CN109281653A - A kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array - Google Patents

A kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array Download PDF

Info

Publication number
CN109281653A
CN109281653A CN201710589629.8A CN201710589629A CN109281653A CN 109281653 A CN109281653 A CN 109281653A CN 201710589629 A CN201710589629 A CN 201710589629A CN 109281653 A CN109281653 A CN 109281653A
Authority
CN
China
Prior art keywords
borehole wall
wall defect
signal
pipe nipple
cylinder
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201710589629.8A
Other languages
Chinese (zh)
Inventor
陈诗文
张碧星
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Institute of Acoustics CAS
Original Assignee
Institute of Acoustics CAS
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Institute of Acoustics CAS filed Critical Institute of Acoustics CAS
Priority to CN201710589629.8A priority Critical patent/CN109281653A/en
Publication of CN109281653A publication Critical patent/CN109281653A/en
Pending legal-status Critical Current

Links

Classifications

    • EFIXED CONSTRUCTIONS
    • E21EARTH OR ROCK DRILLING; MINING
    • E21BEARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
    • E21B47/00Survey of boreholes or wells
    • E21B47/002Survey of boreholes or wells by visual inspection
    • EFIXED CONSTRUCTIONS
    • E21EARTH OR ROCK DRILLING; MINING
    • E21BEARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
    • E21B47/00Survey of boreholes or wells
    • E21B47/12Means for transmitting measuring-signals or control signals from the well to the surface, or from the surface to the well, e.g. for logging while drilling
    • E21B47/14Means for transmitting measuring-signals or control signals from the well to the surface, or from the surface to the well, e.g. for logging while drilling using acoustic waves
    • EFIXED CONSTRUCTIONS
    • E21EARTH OR ROCK DRILLING; MINING
    • E21BEARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
    • E21B49/00Testing the nature of borehole walls; Formation testing; Methods or apparatus for obtaining samples of soil or well fluids, specially adapted to earth drilling or wells
    • E21B49/005Testing the nature of borehole walls or the formation by using drilling mud or cutting data

Landscapes

  • Engineering & Computer Science (AREA)
  • Geology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Mining & Mineral Resources (AREA)
  • Environmental & Geological Engineering (AREA)
  • Fluid Mechanics (AREA)
  • General Life Sciences & Earth Sciences (AREA)
  • Geochemistry & Mineralogy (AREA)
  • Geophysics (AREA)
  • Acoustics & Sound (AREA)
  • Remote Sensing (AREA)
  • Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)

Abstract

The invention discloses a kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array, this method realized based on a kind of borehole wall imaging system based on cylinder ultrasonic phase array, the system comprises: ground controller (1), power supply pipe nipple (2), sonic system pipe nipple (4) and circuit pipe nipple (3);The described method includes: the depth conviction neural network of step 1) building and training for the classification of borehole wall defect;Step 2) acquires borehole wall defect sound echo-signal to be sorted;Borehole wall defect sound echo-signal to be sorted is inputted the trained depth conviction neural network for being used for the classification of borehole wall defect by step 3), obtains the classification of borehole wall defect.The method of the present invention is the defect identification method for combining wavelet packet decomposition algorithm and neural network method, and the discrimination of borehole wall defect can be improved, while shortening recognition time, improves recognition efficiency.

