CN105806780A - Multi-vision-based system and method for measuring gas-liquid two-phase flow phase fraction of small channel - Google Patents

Abstract

The invention discloses a multi-vision-based system and method for measuring gas-liquid two-phase flow phase fraction of a small channel. The system comprises a high-speed camera, two light sources, a right-angle prism, a small pipeline, a glycerinum groove and a computer. The method comprises the following steps: firstly, simultaneously obtaining gas-liquid two-phase flow flowing image sequences of the small channel from two vertical angles; secondly, carrying out image pretreatment and optical distortion correction and building phase fraction measurement models of four typical flow patterns, namely bubble flow, slug flow, annular flow and wavy flow; thirdly, extracting feature vectors of images and carrying out flow pattern identification; and finally, selecting the phase fraction measurement model of the corresponding flow pattern and calculating the phase fraction. By a multi-vision technology, the phase fraction measurement model is built for each typical flow pattern, the flow pattern information is introduced into the phase fraction measurement model, and thus the measurement accuracy is improved. The corresponding device has the advantages of simplicity in structure, non-contact measurement, visualization and the like.

Description

Passage aisle gas-liquid two-phase flow containing rate based on many visions measures system and method

Technical field

The present invention relates to multiphase flow parameter measurement field, particularly relate to a kind of passage aisle gas-liquid two-phase flow containing rate based on many visions and measure system and method.

Background technology

In recent years, along with the emergence of micro-processing technology and new material technology, industrial equipment presents the development trend of microminiaturization, miniaturization.Owing to microminiaturized, compact apparatus have efficient, quick, the easily-controllable and advantage such as economize in raw materials, it is widely used in fields such as biology, medical treatment, chemical industry, pharmacy.Therefore, passage aisle gas-liquid two-phase flow parameter measurement becomes a current study hotspot and difficulties.

Compared to regular channels, passage aisle gas-liquid two-phase flow parameter measurement has certain difficulty.Under miniature scale, due to the reduction of hydraulic diameter so that the impact of surface tension and viscosity effect becomes big, and the impact of action of gravity weakens relatively, and theoretical model and empirical equation in regular channels are no longer applicable.

At present, existing passage aisle gas-liquid two-phase flow containing rate measuring method mainly has fast valve method, optical method, electric method and high speed video process etc..By contrast, high speed video process has intuitive, the advantage such as untouchable, is widely used in passage aisle two phase flow phase content is measured.But, the existing phase content measuring method majority based on high-speed camera is to obtain two phase flow image from an angle, and the shape information of gas-liquid two-phase spatial information and different angles is obtained deficiency.Many vision techniques utilize multiple camera or single camera and optics to constitute system and obtain the image information of object of study from multiple angles, efficiently solve the problem that single vision acquisition of information is not enough, provide probability for solving complex parameters measurement problem.

Summary of the invention

Not enough for the existing passage aisle gas-liquid two-phase flow containing rate measuring method acquisition of information based on high-speed photography, the problem that order of accuarcy is not high enough, the present invention is by repeatedly studying, for passage aisle biphase gas and liquid flow, propose a kind of passage aisle gas-liquid two-phase flow containing rate based on many visions and measure system and method, adopt many visual systemes to obtain two images being mutually perpendicular to angle of two phase flow simultaneously, carry out Image semantic classification and optic aberrance revising, set up the phase content measurement model of four kinds of typical flow patterns, extract characteristic vector and carry out meteor trail echoes, select the phase content measurement model of corresponding flow pattern, calculate phase content.

Passage aisle gas-liquid two-phase flow containing rate measurement system based on many visions includes passage aisle pipe-line system, many visions high-speed image sampling unit and microcomputer.Passage aisle pipe-line system is made up of the passage aisle being immersed in glycerol groove.Many visions high-speed image sampling unit is made up of two LED light sources, corner cube prism, high-speed cameras.Corner cube prism is placed near glycerol groove, and one right-angle side is perpendicular to the imaging plane of high-speed camera.Two light sources provide the backlight of two vertical direction, and the high-speed camera being positioned over glycerol groove front photographs two two phase flow images being mutually perpendicular to angle simultaneously, and the image transmitting collected calculates for image procossing and phase content in microcomputer.

