CN105954163A - On-line measuring method for sedimentation rate and denudation rate of dirt in condenser - Google Patents

Abstract

The invention provides an on-line measuring method for the sedimentation rate and denudation rate of dirt in a condenser. The method comprises the following steps: S1, acquiring an image sequence in the process of scaling of a tube wall; S2, recognizing sedimentation and denudation particles: analyzing the acquired image sequence, recognizing dirt particles, carrying out matching between interframe dirt particles and tracing the movement track of the dirt particles in real time; S3, establishing a three-dimensional geometrical model for the particles: calculating the position and distribution of the dirt particles on the tube wall according to the three-dimensional movement track and orientation gradient of the dirt particles, carrying out three-dimensional reconstruction on the dirt particles and the movement track thereof and then establishing the three-dimensional geometrical model for the particles; and S4, calculating the volume of sedimentation particles and the volume of denudation particles so as to obtain the sedimentation rate and denudation rate of dirt. The method provided by the invention has the advantages of good instantaneity, high accuracy, capacity of visually reflect the dynamic changing process of dirt in the tube of the condenser, etc.

Description

Dirt deposition rate and the On-line Measuring Method of denudation rate in condenser

Technical field

Present invention relates generally to condenser field, refer in particular to the online survey of a kind of dirt deposition rate applied in condenser and denudation rate Metering method.

Background technology

At present, condenser is used widely in each big fields such as electric power, chemical industry, metallurgy, machineries.But, in condenser by In the existence of dirt, greatly reduce the heat-transfer capability of heat-transfer surface so that steam turbine exhaust pressure and exhaust temperature raise, output Power reduces, and heat consumption rate increases, thus causes Unit Economic decline in benefits.Additionally, the accumulation of dirt, the most usually cause condensation The local corrosion of pipe or even penetrate, the safe operation of serious threat to condenser.Just because of this, it is necessary to take effective dirt control Measure processed.

Condenser scaling process is the result that dirt deposition and dirt degrade interaction, and deposition makes dirt increase, and degrades and makes dirt Reducing, deposition is dirt scaling rate with the difference of denudation rate, and the essence that dirt controls is to control deposition, denudation rate, And then reach to control the purpose of scaling rate.Therefore, for Fouling in Condenser control, correct method should be to sink according to dirt Long-pending rate, the dynamically change of denudation rate are adaptively adjusted controlled quentity controlled variable.But, owing to being difficult to precisely in real time the interior dirt that holds a writing brush or pen-write Dynamic change, and Fouling in Condenser is the mixture of multiple particulate fouling, so that online to dirt deposition rate, denudation rate The research difficulty of the method measured strengthens, and makes slow progress, does not the most all have good method to be suggested.

Summary of the invention

The technical problem to be solved in the present invention is that the technical problem existed for prior art, and the present invention provides a kind of real-time Property is good, accuracy rate is high, can intuitively reflect condenser tube in dirt dynamic changing process condenser in dirt deposition rate and degrading The On-line Measuring Method of rate.

For solve above-mentioned technical problem, the present invention by the following technical solutions:

In a kind of condenser, dirt deposition rate and the On-line Measuring Method of denudation rate, the steps include:

S1: obtain the image sequence of tube wall incrustation process；

S2: identify deposition and degrade granule；The image sequence collected is analyzed, identifies dirt particles；Carry out interframe The coupling of dirt particles, completes the real-time tracing to dirt particles movement locus；

S3: set up the 3-D geometric model of granule；According to dirt particles 3 D motion trace and direction gradient, try to achieve dirt particles In position and the distribution of tube wall, dirt particles and movement locus thereof are carried out three-dimensionalreconstruction, sets up the 3-D geometric model of granule；

S4: try to achieve deposited particles volume and with degrade granule volume and, and then obtain dirt deposition rate and denudation rate.

Further improvement as the inventive method: in described step S4, according to the 3-D geometric model of dirt particles, try to achieve The volume of deposited particles and degrade the volume of granule, dirt deposition rate, denudation rate are then deposition and the stripping of dirt in the unit time The volume of erosion amount, i.e. deposited particles and with degrade granule volume and.

Further improvement as the inventive method: be by the figure in image collecting device Real-time Collection tube wall in described step S1 Picture, obtains the image sequence of tube wall incrustation process.

Further improvement as the inventive method: in described step S2, identifies that the method for granule is: use dimensional Gaussian before this Function describes the intensity profile of dirt particles, obtains particle geometric form parameter by granule intensity profile carries out Function Fitting, Geometric shape parameters according to granule identifies granule.