Description

A kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array
Technical field
The present invention relates to ultrasonic phase array borehole wall imaging detection technology fields, and in particular to one kind is based on cylinder ultrasound phase-control The borehole wall defect classifying identification method of battle array.
Background technique
In oil drilling and recovery process, casing guarantees that drilling well has been gone on smoothly in protection wellbore and Reinforcing Shaft Important function.But since subsurface environment is severe, casing damage quantity increases year by year, and detection and the protection of casing wall have become The important process of oil field development.Ultrasonic borehole wall imaging detection technology is a kind of common oil-well wall detection technique, passes through utilization Borehole wall reflection echo information, identifies a plurality of types of defects and well week situation, shows the borehole wall in the form of images clear and intuitively Situation and characteristic.
Start the 1990s to have gradually appeared supersonic imaging logger of new generation, the representative are The UBI (Ultra Sonic Imager) of the Schlumberger and CBIL (Circumferential of Baker Atlas Borehole Imaging Log) etc..It is rotary that these conventional borehole wall acoustic logging Image-forming instruments all use single-shot list to receive Sonic probe, when instrument well logging, transducer probe is rotated centered on borehole axis and is moved on borehole axis direction, to realize Looking into for helical manner is carried out to the borehole wall to sweep.However, this working method have the shortcomings that it is following several:
(1) signal-to-noise ratio of detectable signal is low.Due to the working method received using single-shot list, if encountered complicated or more Severe borehole condition, the useful signal amplitude detected are lower, it is difficult to identify, it is some exception shaft wall structures and it is fine change it is past It is past to be difficult to.
(2) resolution ratio detected is low.The acoustic beam of conventional transducer probe radiative acoustic wave is relatively wide, although concave surface Focused transducer can make sound beam focusing, but this focusing capability is very limited, and not only focal position is fixed, it is difficult to according to reality Border situation is adjusted and changes, and focused beam is elongated not enough, so that the image resolution ratio of detection data is lower.
(3) unstability of mechanical rotary scanning.Conventional transducers probe, which must be pivoted, to be just able to achieve to the borehole wall four Week looks into and sweeps imaging, and this mechanical rotation mode can bring the unstability of well logging and the problems such as relevant instrument maintenance. Machinery rotation can also carry out very burden for output transmission.
(4) logging speed is slower.Due to conventional transducer use mechanical type rotating, in order to guarantee borehole wall scanning imagery speed and Longitudinal resolution, logging speed are restricted, and reduce this instrument in the logging efficiency of underground, if well logging target zone compared with Thickness can occupy longer uphole time using this kind of instrument.
Currently, being still only able to detect for inner wall of the pipe detection and the research of the borehole wall image checking in oil drilling The presence of defect, can't type to defect on the borehole wall and property identify, i.e., the defect on the borehole wall cannot be determined Property analysis.
The research for carrying out defect classification and Flaw discrimination analysis for ultrasound detection using neural network is more and more,
In borehole wall image checking, we not only need to detect the presence of defect, but also need to detect defective Matter and classification need to know which type of defect is defect be, be hole, crackle or be mingled with, and need to know the shape of defect Shape and property etc..In a practical situation, due to the complex environment of underground, the borehole wall can deform, burn into is ruptured and perforated etc., it causes Great loss.Therefore, to borehole wall defect carry out qualitative analysis and type identification borehole wall situation can be obtained reliably detecting and Imaging results have important practical significance and economic value.
There are many methods, such as pattern-recognition and neural network for current Flaw discrimination analysis and type identification, but all It is to be analyzed and identified that effect is to be improved based on single transducer probe.In recent years, someone utilizes ultrasonic phase array The qualitative analysis and type identification research to defect are carried out, but using the ultrasonic phased array of plane.For in well The borehole wall defect effect that detection could obtain need to be carried out, so using cylinder ultrasonic phased array due to the shape of the borehole wall And about the research for carrying out qualitative analysis and identification to borehole wall defect using cylinder ultrasonic phased array, it is ground before without any Study carefully and reports.
Summary of the invention
It is an object of the invention to overcome the problems, such as to lack borehole wall defect classification method at present, in borehole wall image checking Flaw discrimination analysis and identification, propose it is a kind of to the borehole wall Flaw discrimination analysis method based on cylinder ultrasonic phase array, mainly Ultrasonic wave direction wellbore region domain radiation required for being converted electrical energy by cylinder ultrasound phase-control array 1 system, and receive and come from the borehole wall The scatter echo signal of defect carries out feature extraction to raw scattered echo by wavelet packet decomposition algorithm, obtains data volume pressure Signal after contracting, then Weight Training is carried out to three layer depth conviction neural networks using the characteristic signal extracted, finally with instruction The defect characteristic signal for the network handles detection perfected carries out Classification and Identification.
To achieve the goals above, the borehole wall defect Classification and Identification based on cylinder ultrasonic phase array that the invention proposes a kind of Method, a kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array, this method are based on a kind of based on cylinder ultrasound The borehole wall imaging system realization of phased array, the system comprises: ground controller 1, power supply pipe nipple 2, sonic system pipe nipple 4 and circuit are short Section 3;
The sonic system pipe nipple 4 includes cylinder ultrasonic phase array probe 11, and cylinder ultrasonic phase array probe 11 is by several A 6 circumferential array of transducer array element is at cylinder;
The circuit pipe nipple 3, the control signal for being sent according to ground controller 1 control cylinder ultrasonic phase array Each transducer array element 6 in probe 11 circumferentially realizes that the focus emission of ultrasonic signal and the focusing of echo-signal receive, and The echo data collected is sent to ground controller 1 after treatment;
The ground controller 1, for being imaged according to the downhole data received;
The power supply pipe nipple 2 is used to be system power supply;
The described method includes:
The depth conviction neural network of step 1) building and training for the classification of borehole wall defect;
Step 2) acquires borehole wall defect sound echo-signal to be sorted;
Borehole wall defect sound echo-signal to be sorted is inputted the trained depth for being used for the classification of borehole wall defect by step 3) Conviction neural network obtains the classification of borehole wall defect.
As a kind of improvement of the above method, the circuit pipe nipple 3 includes: digital circuit board and analog circuit board;
The analog circuit board includes: impulse ejection module, switch array module and reception gain control module;Described Phase control emission high pressure activation signal is sent to switch arrays for generating phase control emission high pressure activation signal by impulse ejection module Module;The switch array module is used for the transducer array element 6 of gated transmission ultrasonic signal and receives echo-signal;Described Reception gain control module is used for the echo-signal gain of modulation switch array module output;