Passage aisle gas-liquid two-phase flow containing rate measuring method step based on many visions is as follows:

The first step, obtains image: utilize many visions high-speed image sampling unit, obtains two passage aisle biphase gas and liquid flow flow image sequences being mutually perpendicular to angle simultaneously；

Second step, sets up phase content measurement model: carry out Image semantic classification and optic aberrance revising, and four kinds of typical flow patterns are modeled respectively；

3rd step, meteor trail echoes: extract the characteristic vector of image, carry out meteor trail echoes；

4th step, calculates phase content: select the phase content measurement model of corresponding flow pattern, calculates phase content.

Wherein, specifically comprising the following steps that of phase content measurement model is set up

1) Image semantic classification.Filled by difference shadow method, medium filtering, rim detection and binaryzation and carry out pretreatment, obtain bianry image.

2) optic aberrance revising.Correction coefficient is obtained, it is achieved the correction of biphase gas and liquid flow flow image by optical path analysis.For circular glass pipeline, its tube wall refraction rate and glycerol approximately equal.Pipeline is immersed in glycerol groove and can reduce the impact that captured image is caused by tube wall.As it is shown on figure 3, the foursquare lucite cuboid glycerol groove adopting cross section to be 25mm × 25mm, being immersed in by pipeline in glycerol groove, the center of pipeline center position and cross-sectional square shape coincides.For object point P in pipeline, (x, y), this point there is first time refraction at liquid phase and tube wall interface in the light sent, and in glycerol groove cell wall and Air Interface, second time refraction occurs.Two adjacent ray sent by P intersection point P'(x', the y' of reverse extending line after second time refraction) be a P to it should be observed that picture point, may thereby determine that the correction coefficient alpha=y/y' at a P place.After obtaining correction coefficient, bianry image is corrected.

3) four kinds of typical flow pattern modelings.Use for reference the thought of " segmentation, approximate summation ", set up bubble flow and slug flow phase content measurement model, annular flow phase content measurement model, wave flow phase content measurement model respectively.

Bubble flow and slug flow phase content measurement model: owing to the hydraulic diameter of passage aisle is less, the same cross section of pipeline only exists a bubble simultaneously, do not have same cross section in conventional pipeline and have the situation of multiple bubble, it is possible to determine the size of bubble in both direction by the image of two angles.Consider the situation of bubble flow and slug flow, at the same section of pipeline, be oval by region equivalent shared by gas phase.Within the scope of the axial length in pixels of pipeline, think that the area of section shared by gas phase is equal, therefore gas phase in this segment limit can be equivalent to an Elliptic Cylinder, and the major and minor axis of this Elliptic Cylinder bottom surface ellipse is determined by two bianry images being mutually perpendicular to angle respectively.Thus, it is possible to each bubble or slug to be equivalent to the superposition of several Elliptic Cylinders, each bubble or the volume of gas phase corresponding to vent plug can be tried to achieve.

As shown in fig. 5, it is assumed that captured single set of image exists n bubble/vent plug.Number of pixels shared by i-th bubble/vent plug length is l_i, number of pixels respectively a shared by the major and minor axis of the jth pixel place bubble section ellipse that i-th bubble/vent plug is corresponding_ijWith b_ij, then the volume V of gas phase in captured image correspondence pipeline section can be tried to achieve_gas1For:

$V_{gas1}=\frac{\mathrm{\π}}{4}{\mathrm{\μ}}^3\underset{i=1}{\overset{n}{\mathrm{\Σ}}}\underset{j=1}{\overset{l_i}{\mathrm{\Σ}}}{\mathrm{\α}}_{ij}{\mathrm{\β}}_{ij}{a}_{ij}{b}_{ij}$

Wherein, μ is the physical length that each square pixel is corresponding, α_ijFor a_ijCorresponding correction coefficient, β_ijFor b_ijCorresponding correction coefficient.The then phase content ε of bubble flow/slug flow in corresponding pipeline section₁For:

${\mathrm{\ϵ}}_1=\frac{V_{gas1}}{V_{pipe}}=\frac{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}\underset{j=1}{\overset{l_i}{\mathrm{\Σ}}}{\mathrm{\α}}_{ij}{\mathrm{\β}}_{ij}{a}_{ij}{b}_{ij}}{{\mathrm{LD}}^2}$

Wherein, V_pipeFor the conduit volume that image taking arrives, L is the duct length that image taking arrives, and D is the pipe diameter that image taking arrives, and the physical length μ that phase content is corresponding with pixel is unrelated, and only to each dimensional parameters of two phase flow image shared by the number of pixel relevant.