Further improvement as the inventive method: in described step S2, the method for grain matches is based on geometric properties similarity The method combined with Normalized moment of inertia feature similarity.

Further improvement as the inventive method: in described step S3, dirt particles and the three-dimensionalreconstruction side of dirt movement locus Method is: before this based on binocular vision, calculates the locus of granule, it is achieved the three-dimensionalreconstruction of granule；Afterwards combine before Grain, in the motion track information of each interframe, carries out the three-dimensionalreconstruction of particle motion trajectory.

Compared with prior art, it is an advantage of the current invention that:

1, dirt deposition rate and the On-line Measuring Method of denudation rate in the condenser of invention, principle is simple, easy and simple to handle, Real-time is high, it is possible to really realize dirt deposition rate, the on-line measurement of denudation rate.

2, dirt deposition rate and the On-line Measuring Method of denudation rate in the condenser of the present invention, accuracy is high, can react intuitively The dynamic changing process of dirt in condenser tube, thus accurately show that dirt particles is in the position of tube wall and distribution.

3, dirt deposition rate and the On-line Measuring Method of denudation rate in the condenser of the present invention, easily realize, the most reliably, and required one-tenth This is low, and can apply in the middle of complex environment, is conducive to the control of the interior dirt of pipe, improves operational efficiency, easily promote, increases Social benefit.

Accompanying drawing explanation

Fig. 1 is the schematic flow sheet of the inventive method.

Fig. 2 is the principle schematic that the present invention carries out when particle mates in concrete application example.

Fig. 3 is the schematic diagram that the present invention carries out three dimensional space coordinate reconstruct in concrete application example to particle.

Fig. 4 be the present invention when dirt particles being mated in concrete application example interframe search the principle schematic of flow process.

Fig. 5 is present invention schematic diagram of image collecting device employed in concrete application example.

Detailed description of the invention

Below with reference to Figure of description and specific embodiment, the present invention is described in further details.

As it is shown in figure 1, dirt deposition rate and the On-line Measuring Method of denudation rate in the condenser of the present invention, step is:

S1: obtain the image sequence of tube wall incrustation process；

By the image in image collecting device (such as: photographic head) Real-time Collection tube wall, obtain the image sequence of tube wall incrustation process Row；

S2: identify deposition and degrade granule；

The image sequence collected is analyzed, identifies dirt particles；Carry out the coupling of interframe dirt particles, thus realize Real-time tracing to dirt particles movement locus.

S3: set up the 3-D geometric model of granule；

According to dirt particles 3 D motion trace and direction gradient, dirt particles can be tried to achieve in the position of tube wall and distribution；That is, right Dirt particles and movement locus thereof carry out three-dimensionalreconstruction, set up the 3-D geometric model of granule；

S4: try to achieve deposited particles volume and with degrade granule volume and, and then obtain dirt deposition rate and denudation rate；

According to the 3-D geometric model of dirt particles, try to achieve deposited particles volume and with degrade the volume of granule, obtain dirt and sink Long-pending rate and denudation rate.Wherein, dirt deposition rate, denudation rate refer to deposition and the denudation of dirt in the unit interval, i.e. deposit The volume of granule and with degrade granule volume and, the dirt 3-D geometric model according to setting up can try to achieve particle volume.

The image information utilizing binocular vision to transmit calculates the locus of granule, in conjunction with before real-time tracing (two to granule Dimension tracing path), it is achieved the three-dimensionalreconstruction of particle motion trajectory.Set up the 3-D geometric model of granule, try to achieve particle volume, from And realize the on-line measurement of dirt deposition rate, denudation rate.

In above-mentioned steps S2, dirt particles movement locus real-time tracing it is critical only that the identification of granule, coupling.In the present invention first Being the intensity profile using two-dimensional Gaussian function to describe dirt particles, obtaining granule by granule intensity profile being carried out Function Fitting Geometric shape parameters, identifies granule according to the geometric shape parameters of granule, and the algorithm realizing grain matches is to propose a kind of base In the algorithm that geometric properties similarity and Normalized moment of inertia feature similarity combine.

Further, the present invention is directed to the granule overlap run in particle recognition, matching process, the problem being blocked, propose one Plant interframe edge prediction algorithm, interframe Gray Correlation correction-compensation algorithm.As for the particle flocculation metaboly existed, then adopt With cluster predictor method.

In above-mentioned steps S3, dirt particles and the three-dimensionalreconstruction of dirt movement locus, before this based on binocular vision, calculate The locus of grain, it is achieved the three-dimensionalreconstruction of granule.Combine above granule afterwards, in the motion track information of each interframe, to carry out Three maintenance three-dimensionalreconstructions of grain movement locus.