The digital circuit board includes: CAN bus drive module, digital control processing module and AD acquisition and processing mould Block;The CAN bus drive module is communicated by CAN bus with ground controller 1;The digital control processing mould Block is used to parse the control command of the transmission of ground controller 1, and completes the control to analog circuit board and AD acquisition and processing module System;The AD acquisition and processing module carries out AD acquisition to the signal for receiving gain control module output, and AD is collected Echo data through digital control processing module carry out data processing after, ground control is sent to by CAN bus drive module Device 1.
As a kind of improvement of the above method, the transducer array element 6 in the cylinder ultrasonic phase array probe 11 is using filling Sealing encapsulating, the end of cylinder ultrasonic phase array probe 11 are provided with bearing shell 8, and the bearing shell 8 is equipped with electricity Pole pigtail splice 9, for connecting the electrode of transducer array element 6.
As a kind of improvement of the above method, the sonic system pipe nipple 4 further includes pressure balance mechanism 12 and pressure-bearing joint 10;The pressure balance mechanism 12 and pressure-bearing joint 10 is set to the both ends of cylinder ultrasonic phase array probe 11, and the cylinder is super The cavity for Silicone oil injection is offered in the middle part of sound phased array probe 11;
The pressure balance mechanism 12 includes: adapter 13, spring 16, piston 17 and piston steel cylinder 15;The work Plug 17 is arranged between adapter 13 and piston steel cylinder 15, which passes through adapter 13 and cylinder ultrasonic phase array probe 11 Cavity connection, be each filled with silicone oil in the two spaces where 17 both ends of piston;The spring 16, which supports, to be set to piston 17 and turns Between connector 13,11 external and internal pressures so that cylinder ultrasonic phase array is popped one's head in are moved by piston 17 and keep balance;
The both ends of the pressure-bearing joint 10 are respectively arranged with adapter 13 and sealing-plug 23, are respectively used to connection circuit pipe nipple 3 and bearing shell 8 on electrode leader connector 9.
As a kind of improvement of the above method, adapter 13 and sealing-plug 23 on the pressure-bearing joint 10 are all made of O shape Sealing ring sealing.
As a kind of improvement of the above method, the sealing-plug 23 is 81 core sealing-plugs.
As a kind of improvement of the above method, the sonic system pipe nipple 4 further includes mud velocity of sound energy converter, for measuring sound It is the velocity of sound of 4 local environment of pipe nipple.
As a kind of improvement of the above method, centralizer is provided on the outer wall of the power supply pipe nipple 2 and circuit pipe nipple 3 5;The centralizer 5 is supporting structure, and outer rim offsets with casing 7.
As a kind of improvement of the above method, the end of the sonic system pipe nipple 4 is additionally provided with pressure block 19, for realizing sonic system The Function Extension of pipe nipple 4.
As a kind of improvement of the above method, the step 1) is specifically included:
Step 1-1) building weight initialization depth conviction neural network;
Step 1-2) establish training set, the training set includes several borehole wall defect original echoed signals and its corresponding Defect classification;
Step 1-3) feature extraction is carried out to the borehole wall defect sound echo-signal of training set, obtain reconstruction signal and feature Amount;
Four layers of WAVELET PACKET DECOMPOSITION are carried out using borehole wall defect sound echo-signal of the Mallat decomposition algorithm to different defects, are obtained To the 4th layer of wavelet packet coefficient, signal characteristic is extracted, the reconstruction signal and characteristic quantity of borehole wall defect are obtained;
Step 1-4) according to reconstruction signal and characteristic quantity is obtained, depth conviction neural network is trained, is trained Good depth conviction neural network.
Present invention has an advantage that
1, detection speed is fast, and phased-array technique is rotated by the way of electronic system control acoustic beam rotation instead of machinery, So that logging speed is improved;
2, the signal-to-noise ratio of detection is improved, phase array focusing emission characteristics can be such that the sound field intensity at target greatly increases, make The reflection echo amplitude for obtaining target is reinforced;Moreover, phase array focusing reception processing can further improve the strong of target echo signal Degree, therefore, focus emission can greatly improve the signal-to-noise ratio of target acquisition with reception technique is focused;
3, image resolution ratio and detection accuracy are improved.The sound beam focusing characteristic of ultrasonic phased array technology can form relatively fine Acoustic beam, generate acoustic beam reach 3dB width, more much smaller than other conventional instrument beam widths, such acoustic beam can be significantly Improve the resolution ratio and detection accuracy of detection target image;
4, when detecting to a certain range of wellhole, no replacement is required pops one's head in, due to the focus of phase array focusing acoustic beam It is adjustable, for different well diameters, the focus emission and reception delay of probe need to be only adjusted, acoustic beam can be realized in the poly- of the borehole wall Coke, so without frequently replacement well-logging probe;
5, the defect identification method that method of the invention combines wavelet packet decomposition algorithm and neural network method, can be with The discrimination of borehole wall defect is improved, while shortening recognition time, improves recognition efficiency.
Detailed description of the invention
Fig. 1 is the borehole wall imaging system external structure schematic diagram in the present invention;
Fig. 2 is the borehole wall imaging system schematic diagram of internal structure in the present invention;
Fig. 3 a is the cylinder ultrasonic phase array structural schematic diagram in the present invention;
Fig. 3 b is cylinder ultrasonic phase array top view shown in Fig. 3 a;
Fig. 4 is the cylinder ultrasonic phase array probe schematic perspective view in the present invention;
Fig. 5 is the schematic diagram of internal structure of the sonic system pipe nipple in the present invention;
Fig. 6 is the pressure balance mechanism schematic diagram of internal structure in sonic system pipe nipple provided by the invention;
Fig. 7 is the pressure-bearing joint schematic diagram of internal structure in sonic system pipe nipple provided by the invention;
Fig. 8 is the flow chart of the borehole wall defect classifying identification method of the invention based on cylinder ultrasonic phase array;
Fig. 9 is the structural schematic diagram of depth conviction neural network of the invention.
Appended drawing reference
1, ground controller 2, power supply pipe nipple 3, circuit pipe nipple
4, sonic system pipe nipple 5, centralizer 6, transducer array element
7, casing 8, bearing shell 9, electrode leader connector
10, pressure-bearing joint 11, cylinder ultrasonic phase array probe 12, pressure balance mechanism
13, adapter 14, oiling screw 15, piston steel cylinder
16, spring 17, piston 18, overflow valve
19, pressure block 20, the first O-ring seals 21, feather key
22, Rapid-turning nut 23, sealing-plug 24, the second O-ring seals
Specific embodiment
The present invention will be described in detail in the following with reference to the drawings and specific embodiments.
A kind of borehole wall imaging system overall structure based on cylinder ultrasonic phase array provided by the invention is as shown in Figure 1, packet Include: ground control is constituted with imaging system ground controller, power supply pipe nipple, circuit pipe nipple, sonic system pipe nipple and external mechanical part. The sonic system pipe nipple 4 is popped one's head in including cylinder ultrasonic phase array, which pops one's head in by several transducer array elements Circumferential array is at cylinder;The control signal that the circuit pipe nipple 3 is sent according to ground controller 1 controls cylinder ultrasound phase Each transducer array element in control battle array probe circumferentially realizes that the focus emission of ultrasonic signal and the focusing of echo-signal receive, and The echo data collected is sent to ground controller 1;The ground controller 1 according to the echo data received into Row imaging.