In actual photographed process, owing to length of pipe section taken in viewing field of camera is shorter, the calculated phase content of single set of image can not accurately reflect the phase content under current flow pattern.Therefore, choose the image construction image sequence being continuously shot by N group, take the meansigma methods of image sequence phase contentAs the phase content measured value of bubble flow/slug flow under corresponding experiment condition, it may be assumed that

${\stackrel{\‾}{\mathrm{\ϵ}}}_{bubble/slug}=\frac{\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\ϵ}}_k}{N}$

Annular flow phase content measurement model: in horizontal pipe, by the impact of action of gravity, duct bottom liquid film is thicker；And liquid film is distributed around even tube wall in vertical pipeline.Similar with slug flow and bubble flow, under annular flow flow pattern, gas phase in pipeline is approximately the superposition of several Elliptic Cylinders, gaseous phase volume can be tried to achieve accordingly, it is hereby achieved that phase content.As shown in Figure 6, in acquired annular flow bianry image, number of pixels respectively a shared by the major and minor axis that gas phase cross section, ith pixel place is oval_iWith b_i, then single set of image correspondence pipeline section inner annular flow containing rate ε₂For:

${\mathrm{\ϵ}}_2=\frac{V_{gas2}}{V_{pipe}}=\frac{\underset{i=1}{\overset{L}{\mathrm{\Σ}}}{\mathrm{\α}}_i{\mathrm{\β}}_i{a}_i{b}_i}{{\mathrm{LD}}^2}$

Wherein, α_iFor a_iCorresponding correction coefficient, β_iFor b_iCorresponding correction coefficient.

Take the N group image construction image sequence being continuously shot, calculate the phase content that often group image is corresponding, then averageAs the measured value of annular flow phase content under corresponding experiment condition, it may be assumed that

${\stackrel{\‾}{\mathrm{\ϵ}}}_{annular}=\frac{\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\ϵ}}_k}{N}$

Wave flow phase content measurement model: under wave flow flow pattern, in pipeline, the volume of gas phase can be obtained by the volume superposition of several cylinders.Different from other three kinds of typical flow patterns, owing to can uniquely determine the area of section shared by gas phase and liquid phase after determining height shared by gas phase and liquid phase, therefore in horizontal pipe, the image of the single angle that the phase content of wave flow can be shot by pipeline side is determined.As it is shown in fig. 7, in acquired wave flow image, the height in gas phase cross section, ith pixel place is h_i, then the volume V of gas phase in single image correspondence pipeline section_gas3For:

$V_{gas3}={\mathrm{\μ}}^3\underset{i=1}{\overset{L}{\mathrm{\Σ}}}\[\frac{\mathrm{\π}}{4}{D}^2-\frac{D^2}{4}\arccos (\frac{2h_i}{D}-1)+(h_i-\frac{D}{2})\sqrt{h_{i}D-h_i^2}\]$

The then phase content ε of wave flow in single image correspondence pipeline section₃For:

${\mathrm{\ϵ}}_3=\frac{V_{gas3}}{V_{pipe}}=\frac{1}{L}\underset{i=1}{\overset{L}{\mathrm{\Σ}}}\[1-\frac{1}{\mathrm{\π}}\arccos (\frac{2h_i}{D}-1)+\frac{2}{\mathrm{\π}D}(2h_i-D)\sqrt{h_{i}D-h_i^2}\]$

Take the N width image construction image sequence being continuously shot, calculate the phase content that each image is corresponding, then averageAs the measured value of annular flow phase content under corresponding experiment condition, it may be assumed that

${\stackrel{\‾}{\mathrm{\ϵ}}}_{wavy}=\frac{\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\ϵ}}_k}{N}.$

Accompanying drawing explanation

Accompanying drawing 1 is the present invention passage aisle gas-liquid two-phase flow containing rate measurement system schematic based on multi-visual information integration technology, in Fig. 1, and 1-passage aisle pipe-line system, 2-many visions high-speed image sampling unit, 3-microcomputer.

Accompanying drawing 2 is many visions high-speed image sampling unit and passage aisle pipe-line system structural representation, in Fig. 2, and 4-the first LED light source, 5-the second LED light source, 6-glycerol groove, 7-passage aisle, 8-corner cube prism, 9-high-speed camera.

Accompanying drawing 3 is the index path that light is propagated in passage aisle and glycerol groove.

Accompanying drawing 4 is based on the passage aisle gas-liquid two-phase flow containing rate measurement procedure figure of many visions.

Accompanying drawing 5 is bubble flow and slug flow phase content measurement model schematic diagram.

Accompanying drawing 6 is annular flow phase content measurement model schematic diagram.

Accompanying drawing 7 is wave flow phase content measurement model schematic diagram.

Accompanying drawing 8 is the phase content measurement result figure of the horizontal pipe of 4.0mm internal diameter and vertical pipeline.

Detailed description of the invention

With reference to accompanying drawing 1, the present invention utilizes the passage aisle gas-liquid two-phase flow containing rate based on many visions to measure system, it is achieved passage aisle gas-liquid two-phase flow containing rate is measured.Passage aisle gas-liquid two-phase flow containing rate measurement system based on many visions includes passage aisle pipe-line system, many visions high-speed image sampling unit and microcomputer three part.

As shown in Figure 2, passage aisle pipe-line system 1 is made up of the passage aisle 7 being immersed in glycerol groove 6, many visions high-speed image sampling unit 2 is made up of high-speed camera 9, corner cube prism the 8, first LED light source the 4, second LED light source 5, many visions high-speed image sampling unit is responsible for two biphase gas and liquid flow flow images being mutually perpendicular to angle of collection simultaneously, many visions high-speed image sampling unit 2 image transmitting collected carries out image procossing to computer 3 and phase content calculates.

The first described LED light source 4 is perpendicular to passage aisle 6 and places, second LED light source 5 is both perpendicular to the first LED light source 5 and passage aisle 6 and in being generally aligned in the same plane with the first LED light source 5, corner cube prism 8 is positioned at passage aisle 6 other side relative to the first LED light source 4, high-speed camera 9 is positioned at passage aisle 6 other side relative to the second LED light source 5, and a right-angle side of corner cube prism 8 is perpendicular to the imaging plane of high-speed camera 6.

With reference to accompanying drawing 4, carry out phase content calculating according to the passage aisle gas-liquid two-phase flow containing rate measurement procedure figure based on many visions, concretely comprise the following steps:

The first step, obtains image: utilize many visions high-speed image sampling unit, obtains two passage aisle biphase gas and liquid flow flow image sequences being mutually perpendicular to angle simultaneously；

Second step, sets up phase content measurement model: carry out Image semantic classification and optic aberrance revising, and four kinds of typical flow patterns are modeled respectively；

3rd step, meteor trail echoes: extract the characteristic vector of image, carry out meteor trail echoes；

4th step, calculates phase content: select the phase content measurement model of corresponding flow pattern, calculates phase content.

Wherein, second step sets up concretely comprising the following steps of phase content measurement model:

1) Image semantic classification: filled by difference shadow method, medium filtering, rim detection and binaryzation and carry out pretreatment, obtain bianry image.

2) optic aberrance revising: obtain correction coefficient by optical path analysis, it is achieved the correction to bianry image.Adopting cross section is foursquare lucite cuboid glycerol groove, is immersed in by pipeline in glycerol groove, and the center of pipeline center position and cross-sectional square shape coincides.For object point P in pipeline, (x, y), this point there is first time refraction at liquid phase and tube wall interface in the light sent, and in glycerol groove cell wall and Air Interface, second time refraction occurs.Two adjacent ray sent by P intersection point P'(x', the y' of reverse extending line after second time refraction) be a P to it should be observed that picture point, may thereby determine that the correction coefficient alpha=y/y' at a P place.After obtaining correction coefficient, to correct image.

3) four kinds of typical flow pattern modelings: use for reference the thought of " segmentation, approximate summation ", set up bubble flow and slug flow phase content measurement model, annular flow phase content measurement model, wave flow phase content measurement model respectively.

Bubble flow and slug flow phase content measurement model: owing to the hydraulic diameter of passage aisle is less, the same cross section of pipeline only exists a bubble simultaneously, do not have same cross section in conventional pipeline and have the situation of multiple bubble, it is possible to determine the size of bubble in both direction by the image of two angles.Consider the situation of bubble flow and slug flow, at the same section of pipeline, be oval by region equivalent shared by gas phase.Within the scope of the axial length in pixels of pipeline, think that the area of section shared by gas phase is equal, therefore gas phase in this segment limit can be equivalent to an Elliptic Cylinder, and the major and minor axis of this Elliptic Cylinder bottom surface ellipse is determined by two bianry images being mutually perpendicular to angle respectively.Thus, it is possible to each bubble or slug to be equivalent to the superposition of several Elliptic Cylinders, each bubble or the volume of gas phase corresponding to vent plug can be tried to achieve.

As shown in fig. 5, it is assumed that captured single set of image exists n bubble/vent plug.Number of pixels shared by i-th bubble/vent plug length is l_i, number of pixels respectively a shared by the major and minor axis of the jth pixel place bubble section ellipse that i-th bubble/vent plug is corresponding_ijWith b_ij, then the volume V of gas phase in captured image correspondence pipeline section can be tried to achieve_gas1For:

$V_{gas1}=\frac{\mathrm{\π}}{4}{\mathrm{\μ}}^3\underset{i=1}{\overset{n}{\mathrm{\Σ}}}\underset{j=1}{\overset{l_i}{\mathrm{\Σ}}}{\mathrm{\α}}_{ij}{\mathrm{\β}}_{ij}{a}_{ij}{b}_{ij}$

Wherein, μ is the physical length that each square pixel is corresponding, α_ijFor a_ijCorresponding correction coefficient, β_ijFor b_ijCorresponding correction coefficient.The then phase content ε of bubble flow/slug flow in corresponding pipeline section₁For:

${\mathrm{\ϵ}}_1=\frac{V_{gas1}}{V_{pipe}}=\frac{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}\underset{j=1}{\overset{l_i}{\mathrm{\Σ}}}{\mathrm{\α}}_{ij}{\mathrm{\β}}_{ij}{a}_{ij}{b}_{ij}}{{\mathrm{LD}}^2}$

Wherein, V_pipeFor the conduit volume that image taking arrives, L is the duct length that image taking arrives, and D is the pipe diameter that image taking arrives, and the physical length μ that phase content is corresponding with pixel is unrelated, and only to each dimensional parameters of two phase flow image shared by the number of pixel relevant.

In actual photographed process, owing to length of pipe section taken in viewing field of camera is shorter, the calculated phase content of single set of image can not accurately reflect the phase content under current flow pattern.Therefore, choose the image construction image sequence being continuously shot by N group, take the meansigma methods of image sequence phase contentAs the phase content measured value of bubble flow/slug flow under corresponding experiment condition, it may be assumed that

${\stackrel{\‾}{\mathrm{\ϵ}}}_{bubble/slug}=\frac{\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\ϵ}}_k}{N}$

Annular flow phase content measurement model: in horizontal pipe, by the impact of action of gravity, duct bottom liquid film is thicker；And liquid film is distributed around even tube wall in vertical pipeline.Similar with slug flow and bubble flow, under annular flow flow pattern, gas phase in pipeline is approximately the superposition of several Elliptic Cylinders, gaseous phase volume can be tried to achieve accordingly, it is hereby achieved that phase content.As shown in Figure 6, in acquired annular flow bianry image, number of pixels respectively a shared by the major and minor axis that gas phase cross section, ith pixel place is oval_iWith b_i, then single set of image correspondence pipeline section inner annular flow containing rate ε₂For:

${\mathrm{\ϵ}}_2=\frac{V_{gas2}}{V_{pipe}}=\frac{\underset{i=1}{\overset{L}{\mathrm{\Σ}}}{\mathrm{\α}}_i{\mathrm{\β}}_i{a}_i{b}_i}{{\mathrm{LD}}^2}$

Wherein, α_iFor a_iCorresponding correction coefficient, β_iFor b_iCorresponding correction coefficient.

Take the N group image construction image sequence being continuously shot, calculate the phase content that often group image is corresponding, then averageAs the measured value of annular flow phase content under corresponding experiment condition, it may be assumed that

${\stackrel{\‾}{\mathrm{\ϵ}}}_{annular}=\frac{\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\ϵ}}_k}{N}$

Wave flow phase content measurement model: under wave flow flow pattern, in pipeline, the volume of gas phase can be obtained by the volume superposition of several cylinders.Different from other three kinds of typical flow patterns, owing to can uniquely determine the area of section shared by gas phase and liquid phase after determining height shared by gas phase and liquid phase, therefore in horizontal pipe, the image of the single angle that the phase content of wave flow can be shot by pipeline side is determined.As it is shown in fig. 7, in acquired wave flow image, the height in gas phase cross section, ith pixel place is h_i, then the volume V of gas phase in single image correspondence pipeline section_gas3For:

$V_{gas3}={\mathrm{\μ}}^3\underset{i=1}{\overset{L}{\mathrm{\Σ}}}\[\frac{\mathrm{\π}}{4}{D}^2-\frac{D^2}{4}\arccos (\frac{2h_i}{D}-1)+(h_i-\frac{D}{2})\sqrt{h_{i}D-h_i^2}\]$

The then phase content ε of wave flow in single image correspondence pipeline section₃For:

${\mathrm{\ϵ}}_3=\frac{V_{gas3}}{V_{pipe}}=\frac{1}{L}\underset{i=1}{\overset{L}{\mathrm{\Σ}}}\[1-\frac{1}{\mathrm{\π}}\arccos (\frac{2h_i}{D}-1)+\frac{2}{\mathrm{\π}D}(2h_i-D)\sqrt{h_{i}D-h_i^2}\]$

Take the N width image construction image sequence being continuously shot, calculate the phase content that each image is corresponding, then averageAs the measured value of annular flow phase content under corresponding experiment condition, it may be assumed that

${\stackrel{\‾}{\mathrm{\ϵ}}}_{wavy}=\frac{\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\ϵ}}_k}{N}.$

Utilize in the present invention propose system and method internal diameter be 4.0mm horizontally and vertically transparent passage aisle pipe-line system in carried out phase content experiments of measuring.

In experiment, the excursion of gas phase flow rate is 6ml/min-41700ml/min, liquid phase changes in flow rate ranges for 6ml/min-4170ml/min, having carried out the phase content under bubble flow, slug flow, annular flow and four kinds of typical flow patterns of wave flow in horizontal pipe to measure, the phase content having carried out other three kinds of typical flow patterns except wave flow in vertical pipeline is measured.Adopt the phase content that fast valve method records as calibration value.

As shown in Figure 8, a () figure is the phase content experiments of measuring result in 4.0mm horizontal pipe, b () figure is the phase content experiments of measuring result in 4.0mm vertical pipeline, it can be seen that horizontally and vertically the phase content in small pipeline is measured absolute error and is respectively less than 6%.Test result indicate that, it is feasible, effective that the passage aisle gas-liquid two-phase flow containing rate based on many visions proposed by the invention measures system and method.

Claims (5)

1. the passage aisle gas-liquid two-phase flow containing rate based on many visions measures system, it is characterized in that: this system is by passage aisle pipe-line system (1), many visions high-speed image sampling unit (2) and microcomputer (3) composition, passage aisle pipe-line system (1) is made up of the passage aisle (7) being immersed in glycerol groove (6), many visions high-speed image sampling unit (2) are by high-speed camera (9), corner cube prism (8), first LED light source (4), second LED light source (5) forms, many visions high-speed image sampling unit is responsible for two biphase gas and liquid flow flow images being mutually perpendicular to angle of collection simultaneously, the image transmitting collected by many visions high-speed image sampling unit (2) carries out image procossing to computer (3) and phase content calculates；Described the first LED light source (4) is perpendicular to passage aisle (6) and places, second LED light source (5) is both perpendicular to the first LED light source (5) and passage aisle (6) and in being generally aligned in the same plane with the first LED light source (5), corner cube prism (8) is positioned at passage aisle (6) other side relative to the first LED light source (4), high-speed camera (9) is positioned at passage aisle (6) other side relative to the second LED light source (5), and a right-angle side of corner cube prism (8) is perpendicular to the imaging plane of high-speed camera (6).

2. the passage aisle gas-liquid two-phase flow containing rate measuring method based on many visions based on system described in claim 1, it is characterised in that the method comprises the steps:

1) obtain image: utilize many visions high-speed image sampling unit, obtain two passage aisle biphase gas and liquid flow flow image sequences being mutually perpendicular to angle simultaneously；

2) set up phase content measurement model: carry out Image semantic classification and optic aberrance revising, four kinds of typical flow patterns are modeled respectively；

3) meteor trail echoes: extract the characteristic vector of image, carry out meteor trail echoes；

4) calculate phase content: select the phase content measurement model of corresponding flow pattern, calculate phase content.

3. phase content measuring method according to claim 2, it is characterised in that described step 2) in the method setting up phase content measurement model particularly as follows:

1) Image semantic classification: filled by image segmentation, image denoising, rim detection and binaryzation and carry out pretreatment, obtain bianry image；

2) optic aberrance revising: obtain correction coefficient by optical path analysis, it is achieved the correction to bianry image；

3) four kinds of typical flow pattern modelings: set up bubble flow and slug flow phase content measurement model, annular flow phase content measurement model, wave flow phase content measurement model respectively.

4. according to Claims 2 or 3, set up phase content measurement model method, it is characterized in that described step 2) in optic aberrance revising method particularly as follows: adopt cross section be foursquare lucite cuboid glycerol groove, pipeline is immersed in glycerol groove, the center making pipeline center position and cross-sectional square shape coincides, for object point P (x in pipeline, y), first time refraction is there is in the light sent by this point at liquid phase and tube wall interface, second time refraction is there is in glycerol groove cell wall and Air Interface, analysis light path is known, the intersection point P'(x' of two adjacent ray sent by P reverse extending line after second time refraction, y') picture point corresponding to a P it is, so that it is determined that the correction coefficient alpha=y/y' at some P place, after obtaining correction coefficient, bianry image is corrected.

5. according to Claims 2 or 3, set up phase content measurement model method, it is characterised in that described step 3) in four kinds of typical flow patterns modelings method particularly as follows:

1) bubble flow and slug flow phase content measurement model: the hydraulic diameter in passage aisle is less, the same cross section of pipeline only exists a bubble simultaneously, the size of bubble in both direction can be determined by the image of two angles, consider the situation of bubble flow and slug flow, at the same section of pipeline, it is oval by region equivalent shared by gas phase, within the scope of the axial length in pixels of pipeline, think that the area of section shared by gas phase is equal, therefore gas phase in this segment limit can be equivalent to an Elliptic Cylinder, the major and minor axis of this Elliptic Cylinder bottom surface ellipse is determined by two bianry images being mutually perpendicular to angle respectively, thus, each bubble or slug can be equivalent to the superposition of several Elliptic Cylinders, each bubble or the volume of gas phase corresponding to vent plug can be tried to achieve,

Assuming there is n bubble/vent plug in captured single set of image, number of pixels shared by i-th bubble/vent plug length is l_i, number of pixels respectively a shared by the major and minor axis of the jth pixel place bubble section ellipse that i-th bubble/vent plug is corresponding_ijWith b_ij, then the volume V of gas phase in captured image correspondence pipeline section can be tried to achieve_gas1For:

$V_{gas1}=\frac{\mathrm{\π}}{4}{\mathrm{\μ}}^3\underset{i=1}{\overset{n}{\Σ}}\underset{j=1}{\overset{l_i}{\Σ}}{\mathrm{\α}}_{ij}{\mathrm{\β}}_{ij}{a}_{ij}{b}_{ij}$

Wherein, μ is the physical length that each square pixel is corresponding, α_ijFor a_ijCorresponding correction coefficient, β_ijFor b_ijCorresponding correction coefficient, then the phase content ε of bubble flow/slug flow in corresponding pipeline section₁For:

${\mathrm{\ϵ}}_1=\frac{V_{gas1}}{V_{pipe}}=\frac{\underset{i=1}{\overset{n}{\Σ}}\underset{j=1}{\overset{l_i}{\Σ}}{\mathrm{\α}}_{ij}{\mathrm{\β}}_{ij}{a}_{ij}{b}_{ij}}{{\mathrm{LD}}^2}$

Wherein, V_pipeFor the conduit volume that image taking arrives, L is the duct length that image taking arrives, and D is the pipe diameter that image taking arrives, and the physical length μ that phase content is corresponding with pixel is unrelated, and only to each dimensional parameters of two phase flow image shared by the number of pixel relevant；

In actual photographed process, owing to length of pipe section taken in viewing field of camera is shorter, the calculated phase content of single set of image can not accurately reflect the phase content under current flow pattern, therefore, choose the image construction image sequence being continuously shot by N group, take the meansigma methods of image sequence phase contentAs the phase content measured value of bubble flow/slug flow under corresponding experiment condition, it may be assumed that

${\stackrel{\‾}{\mathrm{\ϵ}}}_{bubble/slug}=\frac{\underset{k=1}{\overset{N}{\Σ}}{\mathrm{\ϵ}}_k}{N}$

2) annular flow phase content measurement model: in horizontal pipe, by the impact of action of gravity, duct bottom liquid film is thicker；And liquid film is distributed around even tube wall in vertical pipeline, similar with slug flow and bubble flow, under annular flow flow pattern, gas phase in pipeline is approximately the superposition of several Elliptic Cylinders, gaseous phase volume can be tried to achieve accordingly, it is hereby achieved that phase content, in acquired annular flow bianry image, number of pixels respectively a shared by the major and minor axis that gas phase cross section, ith pixel place is oval_iWith b_i, then single set of image correspondence pipeline section inner annular flow containing rate ε₂For:

${\mathrm{\ϵ}}_2=\frac{V_{gas2}}{V_{pipe}}=\frac{\underset{i=1}{\overset{L}{\Σ}}{\mathrm{\α}}_i{\mathrm{\β}}_i{a}_i{b}_i}{{\mathrm{LD}}^2}$

Wherein, α_iFor a_iCorresponding correction coefficient, β_iFor b_iCorresponding correction coefficient；

Take the N group image construction image sequence being continuously shot, calculate the phase content that often group image is corresponding, then averageAs the measured value of annular flow phase content under corresponding experiment condition, it may be assumed that

${\stackrel{\‾}{\mathrm{\ϵ}}}_{annular}=\frac{\underset{k=1}{\overset{N}{\Σ}}{\mathrm{\ϵ}}_k}{N}$

3) wave flow phase content measurement model: under wave flow flow pattern, in pipeline, the volume of gas phase can be obtained by the volume superposition of several cylinders, different from other three kinds of typical flow patterns, owing to can uniquely determine the area of section shared by gas phase and liquid phase after determining height shared by gas phase and liquid phase, therefore in horizontal pipe, the image of the single angle that the phase content of wave flow can be shot by pipeline side is determined, in acquired wave flow image, the height in gas phase cross section, ith pixel place is h_i, then the volume V of gas phase in single image correspondence pipeline section_gas3For:

$V_{gas3}={\mathrm{\μ}}^3\underset{i=1}{\overset{L}{\Σ}}\[\frac{\mathrm{\π}}{4}{D}^2-\frac{D^2}{4}\arccos (\frac{2h_i}{D}-1)+(h_i-\frac{D}{2})\sqrt{h_{i}D-h_i^2}\]$

The then phase content ε of wave flow in single image correspondence pipeline section₃For:

${\mathrm{\ϵ}}_3=\frac{V_{gas3}}{V_{pipe}}=\frac{1}{L}\underset{i=1}{\overset{L}{\Σ}}\[1-\frac{1}{\mathrm{\π}}\arccos (\frac{2h_i}{D}-1)+\frac{2}{\mathrm{\π}D}(2h_i-D)\sqrt{h_{i}D-h_i^2}\]$

Take the N width image construction image sequence being continuously shot, calculate the phase content that each image is corresponding, then averageAs the measured value of annular flow phase content under corresponding experiment condition, it may be assumed that

${\stackrel{\‾}{\mathrm{\ϵ}}}_{wavy}=\frac{\underset{k=1}{\overset{N}{\Σ}}{\mathrm{\ϵ}}_k}{N}.$