In concrete application example, dirt deposition rate and the On-line Measuring Method of denudation rate, detailed step in the condenser of the present invention For:

S100: obtain the image sequence of tube wall incrustation process；

As, by camera collection image, thus obtain the image sequence of tube wall incrustation process.

S200: utilize image sequence, it is achieved the three-dimensionalreconstruction of particle motion trajectory；

That is: image sequence is processed, extract the dirt particles in every two field picture, and each frame binocular camera is obtained Three-dimensionalreconstruction is carried out to obtain its locus to dirt particles image.

S300: dirt point in image sequence is carried out real-time tracing, thus obtains the motion path of dirt particles in image sequence.

S400: definition dirt deposition rate and denudation rate are deposition and the denudation of dirt in the unit time.

Seeing Fig. 2～Fig. 4, when specifically applying, the idiographic flow of step S200 is as follows:

S201: first the position of dirt particles is carried out three-dimensionalreconstruction, set up the world coordinate system that dirt particles locus is relied on, It is respectively R with spin matrix and the translation matrix of left camera coordinates system_LAnd T_L, it is with the spin matrix peace of right camera coordinates system Move matrix and be respectively R_RAnd T_R, these matrixes are all fixed numbers matrixes, can be obtained by demarcating.

S202: setting magazine particle image P' in left and right and P is a pair coupling particle, its particle three-dimensional space coordinates three-dimensionalreconstruction is several What principle as it is shown on figure 3, wherein I and I' be respectively the image center of left and right camera, the three-dimensional space position at particle P place is i.e. It it is the intersection point of two straight line PI and P'I'.

If the space coordinates that Space Particle G is in defined world coordinate system isAnd it is magazine in left and right Particle image is respectively P and P', and they coordinates in world coordinate system are set toWith Because easily trying to achieve particle image P and P' respectively at the coordinate of left and right camera coordinates system, it is set to (X_P,Y_P,Z_P) and (X_P',Y_P',Z_P'),

Therefore can get formula calculated as below:

${(X_P^W,Y_P^W,Z_P^W)}^T={(X_P,Y_P,Z_P)}^T{R}_L+T_L---\left(1\right)$

${(X_{P^{,}}^W,Y_{P^{,}}^W,Z_{P^{,}}^W)}^T={(X_{P^{,}},Y_{P^{,}},Z_{P^{,}})}^T{R}_R+T_R---\left(2\right)$

It is apparent from again left and right image center I and I' and is respectively (0,0, Z at the coordinate of left and right camera coordinates system_I) and (0,0, Z_I'), it is converted For above-mentioned world coordinates, set respectively they world coordinate system coordinate asWithThen coordinate Conversion formula is:

${(X_I^W,Y_I^W,Z_I^W)}^T={(0,0,Z_I)}^T{R}_L+T_L---\left(3\right)$

${(X_{I^{,}}^W,Y_{I^{,}}^W,Z_{I^{,}}^W)}^T={(0,0,Z_{I^{,}})}^T{R}_R+T_R---\left(4\right)$

Having been all switched to point by above-mentioned calculating formula, in the world coordinate system of definition, be apparent from, straight line PI and GI is with always Line, the expression formula in world coordinate system is identical.Equally, P'I' and GI' is also same straight line, has identical expression formula.Therefore The coordinate of Space Particle G can be by such as following formula (5) to (9) calculating formula gained:

$X_G^W=X_I^W+{\mathrm{\μ}}_a(X_P^W-X_I^W)---\left(5\right)$

$Y_G^W=Y_I^W+{\mathrm{\μ}}_a(Y_P^W-Y_I^W)---\left(6\right)$

$Z_G^W=Z_I^W+{\mathrm{\μ}}_b(Z_P^W-Z_I^W)---\left(7\right)$

Wherein μ_aAnd μ_bHave respectively:

${\mathrm{\μ}}_a=\frac{(X_{P^{,}}^W-X_{I^{,}}^W)(Y_I^W-Y_{I^{,}}^W)-(Y_{P^{,}}^W-Y_{I^{,}}^W)(X_I^W-X_{I^{,}}^W)}{(Y_{P^{,}}^W-Y_{I^{,}}^W)(X_P^W-X_I^W)-(X_{P^{,}}^W-X_{I^{,}}^W)(Y_P^W-Y_I^W)}---\left(8\right)$

