Orifice wear detection method and device for differential pressure flowmeter based on machine vision

Abstract

The invention discloses a method and device for detecting orifice wear of a differential pressure flowmeter based on machine vision. The method comprises: obtaining a physical image coordinate system and a point cloud image of the orifice of the differential pressure flowmeter, denoising the coordinate points of the physical image coordinate system and the coordinate points of the point cloud image, obtaining a physical image optimized coordinate system and a point cloud optimized image, respectively calculating and comparing the noise ratio of the physical image optimized coordinate system and the noise ratio of the point cloud optimized image, taking the coordinate points with large noise ratio as the point cloud data of the orifice of the differential pressure flowmeter, splicing and fusing the point cloud data to establish a three-dimensional model of the orifice of the differential pressure flowmeter after wear, and finally performing stress analysis on the three-dimensional model. By finding the abscissa value of the corresponding maximum stress on the ordinate of the S-N curve as the number of cycles of the orifice of the differential pressure flowmeter under water impact, the invention can realize high-precision automatic detection of the orifice of the differential pressure flowmeter, with high detection accuracy and high efficiency.

Description

Technical Field

[0001] The invention belongs to the technical field of oil production equipment component measurement, and in particular relates to a differential pressure flow meter orifice wear detection method and a detection device based on machine vision. Background Art

[0002] The differential pressure flowmeter orifice is the most critical component in the precise regulation of water pressure. During the water injection process, it is easy to cause the differential pressure flowmeter orifice to wear due to the alternating effect of water pressure for a long time. The wear of the differential pressure flowmeter orifice will cause the water pressure to be unable to be accurately controlled, making it impossible to drive crude oil into the production pipeline, which directly affects the oil production.

[0003] At present, the wear detection of the orifice plate of the differential pressure flowmeter is mainly carried out through manual measurement. This measurement method is easily affected by the instrument itself and human subjective judgment, and the measurement accuracy is not high. Therefore, how to realize the intelligent detection of the orifice plate wear of the differential pressure flowmeter is an urgent problem to be solved. Summary of the invention

[0004] In order to solve the above problems, the purpose of the present invention is to provide a method and device for detecting wear of the orifice of a differential pressure flowmeter based on machine vision, by acquiring the point cloud data of the orifice of the differential pressure flowmeter, reconstructing a three-dimensional model of the orifice of the differential pressure flowmeter according to the point cloud data, and then performing stress analysis on the three-dimensional model to obtain the number of cycles of the orifice of the differential pressure flowmeter, so as to solve the problem of low accuracy of existing measurement methods.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A method for detecting orifice plate wear of a differential pressure flowmeter based on machine vision comprises the following steps:

[0007] Step S1: converting the world coordinate system into a CCD (Charge-coupled Device) visual sensor coordinate system, and obtaining the physical image coordinate system of the differential pressure flowmeter orifice plate through the CCD visual sensor coordinate system;

[0008] The orifice plate image of the differential pressure flowmeter is collected by a CCD visual sensor to obtain a point cloud image of the orifice plate of the differential pressure flowmeter;

[0009] Step S2: performing denoising processing on the coordinate points of the physical image coordinate system and the coordinate points of the point cloud image to obtain the physical image optimized coordinate system and the point cloud optimized image respectively;

[0010] Step S3: respectively calculating the noise ratio of the physical image optimized coordinate system and the noise ratio of the point cloud optimized image, and comparing the noise ratios of the two, taking the coordinate point of the physical image optimized coordinate system or the coordinate point of the point cloud optimized image with a larger noise ratio as the point cloud data of the orifice plate of the differential pressure flowmeter;

[0011] Step S4: performing point cloud fusion after splicing the point cloud data to establish a three-dimensional model of the differential pressure flowmeter orifice plate after wear;

[0012] Step S5: Perform stress analysis on the three-dimensional model to obtain the maximum stress of the three-dimensional model, and find the corresponding maximum stress on the ordinate of the SN curve. The abscissa value of the SN curve corresponding to the maximum stress is the number of cycles of the differential pressure flowmeter orifice plate.

