阅读说明：本技术 基于改进的dpiv垂直井油水两相流流速场测量方法 (Improved DPIV vertical well-based oil-water two-phase flow velocity field measurement method ) 是由 韩连福 丛垚 付长凤 刘超 刘兴斌 姜继玉 于 2019-12-09 设计创作，主要内容包括：本发明涉及的是基于改进的DPIV垂直井油水两相流流速场测量方法,这种基于改进的DPIV垂直井油水两相流流速场测量方法采用迭代最近点(Iterative Closest Point,ICP)代替传统互相关进行图像匹配,同时利用移动最小二乘(Moving least squares,MLS)综合整个流速场信息进行边界位移值补充,改善了DPIV的图像匹配效果,提高了垂直井油水两相流的测量精度。本发明在匹配的过程中同时考虑了平面内油滴的平移和旋转,能够提高两相流灰度图像的匹配精度,解决了在垂直井油水两相流流速场测量中由于图像匹配效果差、位移场边界缺失导致的测量精度降低问题。(The invention relates to an improved DPIV-based vertical well oil-water two-phase flow velocity field measurement method, which adopts Iterative Closest Point (ICP) to replace the traditional cross correlation for image matching, and meanwhile, uses Moving Least Squares (MLS) to synthesize the whole flow velocity field information for boundary displacement value supplement, thereby improving the image matching effect of DPIV and improving the measurement precision of the vertical well oil-water two-phase flow. The invention simultaneously considers the translation and rotation of oil drops in a plane in the matching process, can improve the matching precision of two-phase flow gray images, and solves the problem of measurement precision reduction caused by poor image matching effect and displacement field boundary loss in the measurement of the oil-water two-phase flow velocity field of the vertical well.)

1. An improved DPIV vertical well-based oil-water two-phase flow velocity field measurement method is characterized by comprising the following steps:

the method comprises the following steps: selecting two frames of oil-water two-phase flow images with the time interval delta t, wherein the image size is expressed as: length is multiplied by width, the image size is set to be M pixel multiplied by N pixel, M is the value of the image length, N is the value of the image width, after image denoising and image contrast enhancement are carried out on the image width, the size of an initial query window is determined, and the size of the initial query window is expressed as: setting the length multiplied by the width, setting the size of an initial query window as Wpixel multiplied by Wpixel, setting W as the value of the length and the width of the query window, setting the query step length as W/2pixel, dividing two frames of oil-water two-phase flow images into (2M-W)/Wx (2N-W)/W grids with the coverage rate of 50% according to the size of the corresponding initial query window, setting the current query window, and making the size of the current query window equal to the size of the initial query window;

step two: region selection is performed in two images, and the selected region is expressed as: [ X coordinate Range Start: end of X-coordinate range, start of Y-coordinate range: end of range of Y coordinates]Setting the selected query Area in the first frame image₁＝[i:i+W-1,j:j+W-1]Selecting a query Area in the second frame image₂＝[i:i+W-1,j:j+W-1]I and j respectively represent X and Y coordinate values in the image, i is 1+ W (n-1), j is 1+ W (M-1), n is 1,2, …, (2M-W)/W, M is 1,2, …, (2M-W)/WN-W)/W, m is the sequence number of the query area in the X direction, and N is the sequence number of the query area in the Y direction; for the Area of query₁And query Area₂Performing ICP registration on the two areas to obtain an Area to be inquired₁Average X-direction displacement u_(i',j')Average Y-direction displacement v_(i',j')I ═ i + W/2 is the query Area₁The central X coordinate, j' ═ j + W/2 is the Area of inquiry₁A center Y coordinate;

step three: traversing two frames of images by the current query window by step length W/2pixel to obtain an initial X-direction displacement field U_initialInitial Y-direction displacement field V_initialSetting a current X-direction displacement field as U and a current Y-direction displacement field as V;

step four: deforming the current query window according to the current X-direction displacement field U and the current Y-direction displacement field V, and performing ICP registration again to obtain a secondary iteration X-direction displacement field U_newSecond iteration Y direction displacement field V_new；

