval, and calculating the seven corresponding statistical characteristic values to obtain the upper limit value and lower limit value corresponding to each statistical characteristic value as the limit of the characteristic value of each statistical characteristic of the on-line output; finally, comparing the seven statistical characteristic values of the on-line output with the corresponding upper limit and lower limit, if at least one of theoutput seven characteristic values exceeds the corresponding lower limit value, an abnormal state exists in the output. The method is applied to the field of on-line abnormal detection of the direct current power supply.

Description

technical field [0001] The invention relates to an online abnormal detection method of a DC power supply. Background technique [0002] DC power supplies are widely used in many electronic products, such as test instruments, large computers and biomedical instruments. It provides a stable DC power supply, so it is the basis to ensure the normal operation of all electronic products, and its reliability plays a decisive role in the normal operation of subsequent components and the entire electronic product. When the DC power supply is abnormal, it will lead to abnormal operation status of the entire electronic system, serious or even shut down. It will bring huge economic loss and cause inevitable impact. Therefore, online anomaly detection for DC power supply is very necessary and urgent. [0003] The traditional DC power supply anomaly detection method usually adopts the modeling method to obtain the system equation of the DC power supply, and evaluates and identifies the...

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): G06F17/50G01R31/40
CPCG01R31/40G06F30/367
Inventor 俞洋姜月明杨智明彭喜元张晔
Owner HARBIN INST OF TECH
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com