[0013] Furthermore, in step S1, the world coordinate system is converted into the CCD vision sensor coordinate system, specifically: by using the rotation matrix R and the translation matrix t The world coordinate system is converted into the CCD vision sensor coordinate system as shown in the following formula: ;

[0014] Among them, the rotation matrix R is a 3×3 orthogonal matrix, the translation matrix t is a 3×1 translation vector, T is the transposed vector, P W is a point in the world coordinate system, X W 、Y W 、Z W for P W The coordinates of P C is a point in the CCD vision sensor coordinate system, X C 、Y C 、Z C for P C The coordinates of

[0015] In the step S1, the orifice plate image of the differential pressure flowmeter is collected by a CCD visual sensor to obtain a point cloud image of the orifice plate of the differential pressure flowmeter, as shown in the following formula:

[0016] ;

[0017] in, is the homogeneous coordinate system of the physical image coordinate system, f x , f y is the focal length of the CCD vision sensor coordinate system, c x 、c yIt is the coordinate origin of the physical image coordinate system of the differential pressure flowmeter orifice plate.

[0018] Furthermore, in step S2, the denoising process is specifically: using a statistical filtering method to remove coordinate noise points in the physical image coordinate system or noise points in the point cloud image.

[0019] Further, in step S4, the point cloud data are spliced ​​and then point cloud fused, specifically: the point cloud data that does not correspond in the point cloud data of the orifice plate of the differential pressure flowmeter is eliminated using an error point elimination algorithm to obtain the remaining point cloud data, the remaining point cloud data is spliced ​​using the least squares method to obtain spliced ​​data, and the spliced ​​data is point cloud fused.

[0020] Further, in the step S4, the point cloud fusion is performed by Cloud Compare software;

[0021] In the step S5, stress analysis is performed on the three-dimensional model using ANSYS software.

[0022] The present invention also provides a pressure differential flowmeter orifice wear detection device based on machine vision, and applies the pressure differential flowmeter orifice wear detection method based on machine vision of the present invention. The detection device includes a host computer, a linear motion module, a two-dimensional linear motion table, a support platform and a point cloud image acquisition mechanism. The linear motion module and the two-dimensional linear motion table are both arranged on the support platform. The point cloud image acquisition mechanism is arranged on the linear motion module and is spaced relative to the two-dimensional linear motion table in an upper and lower interval. The linear motion module drives the point cloud image acquisition mechanism to move up and down to acquire a point cloud image of the pressure differential flowmeter orifice arranged on the two-dimensional linear motion table. The point cloud image acquisition mechanism is electrically connected to the host computer for transmitting the point cloud image to the host computer.

[0023] Furthermore, the point cloud image acquisition mechanism includes a stepper motor and a CCD vision sensor, the stepper motor is arranged on the linear motion module, the stepper motor is connected to the CCD vision sensor, and the CCD vision sensor is spaced relative to the two-dimensional linear motion platform in the upper and lower directions for acquiring the point cloud image of the differential pressure flowmeter orifice plate.

[0024] Furthermore, the linear motion module includes a linear displacement slide and a support frame, the support frame is vertically slidably connected to the linear displacement slide, and the point cloud image acquisition mechanism is arranged at one end of the support frame extending out of the linear displacement slide.

[0025] Furthermore, a photoelectric limit switch is provided on the linear motion module, and the photoelectric limit switch is used to limit the up and down movement distance of the point cloud image acquisition mechanism.

[0026] Furthermore, the CCD visual sensor is connected to the stepping motor via a rotating pan-tilt platform.

[0027] The present invention adopts the above technical solution, which has the following advantages and effects:

[0028] The machine vision-based orifice wear detection method and device for a differential pressure flowmeter provided by the present invention first obtain a point cloud image of the orifice of the differential pressure flowmeter, optimize the point cloud image to obtain point cloud data, then construct a three-dimensional model of the orifice of the differential pressure flowmeter through the point cloud data, and then perform stress simulation analysis on the three-dimensional model to obtain stress changes of the orifice of the differential pressure flowmeter under alternating water pressure, select multiple axial pressure nodes of the orifice of the differential pressure flowmeter to solve the maximum stress, and find the corresponding abscissa value of the maximum stress on the ordinate of the SN curve as the number of cycles of the orifice of the differential pressure flowmeter under water impact, thereby realizing high-precision automatic detection of the orifice of the differential pressure flowmeter. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 The present invention is a flow chart of a method for detecting orifice wear of a differential pressure flowmeter based on machine vision.