Borehole wall imaging system based on above structure needs in system work process by power supply pipe nipple 2, circuit pipe nipple 3 It is threaded through in casing with sonic system pipe nipple 4, and is moved along casing to underground.The all parts of the system are illustrated below:
1) it ground control and processing display system: is mainly communicated by CAN bus with underground equipment, for being realized on PC To the parameter setting of underground equipment, the preservation and imaging of system failure investigation and Echo Processing data, are main man-machine friendships Mutual approach.
2) power supply pipe nipple 2: including digital power plate, simulated high-pressure power panel and analog low voltage electric power plate as shown in Fig. 2, It is mainly used for voltage needed for being converted to subsequent conditioning circuits at different levels from ground service cable voltage.Wherein digital power plate is to circuit The digital circuit board of pipe nipple is powered, and simulated high-pressure power panel and analog low voltage electric power plate are supplied to the analog circuit board of circuit pipe nipple Electricity.
3) it circuit pipe nipple 3: is made of digital processing circuit plate and analog circuit board.
As shown in Fig. 2, the digital circuit board includes CAN bus drive module, digital control processing module, AD acquisition With processing module.The CAN bus drive module is communicated by CAN bus with ground controller;The number control Processing module processed is used to parse the control command of ground controller transmission, and completes to impulse ejection module, switch array module, connects Receive the control of gain control module, AD acquisition and processing module;Its concrete function has: explaining the various lives that ground controller issues The control enable, complete transmit circuit fire pulse width, emitting array element, emission delay, is completed at the same time the automatic control of signal gain System, signal acquisition, different preliminary treatment is carried out to acquisition data according to task and is further processed, treated result It is sent to communication module.Wherein preliminary treatment includes being delayed to collected original echoed signals according to phase array focusing rule Superposition calculation, and envelope is taken using Hilbert transform pairs superposition back echo signal;It is further processed the envelope letter referred to acquisition It number carries out peak value judgement and extracts amplitude data, then these amplitude datas are subjected to splicing to obtain final required imaging number According to.The AD acquisition and processing module carries out AD acquisition to the signal for receiving gain control module output, and AD is collected Echo data through digital control processing module carry out data processing after, ground control is sent to by CAN bus drive module Device.
As shown in Fig. 2, the analog circuit board includes impulse ejection module, switch array module, reception gain control Module.The impulse ejection module is for generating phase control emission high pressure activation signal, and by phase control emission high pressure activation signal It is sent to switch array module;The switch array module is used for the energy converter of gated transmission ultrasonic signal and receives echo-signal Array element;The reception gain control module is used for the echo-signal gain of modulation switch array module output.By increasing to automatic The control of beneficial control module, the gain variation range for receiving signal reach 60dB-12dB~48dB, can be to 10mV~10V model Signal in enclosing amplifies and undistorted.
Switch arrays are a functional modules in analog circuit board, and the high-voltage switch gear with 16 autonomous channels can be used Battle array chip MAX4968A carries out the switching between transmission channel and transducer array element and connects, to achieve the purpose that channel multiplexing.It is main It is used for the connection according to control signal gating ultrasound emission receiving channel and transducer array element.Totally two pieces of analog circuit boards, often There is two panels switch arrays chip on block plank, totally four switch arrays chips.System can design 8 transmitting receiving channels, Mei Getong Road is connected with 8 phased array elements respectively by switch arrays chip, to realize 8 channel multiplexings to 64 array element.
Borehole wall imaging system of the invention realizes that energy converter emits array element at work, by being switched fast for electronic switch Switching, to form focused beam along well Zhou Xuanzhuan scanning, need to only be moved up and down by instrument along borehole axis can be realized pair Comprehensive scanning of the borehole wall detects.
4) sonic system pipe nipple 4: including mud velocity of sound energy converter, cylinder ultrasonic phase array probe 11, pressure balance mechanism 12 and Pressure-bearing joint 10.As shown in figure 5, the pressure balance mechanism 12 and pressure-bearing joint 10 is set to cylinder ultrasonic phase array probe 11 both ends.
The mud velocity of sound energy converter is used for the velocity of sound of real-time measurement ultrasonic wave local environment.
As shown in Fig. 3 a, 3b, the cylinder ultrasonic phase array probe 11 is arranged in cylinder by several transducer array elements 6 Shape.Well liquid is full of in casing 7, cylinder ultrasonic phase array probe 11 is placed in 7 center of casing with sonic system pipe nipple, short by circuit The signal transmitting of control circuit control each activity array element of phased array probe of section and reception delay, to realize that ultrasonic phase array is poly- Burnt scanning imagery.Phased acoustic beam angle and focal position etc. can be continuous dynamically adjustable in a certain range, and focusing, reception etc. is more Kind signal processing technology is used for imaging, also can further improve the image resolution ratio of ultrasonic phase array detection.
Wafer energy transducer array element is potted in inside non-metal shell by the present invention by casting glue, realizes integral type encapsulation Probe, probe middle part offers cavity, for pouring into silicone oil, is then sealed by O-ring seal, with traditional non-Unitary gold Belong to shell probe to compare, this encapsulating structure reduces the reflection and decaying popped one's head in signal.As shown in figure 4, sonde configuration sheet Body is that there are a bearing shells 8, positioned at the end of probe.The bearing shell 8 is equipped with electrode leader connector 9, is used for Connect the electrode of transducer array element.
In the specific implementation process, the major parameter of cylinder ultrasonic phase array probe may be designed as: array number 64, outer diameter 80mm, 150 DEG C of heatproof, pressure-resistant 100Mpa.There are bearing course, probes to connect with front and back mechanical structure and use O shape close for itself Sealing is realized in seal.Each array element chip in the cylinder ultrasonic phase array probe all has internal loopback function, for real The transmitting and reception of existing ultrasound phase-control signal.
As shown in fig. 6, the pressure balance mechanism include: 13, two oiling screws 14 of adapter, piston steel cylinder 15, Spring 16, piston 17, overflow valve 18, pressure block 19, Multiple Type O-ring seals.12 core pressure-bearings can be used in the pressure block 19 Block is set to the end of sonic system pipe nipple 4, for realizing the Function Extension of sonic system pipe nipple 4.Two oiling screws 14 difference In the space at 17 both ends of piston.
Piston balance mechanism is used to adjust the probe chip both sides fluid pressure balance being immersed in well liquid, realizes total pressure Balance.Piston balance mechanism realizes internal and external pressure balance by a piston movement, and when well liquid pressure is big, piston is mobile to be squeezed Piston liquid in pipe is pressed, rises to intraductal pressure identical as external pressure, to realize the pressure balance on both sides.
