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

A multi-source sensor fault detection method based on ICA and kNN

A source sensor and fault detection technology, applied in the direction of instruments, character and pattern recognition, computer components, etc., can solve the problem of poor real-time performance of sensors and achieve the effect of improving real-time performance

Inactive Publication Date: 2019-01-11
HARBIN INST OF TECH
View PDF0 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the problem of poor real-time performance of existing sensor fault detection, and propose a multi-source sensor fault detection method based on ICA and kNN

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 multi-source sensor fault detection method based on ICA and kNN
  • A multi-source sensor fault detection method based on ICA and kNN
  • A multi-source sensor fault detection method based on ICA and kNN

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0023] A kind of multi-source sensor fault detection method based on ICA and kNN of the present embodiment, described method comprises the following steps:

[0024] Step 1. Under normal working conditions, collect multiple sets of training sample data through multi-source sensors;

[0025] Step 2. Perform centralized and standardized data preprocessing on the training sample data;

[0026] Step 3, using the FastICA algorithm to perform ICA decomposition on the training samples obtained after data preprocessing to obtain independent component components, composite matrices and separation matrices;

[0027] Step 4. Sorting the negentropy values ​​of each independent component component in descending order, taking the independent component components whose negentropy value is greater than the non-Gaussian degree threshold as the main part, and determining the separation matrix corresponding to the main part;

[0028] Step 5, using the main part as an independent component compon...

specific Embodiment approach 2

[0035] The difference from the first embodiment is that in this embodiment, a multi-source sensor fault detection method based on ICA and kNN, the process of collecting multiple sets of training sample data through multi-source sensors in the first step is:

[0036] For a multi-source sensor with m sensitive units, n sets of training sample data collected under normal working conditions are:

[0037]

specific Embodiment approach 3

[0038] Different from the second specific embodiment, a multi-source sensor fault detection method based on ICA and kNN in this embodiment, the process of centralizing and standardizing the data preprocessing of the training sample data in the second step is:

[0039] Centralized processing through formula (2):

[0040]

[0041] In the formula, m represents the number of sensitive units; x ij Indicates the data of the j-th sensitive unit in the training samples collected by the i-th group; i indicates the serial number of the training sample; j indicates the serial number of the sensitive unit; Indicates the mean value of the training sample data of the jth sensitive unit;

[0042] Afterwards, the standardization process is carried out through formula (3):

[0043]

[0044] In the formula, σ j Indicates the standard deviation of the training sample data of the jth sensitive unit.

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 a multi-source sensor fault detection method based on ICA and kNN, belonging to the sensor detection field. The problem of poor real-time performance of sensor fault detectionis solved. The multi-source sensor fault detection method based on ICA and kNN comprises the steps of: collecting and pretreating many groups of normal training sample data by a multi-source sensor;carrying out ICA decomposition to obtain independent component and separation matrix; taking the independent component whose negative entropy value is greater than the non-Gaussian threshold value asthe main part, and taking the main part as a new training sample data set and calculating the statistical quantity; calculating the control line of the statistical quantity; performing centralized andstandardized data preprocessing on the real-time data to be detected; calculating an independent element component corresponding to the real-time data to be detected; calculating the statistical quantity of real-time data to be detected; when the statistical quantity of the real-time data is larger than the control line of the significance level, determining that the multi-source sensor is faulty. Otherwise, the sensor is working properly. The method can remarkably improve the real-time performance of multi-source sensor fault detection.

Description

technical field [0001] The invention relates to a sensor fault detection method, in particular to a multi-source sensor fault detection method based on ICA and kNN. Background technique [0002] In the industrial field, sensors are the source of information acquisition, and the accuracy and reliability of the output data are extremely critical. Output values ​​with high accuracy and reliability are necessary conditions for online monitoring and control of industrial production processes. A multi-source sensor integrates or combines a variety of different types of sensitive units to measure multiple measurands. Compared with traditional single-source sensors, multi-source sensors with smaller size and more complete functions have been developed faster and more widely used. [0003] Due to the increase in the number of sensitive components contained in the multi-source sensor and the existence of complex signal conditioning circuits, the probability of failure is relatively h...

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): G06K9/62
CPCG06F18/2134G06F18/214
Inventor 杨京礼林连雷姜守达
Owner HARBIN INST OF TECH
Features
  • Generate Ideas
  • 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