Step five: updating the current X-direction displacement field U and the current Y-direction displacement field V, and updating the displacement fields according to the following formula:

U＝U_initial+U_new，V＝V_initial+V_new；

step six: fitting the current X-direction displacement field U and the current Y-direction displacement field V by adopting an MLS (Multi-level modeling System) to obtain an X-direction edge supplementary displacement field U 'and a Y-direction edge supplementary displacement field V'; the X-direction edge supplementary displacement field U 'and the Y-direction edge supplementary displacement field V' are calculated according to the following method:

x-direction edge supplementary displacement field U' curved surface fitting function f_u(x, Y) and Y-direction edge complementary displacement field V' surface fitting function f_v(X, Y), wherein X is an X-direction coordinate variable, Y is a Y-direction coordinate variable, and k is a polynomial serial number:

wherein the surface fitting function f_uCoefficient array α of (x, y)_u(x,y)＝[α_u1(x,y),α_u2(x,y),…,α_uk(x,y)]，α_uk(x, y) is a surface fitting function f_u(x, y) kth coefficient, surface fitting function f_vCoefficient array α of (x, y)_v(x,y)＝[α_v1(x,y),α_v2(x,y),…,α_vk(x,y)]，α_vk(x, y) is a surface fitting function f_v(x, y) k-th coefficient, and variable array μ (x, y) ═ μ₁(x,y),μ₂(x,y),...,μ_k(x,y)]＝[1,x,y,x²,xy,y²]，μ_k(x, y) is the kth variable of two surface fitting functions, and T represents a matrix transposition symbol;

α_u(x,y)、α_v(x, y) is calculated as:

wherein the known X-direction displacement array Z_u＝[u_{(W/2+1,W/2+1)},u_{(3W/2+1,3W/2+1)},…,u_(i',j')]I ═ 1+ W (N-1/2), j ═ 1+ W (M-1/2), N ═ 1,2, …, (2M-W)/W, M ═ 1,2, …, (2N-W)/W, and it is known that the Y-direction displacement array Z is a linear displacement array_v＝[v_{(W/2+1,W/2+1)},v_{(3W/2+1,3W/2+1)},…,v_(i',j')]I ═ 1+ W (N-1/2), j ═ 1+ W (M-1/2), N ═ 1,2, …, (2M-W)/W, M ═ 1,2, …, (2N-W)/W, and the parameter array G ═ μ M^T(_W/2+1,W/2+1),μ^T(_{3W/2+1,3W/2+1}),…,μ^T(i',j')]I ═ 1+ W (N-1/2), j ═ 1+ W (M-1/2), N ═ 1,2, …, (2M-W)/W, M ═ 1,2, …, (2N-W)/W, weight diagonal matrix

α will be mixed_u(x,y)、α_vSubstitution of (x, y) into f_u(x,y)、f_vIn the step (X, Y), obtaining a fitting surface equation of an X-direction edge supplementary displacement field U 'and a Y-direction edge supplementary displacement field V':

the X-direction edge supplemental displacement field U 'and the Y-direction edge supplemental displacement field V' are now expressed as:

at this time, the current X-direction displacement field U is equal to U ', and the current Y-direction displacement field V is equal to V';

step seven: expanding the current X-direction displacement field U and the current Y-direction displacement field V by adopting bicubic uniform B spline interpolation to be 4 times of the original displacement field V;

step eight: reducing the size of the current query window to 1/4 of the original size to obtain a new-size query window, enabling the size of the current query window to be equal to that of the new-size query window, deforming the current query window according to the current X-direction displacement field U and the current Y-direction displacement field V, and performing ICP (inductively coupled plasma) registration again to obtain a three-time iteration X-direction displacement field U'_newAnd V 'in Y direction in three iterations'_new；