[0030] Figure 2 It is a structural schematic diagram of a differential pressure flowmeter orifice wear detection device based on machine vision of the present invention.

[0031] Figure 3 yes Figure 2 Top view of the .

[0032] Figure 4 yes Figure 2 Side view of.

[0033] The accompanying drawings are marked as: 1-host computer, 2-linear motion module, 201-linear displacement slide, 202-support frame, 3-two-dimensional linear motion table, 4-support platform, 5-point cloud image acquisition mechanism, 501-stepping motor, 502-CCD visual sensor, 503-rotating pan-tilt head, 6-differential pressure flowmeter orifice plate. DETAILED DESCRIPTION

[0034] The embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings so that the purpose, features and advantages of the present invention can be more clearly understood. It should be understood that the embodiments shown in the accompanying drawings are not intended to limit the scope of the present invention, but are only intended to illustrate the essential spirit of the technical solution of the present invention.

[0035] like Figure 1 The machine vision-based differential pressure flow meter orifice wear detection method provided by the present invention comprises the following steps:

[0036] Step S1: First, the world coordinate system is converted into the CCD vision sensor coordinate system, and the physical image coordinate system of the differential pressure flowmeter orifice plate is obtained through the CCD vision sensor coordinate system;

[0037] Secondly, the orifice plate image of the differential pressure flowmeter is collected by the CCD visual sensor to obtain the point cloud image of the orifice plate of the differential pressure flowmeter.

[0038] Specifically, the world coordinate system is a real world coordinate system, and the CCD vision sensor coordinate system is a three-dimensional coordinate system with the CCD vision sensor as the origin.

[0039] Convert the world coordinate system to the CCD vision sensor coordinate system, specifically:

[0040] Through the rotation matrix R and translation matrix t The world coordinate system is converted into the CCD vision sensor coordinate system as shown in the following formula:

[0041] ;

[0042] Among them, the rotation matrix R is a 3×3 orthogonal matrix, the translation matrix t is a 3×1 translation vector, T is the transposed vector, P W is a point in the world coordinate system, X W 、Y W 、Z W for P W The coordinates of P C is a point in the CCD vision sensor coordinate system, X C 、Y C 、Z C for P C The coordinates of .

[0043] The physical image coordinate system of the differential pressure flowmeter orifice plate is obtained through the CCD visual sensor coordinate system, as shown in the following formula:

[0044] ;

[0045] in, is the homogeneous coordinate system in the physical image coordinate system of the differential pressure flowmeter orifice plate, f x 、f yis the focal length of the CCD vision sensor coordinate system, c x 、c y It is the coordinate origin of the physical image coordinate system of the differential pressure flowmeter orifice plate.

[0046] Step S2: De-noising is performed on the coordinate points of the physical image coordinate system and the coordinate points of the point cloud image to obtain the physical image optimized coordinate system and the point cloud optimized image respectively.

[0047] Specifically, the coordinate points of the physical image coordinate system of the differential pressure flowmeter orifice plate are denoised by using a statistical filtering method, the coordinate noise points in the physical image coordinate system of the differential pressure flowmeter orifice plate are removed, and the physical image optimized coordinate system of the differential pressure flowmeter orifice plate is obtained;

[0048] The point cloud image of the orifice plate of the differential pressure flowmeter is denoised by using the statistical filtering method to remove the noise points of the point cloud image of the orifice plate of the differential pressure flowmeter and obtain the point cloud optimized image of the orifice plate of the differential pressure flowmeter.