${\mathrm{\μ}}_h=\frac{(X_{P^{,}}^W-X_{I^{,}}^W)(Z_I^W-Z_{I^{,}}^W)-(Z_{P^{,}}^W-Z_{I^{,}}^W)(X_I^W-X_{I^{,}}^W)}{(Z_{P^{,}}^W-Z_{I^{,}}^W)(X_P^W-X_I^W)-(X_{P^{,}}^W-X_{I^{,}}^W)(Z_P^W-Y_I^W)}---\left(9\right)$

When specifically applying, the idiographic flow of step S300 is as follows:

As shown in Figure 4, to dirt particles position and the dirt particles image that extracts in the image sequence obtained in above-mentioned steps S2 Process, to the first two field picture detection dirt particles and location point thereof in image sequence, attempt in the next frame following the tracks of these Grain, i.e. finds these dirt particles position in new picture frame, the image sequence that so tracking is complete, as shown in Figure 4, Image I₁And I₂It it is consecutive frame.Meanwhile, the dirt particles without coupling is marked according to its present position and movement tendency, district It is divided into deposited particles and degrades particle.

After completing according to aforesaid operations, according to actual needs, can select granulomere to be divided into deposited particles, degrades granule, escape Granule.Wherein, escaping particles has coupling granule and without deposition tendency in referring to all frames.Owing to dirt particles point can move with current Dynamic, and the phenomenon such as rotate.Therefore, in order at frame matching dirt particles, the present invention adopts the following technical scheme that:

First, determining query region in the next frame according to work condition environment, in application initial frame, particle characteristic parameter carries out similar Joining, wherein characteristic parameter includes color, geometry, picture quality, image rotary inertia, frontal projected area and angle point Position relatively.When completing flow process as shown in Figure 4 and carrying out particle characteristic compatible portion, need the coupling granule according to a new frame The characteristic parameter of granule is updated, has obtained the complete characterization of dirt particles.

Wherein, picture quality and image rotary inertia scheme are as follows:

The two width images obtained by dual camera carry out gray scale stretch processing to extract the binary image of dirt particles.Such as Fig. 5 institute Showing, in concrete application example, image collecting device includes light source 1, computer 3 and binocular camera 4, and wherein light source 1 has Cross section 2 as schemed, above-mentioned image collecting device is used for monitoring condensing tube 5 in real time.Wherein, the gray scale of pixel Value should be with the far and near negative correlation of distance light source, i.e. distance light source is the most remote, and gray value is the lowest.If the image slices of the dirt particles extracted Element regional extent is 1≤x≤M, 1≤y≤N, and (x y) represents respective pixel gray value to make f.

Definition average gray is picture quality m:

$m=\underset{x=1}{\overset{M}{\Σ}}\underset{y=1}{\overset{N}{\Σ}}f(x,y)---\left(10\right)$

The image reform of definition dirt particlesSolve mode as follows:

$\stackrel{\‾}{x}=\left(\underset{x=1}{\overset{M}{\Σ}}\underset{y=1}{\overset{N}{\Σ}}xf\right(x,y\left)\right)/\left(\underset{x=1}{\overset{M}{\Σ}}\underset{y=1}{\overset{N}{\Σ}}f\right(x,y\left)\right)---\left(11\right)$

$\stackrel{\‾}{y}=\left(\underset{x=1}{\overset{M}{\Σ}}\underset{y=1}{\overset{N}{\Σ}}yf\right(x,y\left)\right)/\left(\underset{x=1}{\overset{M}{\Σ}}\underset{y=1}{\overset{N}{\Σ}}f\right(x,y\left)\right)---\left(12\right)$

The image rotary inertia J of definition dirt particles:

$\stackrel{\‾}{x}=\underset{x=1}{\overset{M}{\Σ}}\underset{y=1}{\overset{N}{\Σ}}\[{(x-\stackrel{\‾}{x})}^2+{(y-\stackrel{\‾}{y})}^2\]f(x,y)---\left(13\right)$

In whole tracking dirt particles motion path, for intergranular overlap, be blocked phenomenon, intends using interframe Gray Correlation Correction-compensation algorithm, the sequence image i.e. moved by granule, is shifted the pixel of previous frame image and interpolation compensates and works as The edge crumbling of front frame.And granule being there is at the volley to flocculation metaboly, present case uses the method that cluster is estimated, with The particle followed the trail of is cluster centre, and crystallize granule is cluster data, chooses suitable distance metric, uses k means clustering algorithm, Obtain granule subgroup, estimate using this granule subgroup as to tracked granule floc sedimentation shape, obtain the form parameter of granule subgroup, And the floc sedimentation image obtained with reality carries out geometric similarity degree feature and compares, thus solve the grain matches problem caused of flocculating.