Step nine: updating the current X-direction displacement field U and the current Y-direction displacement field V;

step ten: iterating the current query window to carry out the steps four to nine until the current query window is reduced to the size of the specified query window, and determining the final X-direction displacement field U_finalFinal Y-direction displacement field V_finalAccording to the time interval Δ t and the displacement field U_final、V_finalObtaining a flow velocity field f; the oil-water two-phase flow velocity field is calculated according to the following formula:

2. the improved DPIV vertical well oil-water two-phase flow velocity field measurement method as claimed in claim 1, wherein the method comprises the following steps: the ICP registration method in the second step comprises the following steps: setting W to 128, initial query window size 128 pixels by 128 pixels, query Area₁＝[i:i+127,j:j+127]Query Area₂＝[i:i+127,j:j+127]Setting a query Area₁Is p ═ p_ii1,2, …,16384, ii is the query Area₁Data point number of p_iiQuerying the Area for the ii data point in the set of data points p₂Is q ═ q_jj1,2, …,16384, jj being the query Area₂Data point number of (1), q_jjFor the jj data point in the data point set q, the average X-direction displacement u of the region_(i',j')Average Y-direction displacement v_(i',j')The calculation is as follows:

the data point set transformation relation q' is:

q′＝rp+t

the matching objective function E is:

where r is a rotation matrix, t is a translation vector, q'_jjThe jj data point in the data point set after the data point set p is transformed by the data point set transformation relational expression q';

solving a rotation matrix r and a translational vector t by adopting SVD, and transforming point sets P and q as follows, P_iiIs p_iiTransformed data, Q_jjIs q_jjTransformed data:

comprises the following steps:

performing singular value decomposition on the optimal solution matrix H, and decomposing H into a left singular matrix D, a right singular matrix L and a singular value matrix Lambda:

H＝DΛL^T

the calculation of r and t is as follows:

the average displacement of this region is:

u_(i′,j′)＝t(1)，v_(i′,j′)＝t(2)。

3. the improved DPIV vertical well oil-water two-phase flow velocity field measurement method as claimed in claim 2, wherein the method comprises the following steps: in the third step, the displacement field U in the X direction is initiated_initialInitial Y-direction displacement field V_initialExpressed as:

wherein i ═ 1+128(N-1/2), j ═ 1+128(M-1/2), N ═ 1,2, …, (2M-128)/128, M ═ 1,2, …, (2N-128)/128; the current X-direction displacement field U is equal to U_initialThe current Y-direction displacement field V is equal to V_initial。

4. The improved DPIV vertical well oil-water two-phase flow velocity field measurement method as claimed in claim 3, wherein the method comprises the following steps: after the current query window is deformed in the fourth step, the query Area of the first frame image is Area₁＝[i:i+W-1,j:j+W-1]The query Area of the second frame image is Area₂＝[i+u(i',j'):i+W-1+u(i',j'),j+v(i',j'):j+W-1+v(i',j')]。

5. The improved DPIV vertical well oil-water two-phase flow velocity field measurement method as claimed in claim 4, wherein the method comprises the following steps: in the seventh step, the bicubic uniform B-spline interpolation is calculated according to the following formula:

in the formula (c) (-)_uIs a bicubic uniform B-spline surface patch of the current X-direction displacement field U_vIs a bicubic uniform B-spline surface sheet of the current Y-direction displacement field V, and the X-direction coordinate variable matrix lambda is [ ξ ]³ξ²ξ 1]ξ is an X coordinate variable, and Y-direction coordinate variable matrix γ ═ ζ³ζ²ζ 1]Zeta is a Y coordinate variable, B_wAs a B-spline basis function, C_uIs the control peak, C, of the bicubic uniform B-spline surface of the current X-direction displacement field U_vWhen the X coordinate variable ξ and the Y coordinate variable zeta are in [0,1 ] for the control vertex of the bicubic uniform B-spline curved surface sheet of the current Y-direction displacement field V]And obtaining the displacement value of any point on the curved surface sheet during the process of the motion.

