A Method of Accurately Identifying and Locating Abnormal Variables

A positioning method and variable technology, applied in character and pattern recognition, instruments, testing/monitoring control systems, etc.

Active Publication Date: 2021-04-06
河南工学院
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, there are certain challenges in applying the k-NN method for effective identification of abnormal variables without the knowledge and experience of available fault data, and it also has certain academic research value and practical significance

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 Method of Accurately Identifying and Locating Abnormal Variables
  • A Method of Accurately Identifying and Locating Abnormal Variables
  • A Method of Accurately Identifying and Locating Abnormal Variables

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0109] Please also refer to Figure 5-Figure 10 .

[0110] The embodiment of the present invention provides numerical simulation for verifying the effectiveness and feasibility of the invention. Firstly, through a seven-variable numerical simulation, the accuracy of reconstructed variables using the mean k-NN, k-1NN, weighted k-NN and CNN methods is compared.

[0111] In the embodiment of the present invention, the numerical simulation consists of two latent variables s a and s b The constructed seven variables are composed, as shown in formula (25):

[0112]

[0113] In the formula, e 1 ~e 7 is noise with mean zero and standard deviation 0.01, s a is a random number s from -10 to -7 b Obey normal distribution and N(-15,1). According to formula (25), a total of 500 training samples and 500 samples to be verified are generated. In order to compare the accuracy of the four reconstruction algorithms, the simulation assumes that the sample set to be verified is from ti...

Embodiment 2

[0126] Please also refer to Figure 11-Figure 15 .

[0127] In the embodiment of the present invention, the effectiveness of the proposed method for accurate identification of abnormal variables is verified by a Continuous Stirred Tank Reactor (CSTR) system.

[0128] CSTR is a reactor widely used in polymerization chemical reactions, in which the reaction raw materials enter the reactor at a steady flow rate, and the reactants flow out at the same steady rate. Due to the effect of strong stirring, the fresh material just entering the reactor and the material remaining in the reactor are completely mixed instantly, the concentration and temperature of the material inside the reactor are equal everywhere, and the reaction rate in the continuous stirred tank reactor is determined by The temperature and concentration of the material in the kettle are determined. The principle of the system is as follows: Figure 11 shown. The liquid level and temperature of the reactor are cont...

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 method for accurately identifying and locating abnormal variables, including: modeling based on normal working data, performing fault detection on real-time data through the control threshold of the square distance of the nearest neighbor of the normal model, and then performing abnormal variable detection on the fault data through the variable contribution control threshold of the normal model "Primary" identification, and then reconstruct the abnormal variables excluded in the fault sample and then compare them with the neighbor square distance control threshold again for "secondary" identification of abnormal variables. If the abnormality is still identified, return to the previous step until No more exceptions.

Description

technical field [0001] The invention relates to the field of identification and positioning of abnormal variables, in particular to a method for precise identification and positioning of abnormal variables based on k-nearest neighbor variable contribution analysis and data reconstruction. Background technique [0002] When a fault is detected in an industrial process, extract effective fault information, study the relationship between various faults and abnormal variables, establish an accurate "fault-symptom" table and use it as available knowledge for subsequent fault decision-making and evaluation Library is very necessary, and has become a hot issue in the field of fault diagnosis. Under the framework of research based on statistical theory, the contribution map method is the most commonly used method for outlier variable identification, which can be divided into SPE contribution map and T 2 There are two kinds of contribution graphs. This method visualizes the contribu...

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): G05B23/02G06K9/62
CPCG05B23/0254G06F18/24143
Inventor 王国柱杨晓邢倩杜志勇孟昕元
Owner 河南工学院
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products