When specifically applying, the idiographic flow of step S400 is as follows:

By above-mentioned step, photographic head have recorded the tube wall overall process from smooth to fouling, A_nAnd B_nRepresent n-th frame figure respectively As the deposited particles of wall section with degrade granule, in particles track above, the particle of both trend two are made respectively Different tables is known.After three-dimensionalreconstruction, to deposited particles with degrade granule and make different labelling (as by red-label deposition Grain, blue markings degrades granule), it is distinguish between.If V_AnAnd V_BnAfter representing n-th frame image three-dimensional reconstruction respectively, wall section sinks Long-pending granule and the volume degrading granule, utilize the characteristic parameter of the dirt particles obtained in step S3, utilize its angle point relative to position Reconstruct its geometric model, and then be calculated its volume.

T is made again to represent the difference time of camera acquisition two two field picture, ν_anAnd ν_bnRepresent frame and the deposition of interframe and denudation rate. Then there is a following calculating formula:

${\mathrm{\ν}}_{an}=\frac{V_{An}-V_{A(n-1)}}{T}---\left(14\right)$

${\mathrm{\ν}}_{bn}=\frac{V_{Bn}-V_{B(n-1)}}{T}---\left(15\right)$

And deposition ν after tube wall smooth beginning timing, elapsed time t_aWith denudation rate ν_bHave:

${\mathrm{\ν}}_a=\frac{V_{A\frac{t}{T}}-V_{A1}}{t}---\left(16\right)$

${\mathrm{\ν}}_b=\frac{V_{B\frac{t}{T}}-V_{B1}}{t}---\left(17\right)$

Below being only the preferred embodiment of the present invention, protection scope of the present invention is not limited merely to above-described embodiment, all belongs to Technical scheme under thinking of the present invention belongs to protection scope of the present invention.It should be pointed out that, the ordinary skill for the art For personnel, some improvements and modifications without departing from the principles of the present invention, should be regarded as protection scope of the present invention.

Claims (6)

1. dirt deposition rate and the On-line Measuring Method of denudation rate in a condenser, it is characterised in that step is:

S1: obtain the image sequence of tube wall incrustation process；

S2: identify deposition and degrade granule；The image sequence collected is analyzed, identifies dirt particles；Carry out interframe The coupling of dirt particles, completes the real-time tracing to dirt particles movement locus；

S3: set up the 3-D geometric model of granule；According to dirt particles 3 D motion trace and direction gradient, try to achieve dirt particles In position and the distribution of tube wall, dirt particles and movement locus thereof are carried out three-dimensionalreconstruction, sets up the 3-D geometric model of granule；

S4: try to achieve deposited particles volume and with degrade granule volume and, and then obtain dirt deposition rate and denudation rate.

Dirt deposition rate and the On-line Measuring Method of denudation rate in condenser the most according to claim 1, it is characterised in that In described step S4, according to the 3-D geometric model of dirt particles, try to achieve the volume of deposited particles and degrade the volume of granule, Dirt deposition rate, denudation rate be then the deposition of dirt in the unit time and the volume of denudation, i.e. deposited particles and with degrade Grain volume and.

Dirt deposition rate and the On-line Measuring Method of denudation rate in condenser the most according to claim 1, it is characterised in that Described step S1 is by the image in image collecting device Real-time Collection tube wall, obtains the image sequence of tube wall incrustation process.

4. according to dirt deposition rate and the On-line Measuring Method of denudation rate in the condenser described in claim 1 or 2 or 3, its It is characterised by, in described step S2, identifies that the method for granule is: use two-dimensional Gaussian function to describe the gray scale of dirt particles before this Distribution, obtains particle geometric form parameter, according to the geometric shape parameters of granule by granule intensity profile carries out Function Fitting Identify granule.

5. according to dirt deposition rate and the On-line Measuring Method of denudation rate in the condenser described in claim 1 or 2 or 3, its Being characterised by, in described step S2, the method for grain matches is based on geometric properties similarity and Normalized moment of inertia feature phase The method combined like degree.

6. according to dirt deposition rate and the On-line Measuring Method of denudation rate in the condenser described in claim 1 or 2 or 3, its Being characterised by, in described step S3, the three-dimensional reconstruction method of dirt particles and dirt movement locus is: before this based on binocular vision, Calculate the locus of granule, it is achieved the three-dimensionalreconstruction of granule；Combine afterwards above granule in the motion track information of each interframe, Carry out the three-dimensionalreconstruction of particle motion trajectory.