The piston 17 is arranged between adapter 13 and piston steel cylinder 15, which passes through adapter 13 and cylinder The cavity of ultrasonic phase array probe is connected to, and is each filled with silicone oil in the two spaces where 17 both ends of piston;The spring 16 supports Between piston 17 and adapter 13, moved by piston 17 so that cylinder ultrasonic phase array probe external and internal pressure keeps flat Weighing apparatus;
In order to balance pressure, before instrument is gone into the well, it is first turned on the oiling popped one's head in space close to cylinder ultrasonic phase array Screw 14 injects a certain amount of silicone oil into probe cavity;Then need to open the oiling in another space far from probe Screw 14 injects silicone oil in instrument internal, then spring 16 is compressed, and piston 17 starts to the left using electronic or manual pump Mobile, when the mobile leftmost side of piston, overflow valve begins to overflow, so far completes oil injection operation.
After system is gone into the well, with the variation of temperature and depth, piston 17 will be mobile to the small direction of pressure, until pressure Until dynamic balance.When piston 17 is located at the leftmost side, external pressure is still greater than in internal pressure, and overflow valve 18 starts, and starts pressure release.Due to visiting Head among piston balance mechanism and pressure-bearing joint, inside fill silicone oil, when external pressure causes probe deformation occurs when, Dummy piston will happen movement according to deformation, to maintain probe chip both sides fluid pressure balance.
The pressure-bearing joint realizes sonic system part immersion oil, while guaranteeing electricity for connecting sonic system pipe nipple and circuit pipe nipple The sealing and compressive property of short out section.Because phased array probe is directly immersed in the highly pressurised liquid of wellhole, and circuit system is located at In metal coating shell, belong to normal pressure system, there are huge pressure difference between two spaces, needs to bear sonic system by pressure-bearing joint Pipe nipple bring high pressure.
As shown in fig. 7, pressure-bearing joint 10 includes: adapter 13, first O-ring seals 2063.09mm × 3.53mm, guiding Key 21, Rapid-turning nut 22, second O-ring seals 2455.25mm × 2.65mm, sealing-plug 23.The both ends of the pressure-bearing joint 10 It is respectively arranged with adapter 13 and sealing-plug 23, the electrode leader connector being respectively used on connection circuit pipe nipple 3 and bearing shell 8 9.The screw thread of the Rapid-turning nut 22 at pressure-bearing joint both ends is connect with circuit pipe nipple and sonic system pipe nipple respectively, to realize circuit pipe nipple With the Hard link of sonic system pipe nipple, while silicone oil being isolated in except circuit pipe nipple by sealing-plug 23.On the pressure-bearing joint 10 Adapter 13 sealed using the first O-ring seals 20, the sealing-plug 23 can be used 81 core sealing-plugs, and pass through the 2nd O shape Sealing ring 24 seals.
5) external mechanical part: skeleton structure and centralizer including each pipe nipple.
The skeleton structure of each pipe nipple mainly for underground high temperature and pressure particular surroundings demand and design.It is needed for resistance to compression It asks, sonic system pipe nipple uses immersion oil mode, resistance to compression is realized by pressure balance, material is heat-resisting material, meets resistance to height Warm demand.Other circuit frameworks are placed in special voltage holding circuit cylinder, are linked by pressure-bearing joint and sonic system pipe nipple.
The centralizer is independent bracket component, as shown in Figure 1, the outer wall of the power supply pipe nipple 2 and circuit pipe nipple 3 On be provided with centralizer 5;Its outer rim offsets with casing.The centralizer is detachable, and casing suitable position is fixed to when going into the well It sets.Centralizer is for ensuring that entire mechanical structure is always positioned at the center of oil pipe, it is ensured that phased array imaging effect.
Based on above-mentioned imaging system, as shown in figure 8, the present invention also provides a kind of boreholes wall based on cylinder ultrasonic phase array Defect classifying identification method, which comprises
The depth conviction neural network of step 1) building and training for the classification of borehole wall defect;It specifically includes:
Step 1-1) building weight initialization depth conviction neural network;
Step 1-2) establish training set, the training set includes several borehole wall defect original echoed signals and its corresponding Defect classification;
Choose the these types of defect type such as flat-bottom hole, oblique through hole and the through-holes of various shapes such as round, rectangular and triangle;
Step 1-3) feature extraction is carried out to the borehole wall defect sound echo-signal of training set, obtain reconstruction signal and feature Amount;
During WAVELET PACKET DECOMPOSITION, the decomposition of high and low frequency component can be decomposed into a very delicate water It is flat, and the signal in N layers can be decomposed, to obtain the characteristic information of the signal in frequency band, energy, that is, quilt of original signal Decompose 2NA orthogonal frequency band, energy summation of the signal in each frequency range answered with the energy of original echoed signals it is identical, every Characteristic information in the frequency range for the original echoed signals that signal in a frequency band indicates.It is decomposed in this method using Mallat Algorithm carries out four layers of WAVELET PACKET DECOMPOSITION to the ultrasonic echos of different defects, obtain the 4th layer of wavelet packet coefficient, i.e., and the 4th layer 16 node coefficients.
By cylinder ultrasound phase-control array 1 system to borehole wall radiated sound field, and receive the original sound echo from borehole wall defect Signal intercepts the time-domain signal of 200 sampled points, carries out four layers of small echo to original signal by Mallat wavelet packet decomposition algorithm Packet decomposes and extracts signal characteristic, obtains the reconstruction signal and characteristic quantity of borehole wall defect.
Step 1-4) according to reconstruction signal and characteristic quantity is obtained, depth conviction neural network is trained;
Deepness belief network model is one by the cumulative of multiple RBM (Restricted Boltzmann Machine) RBM includes a hidden layer and a visible layer, there is bi-directional chaining, each unit inside same layer between hidden layer and visible layer unit Between without interconnection.
One typical deepness belief network, is equivalent to a series of stacking for RBM, can also regard as by many Stochastic variable forms directed acyclic graph, as shown in figure 9, these networks are divided into a series of RBM, they are by visible layer and imply Layer composition, implicit layer unit are trained to capture the correlation of the high level data showed in visible layer.In sorter network, Need to be added tag unit when the RBM training of top layer, corresponding tag unit, which is opened, is set as 1, and other be then closed sets It is 0.
Using the borehole wall defect characteristic amount extracted, the depth conviction neural network for having constructed weight initialization is weighed Value training and defect class test.
Step 2) acquires borehole wall defect sound echo-signal to be sorted;
Using the borehole wall imaging system of above-mentioned cylinder ultrasonic phase array, focus is arranged on the borehole wall, utilizes cylinder ultrasound Phased array emits ultrasonic pulsative signal, and receives and acquire echo-signal, obtains borehole wall flaw echoes to be sorted.
Depth conviction neural network after borehole wall defect sound echo-signal input training to be sorted is obtained well by step 3) The classification of wall defect.
Above-described specific embodiment has carried out further the purpose of the present invention, technical scheme and beneficial effects It is described in detail, it should be understood that being not intended to limit the present invention the foregoing is merely a specific embodiment of the invention Protection scope, all within the spirits and principles of the present invention, any modification, equivalent substitution, improvement and etc. done should all include Within protection scope of the present invention.

