Method and device for detecting and evaluating defect

Patent metadata

Field	Value
Publication number	US-11099156-B2
Application number	US-201916403990-A
Country	US
Kind code	B2
Filing date	May 6, 2019
Priority date	Jun 29, 2018
Publication date	Aug 24, 2021
Grant date	Aug 24, 2021

Abstract

The present disclosure provides a method and a device for detecting and evaluating a defect with electromagnetic multi-field coupling. The method includes magnetizing a pipeline with the electromagnetic multi-field coupling; detecting a defect of the pipeline along an axial direction of the pipeline at a constant speed; collecting signals at a position of the defect to obtain magnetic leakage signals in three dimensions and an electrical impedance signal; pre-processing the collected signals; decoupling the pre-processed signals, to obtain decoupled magnetic leakage signals and a decoupled electrical impedance signal; performing impedance analysis on the decoupled electrical impedance signal, and determining a type of the defect based on a phase angle of the decoupled electrical impedance signal; and performing quantification analysis on the decoupled magnetic leakage signals and performing quantification evaluation on a size of the defect using a neural network defect quantification method.

First claim

What is claimed is:

1. A method for detecting and evaluating defect with electromagnetic multi-field coupling, comprising:
magnetizing, by a magnetizing circuit, a pipeline to be detected with the electromagnetic multi-field coupling;
detecting, by a detector, a defect of the pipeline along an axial direction of the pipeline at a constant speed;
collecting, by a signal collecting circuit, signals at a position of the defect to obtain magnetic leakage signals in three dimensions at the position and an electrical impedance signal at the position;
pre-processing, by a signal processing circuit, the magnetic leakage signals in three dimensions and the electrical impedance signal;
decoupling, by a processor, the pre-processed magnetic leakage signals in three dimensions and the electrical impedance signal, to obtain decoupled magnetic leakage signals in three dimensions and a decoupled electrical impedance signal;
performing, by the processor, impedance analysis on the decoupled electrical impedance signal, and determining, by the processor, a type of the defect based on a phase angle of the decoupled electrical impedance signal; and
performing, by the processor, a neural network defect quantification method, comprising:
establishing a first neural network quantification model for the magnetic leakage signals of the corrosion defect; and
establishing a second neural network quantification model for the magnetic leakage signals of the crackle defect, the second neural network quantification model being different from the first neural network quantification model;
wherein each of the first neural network quantification model and the second neural network quantification model have an input of the decoupled magnetic leakage signals, and an output of quantitative values indicating length, width and depth of the defect.

2. The method of claim 1 , wherein magnetizing the pipeline to be detected with the electromagnetic multi-field coupling further comprises:
providing a saturated direct-current magnetization field using a permanent magnet or a direct-current coil, and providing simultaneously a high-frequency alternating-current magnetization field using an alternating-current coil, wherein a frequency of a high-frequency alternating-current excitation signal for providing the high-frequency alternating-current magnetization field is within a range of 20 kHz to 100 kHz, a direction of alternating-current magnetization is along a radial direction of the pipeline, and a direction of direct-current magnetization is along an axial direction of the pipeline.

3. The method of claim 2 , wherein the pre-processed magnetic leakage signals in three dimensions and the electrical impedance signal are decoupled according to:

B
x
′

=

B
x

B
y
′

=

B
y

B
z
′

=

E
/
n

⁢

⁢
s

-

2

⁢

sin
⁡

(

ω
⁢

⁢
t

+
φ

)

·

∇

B
z

-

2

⁢
ω
⁢

⁢

cos
⁡

(

ω
⁢

⁢
t

+
φ

)

⁢

B
z

1
-

(

2

⁢

sin
⁡

(

ω
⁢

⁢
t

+
φ

)

)

·

∇

-

2

⁢

ωcos
⁡

(

ω
⁢

⁢
t

+
φ

)

Assignees

• Univ Tsinghua

Inventors

• Huang Songling
• Zhao Wei
• Wang Shen
• Yu Xinjie
• Peng Lisha
• Zou Jun
• Wang Fuping
• Dong Jiarui
• Gui Lin
• Long Yue

Classifications

• G01N27/83 Primary

  by investigating stray magnetic fields · CPC title

• G01N27/023

  where the material is placed in the field of a coil · CPC title

• G01N27/9046

  by analysing electrical signals · CPC title

• G01N17/006

  of metals · CPC title