[0049] Step S3: using a signal-to-noise ratio evaluation algorithm to respectively calculate the noise ratio of the physical image optimization coordinate system of the differential pressure flowmeter orifice plate and the noise ratio of the point cloud optimization image of the differential pressure flowmeter orifice plate, and compare the noise ratios of the two;

[0050] If the noise ratio of the point cloud optimization image of the differential pressure flow meter orifice plate is greater than the noise ratio of the physical image optimization coordinate system of the differential pressure flow meter orifice plate, the coordinate points of the point cloud optimization image of the differential pressure flow meter orifice plate are used as the point cloud data of the differential pressure flow meter orifice plate;

[0051] If the noise ratio of the physical image optimization coordinate system of the differential pressure flowmeter orifice plate is greater than the noise ratio of the point cloud optimization image of the differential pressure flowmeter orifice plate, the coordinate points of the physical image optimization coordinate system of the differential pressure flowmeter orifice plate are used as the point cloud data of the differential pressure flowmeter orifice plate.

[0052] Step S4: After the point cloud data are spliced, point cloud fusion is performed to establish a three-dimensional model of the differential pressure flowmeter orifice plate after wear.

[0053] Specifically, the incorrect point elimination algorithm is first used to eliminate the irrelevant point cloud data in the point cloud data of the differential pressure flowmeter orifice to obtain the remaining point cloud data. The least squares method is then used to splice the remaining point cloud data to obtain the spliced ​​data. Finally, the Cloud Compare software is used to perform point cloud fusion on the spliced ​​data to establish a three-dimensional model of the worn differential pressure flowmeter orifice.

[0054] Step S5: Perform stress analysis on the three-dimensional model to obtain the maximum stress of the three-dimensional model, and find the corresponding maximum stress on the ordinate of the SN curve. The abscissa value of the SN curve corresponding to the maximum stress is the number of cycles of the differential pressure flowmeter orifice.

[0055] Specifically, the following steps are included:

[0056] Step S51: importing the three-dimensional model into the Structural Mechanics module of the ANSYS software to perform stress simulation analysis, solving the stress change of the three-dimensional model under the alternating action of water pressure, and solving the maximum stress of the three-dimensional model after selecting multiple axial pressure nodes of the three-dimensional model;

[0057] Step S52: Estimate the number of cycles of the three-dimensional model according to the SN curve (strain-fatigue life curve). When estimating, find the corresponding maximum stress on the ordinate of the SN curve. The abscissa value of the SN curve corresponding to the maximum stress is the number of cycles of the differential pressure flowmeter orifice under water impact.

[0058] like Figure 2-Figure 4 The present invention also provides a pressure differential flow meter orifice plate detection device based on machine vision, and the detection device applies the detection method of the present invention.

[0059] The detection device includes a host computer 1, a linear motion module 2, a two-dimensional linear motion platform 3, a support platform 4 and a point cloud image acquisition mechanism 5. The linear motion module 2 and the two-dimensional linear motion platform 3 are both arranged on the support platform 4. The point cloud image acquisition mechanism 5 is arranged on the linear motion module 2 and is spaced relative to the two-dimensional linear motion platform 3. The linear motion module 2 drives the point cloud image acquisition mechanism 5 to move up and down to acquire a point cloud image of a differential pressure flow meter orifice 6 arranged on the two-dimensional linear motion platform 3. The point cloud image acquisition mechanism 5 is electrically connected to the host computer 1 to transmit the point cloud image to the host computer 1, and the three-dimensional model of the differential pressure flow meter orifice 6 is obtained through the host computer 1. After stress analysis of the three-dimensional model, the number of cycles of the differential pressure flow meter orifice 6 under water impact is obtained.

[0060] Specifically, the linear motion module 2 is vertically mounted on the upper surface of one end of the support platform 4, the host computer 1 is arranged on the upper surface of the other end of the support platform 4, and the point cloud image acquisition mechanism 5 is mounted on the linear motion module 2 and driven by the linear motion module 2 to move vertically up and down along the linear motion module 2. The point cloud image acquisition mechanism 5 is used to acquire the point cloud image of the differential pressure flow meter orifice 6. The two-dimensional linear motion platform 3 is arranged on the support platform 4 at the lower end of the point cloud image acquisition mechanism 5. The differential pressure flow meter orifice 6 is placed on the two-dimensional linear motion platform 3, and the horizontal position of the differential pressure flow meter orifice 6 can be adjusted by the two-dimensional linear motion platform 3.

[0061] Furthermore, the linear motion module 2 includes a linear displacement slide 201 and a support frame 202 . The support frame 202 is vertically connected to the linear displacement slide 201 . A point cloud image acquisition mechanism 5 is disposed at one end of the support frame 202 extending out of the linear displacement slide 201 .