6. The improved DPIV vertical well oil-water two-phase flow velocity field measurement method as claimed in claim 5, wherein the method comprises the following steps: step eight, after the current query window is deformed, the query Area of the first frame image₁＝[i:i+W-1,j:j+W-1]Query Area of second frame image₂＝[i+u(i',j'):i+W-1+u(i',j'),j+v(i',j'):j+W-1+v(i',j')]。

7. The improved DPIV vertical well oil-water two-phase flow velocity field measurement method according to claim 6, wherein the method comprises the following steps: in the ninth step, the displacement field is updated according to the following formula:

U＝U′+U′_new，V＝V′+V′_new。

The technical field is as follows:

the invention relates to the technical field of petroleum engineering and measurement, in particular to an improved DPIV vertical well oil-water two-phase flow velocity field measurement method.

Background art:

the flow velocity measurement is an essential part for dynamic monitoring of oil field development, and the oil-water two-phase flow velocity measurement methods include an ultrasonic method, a capacitance method, an electric conductivity method, an electromagnetic method, a heat tracing method, an optical fiber method and the like. Oilfield flow rate measurements of the current time period are more favored over zonal flow rate profile measurements because multiphase flow characteristics are more easily observed from zonal flow rate profiles, thereby guiding oilfield development. The traditional flow rate measurement method cannot meet the requirement, and a new flow rate measurement method is adopted. DPIV is an effective full-field speed measurement method in the field of hydrodynamics, and the characteristics of no disturbance, no contact and full-field speed measurement can well meet the requirements of multiphase flow speed measurement of oil fields, so that the DPIV has considerable application prospect in oil field speed measurement. The influence factors of the DPIV technology measurement accuracy are generally divided into two categories: one is the effect of external measurement environments such as trace particle distribution, illumination distribution, and various hardware systems, and the other is the implementation of DPIV image post-processing techniques. The Window Iterative Deformation (WIDIM) algorithm is used for improving the followability of the query Window under the condition that large-speed gradient tracer particles exist in a flow field, and is a DPIV algorithm widely used at present.

The oil-water two-phase flow field measurement by applying the WIDIM-based DPIV algorithm in the oil-water two-phase flow measurement of the 125mm vertical well has the following problems: (1) due to the fact that the diameter of the oil well is large, relative motion between oil drops in the query window is large in repeated reading, and the follow-up performance of the query window is high; (2) because the DPIV technology is applied to two-phase flow and adopts LED illumination and oil drop tracing, the gray level image of the two-phase flow to be processed is more complex than the gray level image information of the single-phase flow, so that the cross-correlation gray level calculation is easy to generate error matching to bring measurement errors, the errors can generate an amplification effect in the iteration process to seriously influence the image deformation effect, and the precision of the iteration calculation is improved; (3) the problem of displacement field boundary loss exists in the interpolation process of the WIDIM algorithm, and the window deformation effect is also influenced. Iterative Closest Point (ICP) is a Point cloud data registration method, which can realize accurate registration of a large amount of data, and applies a three-dimensional ICP technique to two-dimensional DPIV image registration, only a DPIV gray image needs to be projected to a three-dimensional space, and compared with cross-correlation matching, ICP registration is less affected by noise, and planar rotational motion of oil droplets is considered in the registration process. And assigning values to the missing displacement field boundary by comprehensively considering the whole displacement field information, so that a DPIV algorithm is further improved by adopting a Moving Least Square (MLS) method, MLS (Moving least square) surface fitting is carried out on the current displacement field before each interpolation, and the missing boundary value is supplemented according to a surface fitting function.

The invention content is as follows:

the invention aims to provide an improved DPIV vertical well oil-water two-phase flow velocity field measuring method which is used for solving the problems of poor image matching effect and inaccurate flow velocity measurement caused by the loss of a displacement field boundary in the vertical well oil-water two-phase flow measurement of a DPIV algorithm based on WIDIM.