Claims (10)

1. a kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array, this method is based on a kind of based on cylinder ultrasound The borehole wall imaging system of phased array realizes, the system comprises: ground controller (1), power supply pipe nipple (2), sonic system pipe nipple (4) and Circuit pipe nipple (3);
The sonic system pipe nipple (4) include cylinder ultrasonic phase array probe (11), the cylinder ultrasonic phase array pop one's head in (11) if by Dry transducer array element (6) circumferential array is at cylinder;
The circuit pipe nipple (3), the control signal for being sent according to ground controller (1), controls cylinder ultrasonic phase array Each transducer array element (6) in probe (11) circumferentially realizes that the focusing of the focus emission and echo-signal of ultrasonic signal connects It receives, and the echo data collected is sent to ground controller (1) after treatment;
The ground controller (1), for being imaged according to the downhole data received;
The power supply pipe nipple (2) is used to be system power supply;
The described method includes:
The depth conviction neural network of step 1) building and training for the classification of borehole wall defect;
Step 2) acquires borehole wall defect sound echo-signal to be sorted;
Borehole wall defect sound echo-signal to be sorted is inputted the trained depth conviction for being used for the classification of borehole wall defect by step 3) Neural network obtains the classification of borehole wall defect.
2. the borehole wall defect classifying identification method according to claim 1 based on cylinder ultrasonic phase array, which is characterized in that The circuit pipe nipple (3) includes: digital circuit board and analog circuit board;
The analog circuit board includes: impulse ejection module, switch array module and reception gain control module;The pulse Phase control emission high pressure activation signal is sent to switch arrays mould for generating phase control emission high pressure activation signal by transmitting module Block;The switch array module is used for the transducer array element (6) of gated transmission ultrasonic signal and receives echo-signal;Described Reception gain control module is used for the echo-signal gain of modulation switch array module output;
The digital circuit board includes: CAN bus drive module, digital control processing module and AD acquisition and processing module; The CAN bus drive module is communicated by CAN bus with ground controller (1);The digital control processing mould Block is used to parse the control command of ground controller (1) transmission, and completes to analog circuit board and AD acquisition and processing module Control;The AD acquisition and processing module carries out AD acquisition to the signal for receiving gain control module output, and AD is acquired The echo data arrived is sent to ground control after digital control processing module carries out data processing, through CAN bus drive module Device (1) processed.
3. the borehole wall defect classifying identification method according to claim 1 based on cylinder ultrasonic phase array, which is characterized in that Transducer array element (6) in the cylinder ultrasonic phase array probe (11) uses casting glue encapsulating, which visits The end of head (11) is provided with bearing shell (8), and the bearing shell (8) is equipped with electrode leader connector (9), for connecting The electrode of transducer array element (6).
4. the borehole wall defect classifying identification method according to claim 3 based on cylinder ultrasonic phase array, which is characterized in that The sonic system pipe nipple (4) further includes pressure balance mechanism (12) and pressure-bearing joint (10);The pressure balance mechanism (12) The both ends of cylinder ultrasonic phase array probe (11) are set to pressure-bearing joint (10), which pops one's head in the middle part of (11) Offer the cavity for Silicone oil injection;
The pressure balance mechanism (12) includes: adapter (13), spring (16), piston (17) and piston steel cylinder (15);Institute The piston (17) stated is arranged between adapter (13) and piston steel cylinder (15), which passes through adapter (13) and cylinder The cavity of ultrasonic phase array probe (11) is connected to, and is each filled with silicone oil in the two spaces where piston (17) both ends;The bullet Spring (16), which supports, to be set between piston (17) and adapter (13), is moved by piston (17) so that cylinder ultrasonic phase array is popped one's head in (11) external and internal pressure keeps balance;
The both ends of the pressure-bearing joint (10) are respectively arranged with adapter (13) and sealing-plug (23), and it is short to be respectively used to connection circuit Save the electrode leader connector (9) on (3) and bearing shell (8).
5. the borehole wall defect classifying identification method according to claim 4 based on cylinder ultrasonic phase array, which is characterized in that Adapter (13) and sealing-plug (23) on the pressure-bearing joint (10) are all made of O-ring seals sealing.
6. the borehole wall defect classifying identification method according to claim 5 based on cylinder ultrasonic phase array, which is characterized in that The sealing-plug (23) is 81 core sealing-plugs.
7. the borehole wall defect classifying identification method according to claim 1 based on cylinder ultrasonic phase array, which is characterized in that The sonic system pipe nipple (4) further includes mud velocity of sound energy converter, for measuring the velocity of sound of sonic system pipe nipple (4) local environment.
8. the borehole wall defect classifying identification method according to claim 1 based on cylinder ultrasonic phase array, which is characterized in that Centralizer (5) are provided on the outer wall of the power supply pipe nipple (2) and circuit pipe nipple (3);The centralizer (5) is bracket knot Structure, outer rim offset with casing (7).
9. the borehole wall defect classifying identification method according to claim 1 based on cylinder ultrasonic phase array, which is characterized in that The end of the sonic system pipe nipple (4) is additionally provided with pressure block (19), for realizing the Function Extension of sonic system pipe nipple (4).
10. the borehole wall defect classifying identification method according to claim 1 based on cylinder ultrasonic phase array, feature exist In the step 1) specifically includes:
Step 1-1) building weight initialization depth conviction neural network;
Step 1-2) training set is established, the training set includes several borehole wall defect original echoed signals and its corresponding defect Classification;
Step 1-3) feature extraction is carried out to the borehole wall defect sound echo-signal of training set, obtain reconstruction signal and characteristic quantity;
Four layers of WAVELET PACKET DECOMPOSITION are carried out to the borehole wall defect sound echo-signals of different defects using Mallat decomposition algorithm, obtain the Four layers of wavelet packet coefficient extracts signal characteristic, obtains the reconstruction signal and characteristic quantity of borehole wall defect;
Step 1-4) according to reconstruction signal and characteristic quantity is obtained, depth conviction neural network is trained, is obtained trained Depth conviction neural network.
CN201710589629.8A 2017-07-19 2017-07-19 A kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array Pending CN109281653A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710589629.8A CN109281653A (en) 2017-07-19 2017-07-19 A kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710589629.8A CN109281653A (en) 2017-07-19 2017-07-19 A kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array