[0062] Furthermore, the point cloud image acquisition mechanism 5 includes a stepper motor 501, a CCD visual sensor 502 and a rotating pan-tilt table 503. The stepper motor 501 is arranged at one end of the support frame 202 extending from the linear displacement slide 201. The output end of the stepper motor 501 faces the side of the support platform 4. The output end of the stepper motor 501 is connected to the CCD visual sensor 502 through the rotating pan-tilt table 503. The CCD visual sensor 502 is spaced apart from the two-dimensional linear moving table 3 in an upper and lower manner. The stepper motor 501 drives the rotating pan-tilt table 503 to drive the CCD visual sensor 502 to rotate. The CCD visual sensor 502 collects the point cloud image of the differential pressure flowmeter orifice plate 6, and the CCD visual sensor 502 transmits the collected point cloud image to the host computer 1.

[0063] Specifically, the stepper motor 501 is vertically embedded in the end of the extended end of the support frame 202, the output end of the stepper motor 501 is connected to the rotating platform 503, the rotating platform 503 is located at the lower end of the support frame 202, and the lower end of the rotating platform 503 is connected to the CCD visual sensor 502.

[0064] Furthermore, a photoelectric limit switch is provided on the linear motion module 2 , and the photoelectric limit switch is used to limit the vertical movement distance of the point cloud image acquisition mechanism 5 .

[0065] Specifically, a pair of photoelectric limit switches are respectively arranged at the upper and lower ends of the linear displacement slide 201. When the slider of the linear displacement slide 201 drives the support frame 202 to move up and down, a pair of photoelectric limit switches can be triggered to control the distance of the support frame 202 moving up and down.

[0066] When the machine vision-based pressure differential flowmeter orifice detection device of the present invention is in use, the pressure differential flowmeter orifice 6 is placed on the two-dimensional linear moving table 3, and the pressure differential flowmeter orifice 6 is located directly below the CCD vision sensor 502 by adjusting the two-dimensional linear moving table 3. Then, the linear moving module 2 is controlled to drive the point cloud image acquisition mechanism 5 to move vertically to collect the point cloud image of the pressure differential flowmeter orifice 6. The point cloud image acquisition mechanism 5 transmits the collected point cloud image to the host computer 1, and the host computer 1 processes it to obtain the number of cycles of the pressure differential flowmeter orifice 6 under water impact.

Claims

1. A method for detecting orifice wear of a differential pressure flowmeter based on machine vision, characterized in that: The following steps are involved: Step S1: converting the world coordinate system into the CCD vision sensor coordinate system, and obtaining the physical image coordinate system of the differential pressure flowmeter orifice plate through the CCD vision sensor coordinate system; The orifice plate image of the differential pressure flowmeter is collected by a CCD visual sensor to obtain a point cloud image of the orifice plate of the differential pressure flowmeter; Step S2: performing denoising processing on the coordinate points of the physical image coordinate system and the coordinate points of the point cloud image to obtain the physical image optimized coordinate system and the point cloud optimized image respectively; Step S3: respectively calculating the noise ratio of the physical image optimized coordinate system and the noise ratio of the point cloud optimized image, and comparing the noise ratios of the two, taking the coordinate point of the physical image optimized coordinate system or the coordinate point of the point cloud optimized image with a larger noise ratio as the point cloud data of the orifice plate of the differential pressure flowmeter; Step S4: performing point cloud fusion after splicing the point cloud data to establish a three-dimensional model of the differential pressure flowmeter orifice plate after wear; Step S5: Perform stress analysis on the three-dimensional model to obtain the maximum stress of the three-dimensional model, and find the corresponding maximum stress on the ordinate of the SN curve. The abscissa value of the SN curve corresponding to the maximum stress is the number of cycles of the differential pressure flowmeter orifice plate.