The technical scheme adopted by the invention for solving the technical problems is as follows: the improved DPIV vertical well-based oil-water two-phase flow velocity field measurement method comprises the following steps:

the method comprises the following steps: selecting two frames of oil-water two-phase flow images with the time interval delta t, wherein the image size is expressed as: length is multiplied by width, the image size is set to be M pixel multiplied by N pixel, M is the value of the image length, N is the value of the image width, after image denoising and image contrast enhancement are carried out on the image width, the size of an initial query window is determined, and the size of the initial query window is expressed as: setting the length multiplied by the width, setting the size of an initial query window as W pixel multiplied by W pixel, setting W as the value of the length and the width of the query window, setting the query step length as W/2pixel, dividing two frames of oil-water two-phase flow images into (2M-W)/Wx (2N-W)/W grids with the coverage rate of 50% according to the size of the corresponding initial query window, setting the current query window, and making the size of the current query window equal to the size of the initial query window;

step two: region selection is performed in two images, and the selected region is expressed as: [ X coordinate Range Start: end of X-coordinate range, Y-coordinate rangeStarting point: end of range of Y coordinates]Setting the selected query Area in the first frame image₁＝[i:i+W-1,j:j+W-1]Selecting a query Area in the second frame image₂＝[i:i+W-1,j:j+W-1]I and j respectively represent X and Y coordinate values in the image, i is 1+ W (N-1), j is 1+ W (M-1), N is 1,2, …, (2M-W)/W, M is 1,2, …, (2N-W)/W, M is an X-direction query region number, and N is a Y-direction query region number; for the Area of query₁And query Area₂Performing ICP registration on the two areas to obtain an Area to be inquired₁Average X-direction displacement u_(i',j')Average Y-direction displacement v_(i',j')I ═ i + W/2 is the query Area₁The central X coordinate, j' ═ j + W/2 is the Area of inquiry₁A center Y coordinate;

step three: traversing two frames of images by the current query window by step length W/2pixel to obtain an initial X-direction displacement field U_initialInitial Y-direction displacement field V_initialSetting a current X-direction displacement field as U and a current Y-direction displacement field as V;

step four: deforming the current query window according to the current X-direction displacement field U and the current Y-direction displacement field V, and performing ICP registration again to obtain a secondary iteration X-direction displacement field U_newSecond iteration Y direction displacement field V_new；

Step five: updating the current X-direction displacement field U and the current Y-direction displacement field V, and updating the displacement fields according to the following formula:

U＝U_initial+U_new，V＝V_initial+V_new；

step six: fitting the current X-direction displacement field U and the current Y-direction displacement field V by adopting an MLS (Multi-level modeling System) to obtain an X-direction edge supplementary displacement field U 'and a Y-direction edge supplementary displacement field V'; the X-direction edge supplementary displacement field U 'and the Y-direction edge supplementary displacement field V' are calculated according to the following method:

x-direction edge supplementary displacement field U' curved surface fitting function f_u(x, Y) and Y-direction edge complementary displacement field V' surface fitting function f_v(X, Y), wherein X is an X-direction coordinate variable, Y is a Y-direction coordinate variable, and k is a polynomial serial number:

wherein the surface fitting function f_uCoefficient array α of (x, y)_u(x,y)＝[α_u1(x,y),α_u2(x,y),…,α_uk(x,y)]，α_uk(x, y) is a surface fitting function f_u(x, y) kth coefficient, surface fitting function f_vCoefficient array α of (x, y)_v(x,y)＝[α_v1(x,y),α_v2(x,y),…,α_vk(x,y)]，α_vk(x, y) is a surface fitting function f_v(x, y) k-th coefficient, and variable array μ (x, y) ═ μ₁(x,y),μ₂(x,y),...,μ_k(x,y)]＝[1,x,y,x²,xy,y²]，μ_k(x, y) is the kth variable of two surface fitting functions, and T represents a matrix transposition symbol;