Publications (1)

Publication Number Publication Date
CN109281653A true CN109281653A (en) 2019-01-29

Family

ID=65184336

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710589629.8A Pending CN109281653A (en) 2017-07-19 2017-07-19 A kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array

Country Status (1)

Country Link
CN (1) CN109281653A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110988981A (en) * 2019-12-23 2020-04-10 山东大学 Phased array sound wave advanced prediction system and method suitable for drilling and blasting method tunnel
CN115680618A (en) * 2021-07-29 2023-02-03 中国石油化工股份有限公司 Array type multi-frequency sound wave oil-gas well casing visualization method and detection device

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1112639A (en) * 1994-05-26 1995-11-29 魏营隆 Drill bit central coordinate ultrasonic wave real-time measurement while drilling for large driller
CN2613758Y (en) * 2003-03-11 2004-04-28 中国石油天然气集团公司 Acoustic system for high-temp. high-pressure and multi-parameter ultrasonic engineering well logging instrument
CN2627214Y (en) * 2003-03-28 2004-07-21 中国石油天然气集团公司 Multiple parameter ultrasonic engineering logging device
CN101493438A (en) * 2009-02-18 2009-07-29 宁波工程学院 Phased array ultrasonic detection, data acquisition and process device
CN202031567U (en) * 2011-01-05 2011-11-09 中国海洋石油总公司 Circuit used for measuring annulus instantaneous flow in marine riser
CN102877827A (en) * 2012-10-10 2013-01-16 中国石油天然气集团公司 Acoustic system module of acoustic logging while drilling
CN104330480A (en) * 2013-07-22 2015-02-04 中国科学院声学研究所 Fault diagnosis method for ultrasonic phased array imaging detector
CN204283410U (en) * 2014-11-28 2015-04-22 中国石油天然气集团公司 A kind of transmitting sonic system of multipolar array acoustic tool PEEK material package
CN205117265U (en) * 2015-11-19 2016-03-30 北京美高科技发展有限公司 Cable formula electrohydraulic control cuts piping tool
CN206091977U (en) * 2016-08-22 2017-04-12 西安石竹能源科技有限公司 Four multi -functional arm hole diameters

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1112639A (en) * 1994-05-26 1995-11-29 魏营隆 Drill bit central coordinate ultrasonic wave real-time measurement while drilling for large driller
CN2613758Y (en) * 2003-03-11 2004-04-28 中国石油天然气集团公司 Acoustic system for high-temp. high-pressure and multi-parameter ultrasonic engineering well logging instrument
CN2627214Y (en) * 2003-03-28 2004-07-21 中国石油天然气集团公司 Multiple parameter ultrasonic engineering logging device
CN101493438A (en) * 2009-02-18 2009-07-29 宁波工程学院 Phased array ultrasonic detection, data acquisition and process device
CN202031567U (en) * 2011-01-05 2011-11-09 中国海洋石油总公司 Circuit used for measuring annulus instantaneous flow in marine riser
CN102877827A (en) * 2012-10-10 2013-01-16 中国石油天然气集团公司 Acoustic system module of acoustic logging while drilling
CN104330480A (en) * 2013-07-22 2015-02-04 中国科学院声学研究所 Fault diagnosis method for ultrasonic phased array imaging detector
CN204283410U (en) * 2014-11-28 2015-04-22 中国石油天然气集团公司 A kind of transmitting sonic system of multipolar array acoustic tool PEEK material package
CN205117265U (en) * 2015-11-19 2016-03-30 北京美高科技发展有限公司 Cable formula electrohydraulic control cuts piping tool
CN206091977U (en) * 2016-08-22 2017-04-12 西安石竹能源科技有限公司 Four multi -functional arm hole diameters

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
施成龙等: "利用深度神经网络和小波包变换进行缺陷类型分析", 《声学学报》 *
李健等: "超声相控阵检测CFRP缺陷识别方法", 《天津大学学报》 *
董晗等: "石油钻井中超声相控阵井壁成像检测技术研究", 《机械工程学报》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110988981A (en) * 2019-12-23 2020-04-10 山东大学 Phased array sound wave advanced prediction system and method suitable for drilling and blasting method tunnel
CN110988981B (en) * 2019-12-23 2021-09-14 山东大学 Phased array sound wave advanced prediction system and method suitable for drilling and blasting method tunnel
CN115680618A (en) * 2021-07-29 2023-02-03 中国石油化工股份有限公司 Array type multi-frequency sound wave oil-gas well casing visualization method and detection device

Similar Documents

Publication Publication Date Title
CN109281652A (en) A kind of borehole wall imaging system based on cylinder ultrasonic phase array
AU2011382521B2 (en) Acoustic transducer apparatus, systems, and methods
BR112021000838A2 (en) EVALUATION OF CEMENT THROUGH PIPING WITH THE USE OF SEISMIC METHODS
CN101782553B (en) System for detecting thickness and corrosion of mineshaft of gas storage well
CN103777247B (en) Transient electromagnetic radar detection system and detection method
US20150338378A1 (en) Ultrasonic signal time-frequency decomposition for borehole evaluation or pipeline inspection
WO2015105977A1 (en) Devices and methods for downhole acoustic imaging
BR112020009277A2 (en) well profiling excitation optimizer by oriented wave attenuation based on waveform modeling
CN102967657A (en) Nondestructive testing method based on synthetic aperture ultrasonic imaging technique
CA2632759A1 (en) Programmable data acquisition for tubular objects
CN109281650A (en) A kind of borehole wall real time imagery method based on cylinder ultrasonic phase array
NO343125B1 (en) Method and apparatus for detecting echo maximum when logging acoustic images of wellbore feeding tubes
CN105445371B (en) The ultrasonic phase array fast checking device and detection method of electron beam weld
CN101122228A (en) Down-hole forward looking phase controlled sound wave imaging method and imaging device
CN107907597A (en) A kind of penetration type ultrasonic wave soil moisture content test device and method
CN104749253A (en) Ultrasonic back scattering imaging method and device for inner defects of cylindrical workpiece
US10718865B2 (en) Method of compressing beamformed sonar data
CN105888647A (en) Calibration method and device of acoustic variable density logging instrument
CN109281653A (en) A kind of borehole wall defect classifying identification method based on cylinder ultrasonic phase array
CN106501285B (en) The equipment and detection method of the mud jacking compactness of non-destructive testing prestress pipe
CN109115892A (en) Tube wall detection device, detection system and the detection method of vertical pipeline
US6002639A (en) Sensor configuration for nulling reverberations to image behind reflective layers
Lu et al. Azimuthally acoustic logging tool to evaluate cementing quality
CN108508093A (en) A kind of detection method and system of workpiece, defect height
CN206756749U (en) The data acquisition device of bottom sediment original position acoustic measurement system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
CB03 Change of inventor or designer information
CB03 Change of inventor or designer information

Inventor after: Chen Shiwen

Inventor after: Zhang Bixing

Inventor after: Yan Shouguo

Inventor after: Huang Juan

Inventor before: Chen Shiwen

Inventor before: Zhang Bixing

RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20190129