2. The method for detecting orifice plate wear of a differential pressure flowmeter based on machine vision according to claim 1, characterized in that: In step S1, the world coordinate system is converted to the CCD vision sensor coordinate system, specifically: by rotating the matrix R and translating the matrix t The world coordinate system is converted into the CCD vision sensor coordinate system as shown in the following formula: ； Among them, the rotation matrix R is a 3×3 orthogonal matrix, the translation matrix t is a 3×1 translation vector, T is the transposed vector, P W is a point in the world coordinate system, X W 、Y W 、Z W for P W The coordinates of P C is a point in the CCD vision sensor coordinate system, X C 、Y C 、Z C for P C The coordinates of In the step S1, the orifice plate image of the differential pressure flowmeter is collected by a CCD visual sensor to obtain a point cloud image of the orifice plate of the differential pressure flowmeter, as shown in the following formula; ； in, is the homogeneous coordinate system of the physical image coordinate system, f x 、f y is the focal length of the CCD vision sensor coordinate system, c x 、c y It is the coordinate origin of the physical image coordinate system of the differential pressure flowmeter orifice plate.

3. The method for detecting orifice plate wear of a differential pressure flowmeter based on machine vision according to claim 2, characterized in that: In the step S2, the denoising process is to use a statistical filtering method to remove coordinate noise points in the physical image coordinate system or noise points in the point cloud image.

4. The method for detecting orifice plate wear of a differential pressure flowmeter based on machine vision according to claim 3 is characterized in that: In the step S4, the point cloud data are spliced ​​and then point cloud fusion is performed, specifically: the point cloud data that does not correspond in the point cloud data of the differential pressure flowmeter orifice plate is eliminated using an error point elimination algorithm to obtain the remaining point cloud data, the remaining point cloud data is spliced ​​using the least squares method to obtain spliced ​​data, and the spliced ​​data is point cloud fused.

5. The method for detecting orifice plate wear of a differential pressure flowmeter based on machine vision according to claim 4, characterized in that: In the step S4, the point cloud fusion is performed by Cloud Compare software; In the step S5, stress analysis is performed on the three-dimensional model using ANSYS software.

6. A pressure differential flowmeter orifice wear detection device based on machine vision, using the pressure differential flowmeter orifice wear detection method based on machine vision as described in any one of claims 1 to 5, characterized in that: The invention comprises a host computer (1), a linear motion module (2), a two-dimensional linear motion platform (3), a supporting platform (4) and a point cloud image acquisition mechanism (5), wherein the linear motion module (2) and the two-dimensional linear motion platform (3) are both arranged on the supporting platform (4), the point cloud image acquisition mechanism (5) is arranged on the linear motion module (2) and is spaced apart from the two-dimensional linear motion platform (3) in an upper and lower direction, the linear motion module (2) drives the point cloud image acquisition mechanism (5) to move up and down to acquire a point cloud image of a differential pressure flow meter orifice plate (6) arranged on the two-dimensional linear motion platform (3), and the point cloud image acquisition mechanism (5) is electrically connected to the host computer (1) to transmit the point cloud image to the host computer (1).

7. The machine vision-based differential pressure flowmeter orifice wear detection device according to claim 6, characterized in that: The point cloud image acquisition mechanism (5) comprises a stepper motor (501) and a CCD visual sensor (502); the stepper motor (501) is arranged on the linear motion module (2); the stepper motor (501) is connected to the CCD visual sensor (502); the CCD visual sensor (502) and the two-dimensional linear motion platform (3) are spaced relative to each other in an upper and lower manner and are used to acquire a point cloud image of the differential pressure flow meter orifice plate (6).

8. The machine vision-based differential pressure flowmeter orifice wear detection device according to claim 7, characterized in that: The linear motion module (2) comprises a linear displacement slide (201) and a support frame (202), wherein the support frame (202) is vertically slidably connected to the linear displacement slide (201), and the point cloud image acquisition mechanism (5) is arranged at one end of the support frame (202) extending out of the linear displacement slide (201).

9. The machine vision-based differential pressure flowmeter orifice wear detection device according to claim 8, characterized in that: The linear motion module (2) is also provided with a photoelectric limit switch, and the photoelectric limit switch is used to limit the distance of the upward and downward movement of the point cloud image acquisition mechanism (5).

10. The machine vision-based differential pressure flowmeter orifice wear detection device according to claim 7, characterized in that: The CCD visual sensor (502) is connected to the stepping motor (501) via a rotating pan-tilt platform (503).

Images