α_u(x,y)、α_v(x, y) is calculated as:

wherein the known X-direction displacement array Z_u＝[u_{(W/2+1,W/2+1)},u_{(3W/2+1,3W/2+1)},…,u_(i',j')]I ═ 1+ W (N-1/2), j ═ 1+ W (M-1/2), N ═ 1,2, …, (2M-W)/W, M ═ 1,2, …, (2N-W)/W, and it is known that the Y-direction displacement array Z is a linear displacement array_v＝[v_{(W/2+1,W/2+1)},v_{(3W/2+1,3W/2+1)},…,v_(i',j')]I ═ 1+ W (N-1/2), j ═ 1+ W (M-1/2), N ═ 1,2, …, (2M-W)/W, M ═ 1,2, …, (2N-W)/W, and the parameter array G ═ μ M^T(W/2+1,W/2+1),μ^T(3W/2+1,3W/2+1),…,μ^T(i',j')]I ═ 1+ W (N-1/2), j ═ 1+ W (M-1/2), N ═ 1,2, …, (2M-W)/W, M ═ 1,2, …, (2N-W)/W, weight diagonal matrixIs a weight function with tightly-supported characteristics;

α will be mixed_u(x,y)、α_vSubstitution of (x, y) into f_u(x,y)、f_vIn the step (X, Y), obtaining a fitting surface equation of an X-direction edge supplementary displacement field U 'and a Y-direction edge supplementary displacement field V':

the X-direction edge supplemental displacement field U 'and the Y-direction edge supplemental displacement field V' are now expressed as:

at this time, the current X-direction displacement field U is equal to U ', and the current Y-direction displacement field V is equal to V';

step seven: expanding the current X-direction displacement field U and the current Y-direction displacement field V by adopting bicubic uniform B spline interpolation to be 4 times of the original displacement field V;

step eight: reducing the size of the current query window to 1/4 of the original size to obtain a new-size query window, enabling the size of the current query window to be equal to the size of the new-size query window, deforming the current query window according to the current X-direction displacement field U and the current Y-direction displacement field V, and performing ICP (inductively coupled plasma) registration again to obtain a three-time iteration X-direction displacement field U'_newAnd V 'in Y direction in three iterations'_new；

Step nine: updating the current X-direction displacement field U and the current Y-direction displacement field V;

step ten: iterating the current query window processed in the step nine to perform the step four to the step nine until the current query window is reduced to the size of the specified query window, and determining the final X-direction displacement field U_finalFinal Y-direction displacement field V_finalAccording to the time interval Δ t and the displacement field U_final、V_finalObtaining a flow velocity field f; the oil-water two-phase flow velocity field is calculated according to the following formula:

the ICP registration method in step two of the above scheme: set W-128, initial query window size 128 pixels by 128 pixels, query Area₁＝[i:i+127,j:j+127]Query Area₂＝[i:i+127,j:j+127]Setting a query Area₁Is p ═ p_ii1,2, …,16384, ii is the query Area₁Data point number of p_iiQuerying the Area for the ii data point in the set of data points p₂Is q ═ q_jj1,2, …,16384, jj being the query Area₂Data point number of (1), q_jjFor the jj data point in the data point set q, the average X-direction displacement u of the region_(i',j')Average Y-direction displacement v_(i',j')The calculation is as follows:

the data point set transformation relation q' is:

q′＝rp+t

the matching objective function E is:

where r is a rotation matrix, t is a translation vector, q'_jjThe jj data point in the data point set after the data point set p is transformed by the data point set transformation relational expression q';

solving a rotation matrix r and a translational vector t by adopting SVD, and transforming point sets P and q as follows, P_iiIs p_iiTransformed data, Q_jjIs q_jjTransformed data:

comprises the following steps:

performing singular value decomposition on the optimal solution matrix H, and decomposing H into a left singular matrix D, a right singular matrix L and a singular value matrix Lambda, T representing a matrix transposition symbol:

H＝DΛL^T

the calculation of r and t is as follows:

the average displacement of this region is:

u_(i′,j′)＝t(1)，v_(i′,j′)＝t(2)。

in the third step of the scheme, the displacement field U in the X direction is initialized_initialInitial Y-direction displacement field V_initialExpressed as:

wherein i ═ 1+128(N-1/2), j ═ 1+128(M-1/2), N ═ 1,2, …, (2M-128)/128, M ═ 1,2, …, (2N-128)/128; the current X-direction displacement field U is equal to U_initialThe current Y-direction displacement field V is equal to V_initial。

After the current query window is deformed in the fourth step of the scheme, the query Area of the first frame image is Area₁＝[i:i+W-1,j:j+W-1]The query Area of the second frame image is Area₂＝[i+u(i',j'):i+W-1+u(i',j'),j+v(i',j'):j+W-1+v(i',j')]。

The bicubic uniform B-spline interpolation in the seventh step of the scheme is calculated according to the following formula:

in the formula (c) (-)_uIs a bicubic uniform B-spline surface patch of the current X-direction displacement field U_vIs a bicubic uniform B-spline surface sheet of the current Y-direction displacement field V, and the X-direction coordinate variable matrix lambda is [ ξ ]³ξ²ξ1]And ξ is an X coordinate variable,y-direction coordinate variable matrix γ ═ ζ³ζ²ζ1]Zeta is a Y coordinate variable, B_wAs a B-spline basis function, C_uIs the control peak, C, of the bicubic uniform B-spline surface of the current X-direction displacement field U_vWhen the X coordinate variable ξ and the Y coordinate variable zeta are in [0,1 ] for the control vertex of the bicubic uniform B-spline curved surface sheet of the current Y-direction displacement field V]And obtaining the displacement value of any point on the curved surface sheet during the process of the motion.

In the eighth step of the scheme, after the current query window is deformed, the query Area of the first frame image₁＝[i:i+W-1,j:j+W-1]Query Area of second frame image₂＝[i+u(i',j'):i+W-1+u(i',j'),j+v(i',j'):j+W-1+v(i',j')]。

In the above scheme step nine, the displacement field is updated according to the following formula:

U＝U′+U′_new，V＝V′+V′_new。

the invention has the following beneficial effects:

(1) according to the invention, the LED backlight light source is used for replacing the traditional laser sheet light source in the measuring process, and oil drops are used as tracer particles, so that the problem that the traditional tracer is shielded by the oil drops to influence the measurement in the oil-water two-phase flow is solved, and the measurement precision of the flow field and the flow rate is improved;

(2) the invention adopts ICP to replace cross-correlation algorithm to carry out image matching, simultaneously considers the translation and rotation of oil drops in a plane in the matching process, and can improve the matching precision of two-phase flow gray level images;

(3) the invention adopts MLS algorithm to synthesize the information of the whole flow field, supplements the missing displacement field boundary value, improves the follow-up property of the query window of the DPIV algorithm based on WIDIM, and improves the measurement precision.

Description of the drawings:

fig. 1 is a schematic diagram of image coordinates.

Fig. 2 is a schematic diagram of an image traversed by a query window, wherein the query window traverses the whole image from top to bottom from left to right.

FIG. 3 is a vertical well oil-water two-phase flow velocity field measured using a modified DPIV algorithm.

FIG. 4 shows the flow velocity of the two-phase flow of oil and water in the vertical well measured by the improved DPIV algorithm and compared with the classical DPIV measurement results, wherein (a) is the measurement result of the average flow velocity of the two-phase flow of oil and water in the vertical well under the condition that the water content is 90%, and (b) is the measurement result of the average flow velocity of the two-phase flow of oil and water in the vertical well under the condition that the water content is 80%.

Detailed Description

The invention is further described below with reference to the accompanying drawings: