A solid phase content calculation method and device based on resistive tomography and markov clustering

Abstract

The application provides a solid phase content calculation method and device based on resistive tomography and Markov clustering, and the method comprises the following steps: acquiring time series measurement data of a resistive tomography sensor in a solid-liquid two-phase flow pipeline detection scene, combining pre-acquired electrode spatial position information to construct a measurement vector containing space-time dimensions; using a calibrated Markov clustering algorithm to perform unsupervised clustering segmentation on the space-time measurement vector, dividing pixel points in a reconstructed image into solid phase classes and liquid phase classes; and calculating the solid phase content of the solid-liquid two-phase flow by combining the number proportion of the solid phase class pixel points in the overall imaging area and a pixel weight coefficient.

Description

Technical Field

[0001] This application belongs to the field of multiphase flow process parameter detection and automation technology, and in particular, relates to a method and apparatus for calculating solid phase content based on resistance tomography and Markov clustering. Background Technology

[0002] With the widespread application of solid-liquid two-phase flow in petrochemical, mining, water conservancy, and municipal environmental protection fields, industrial production has placed more stringent demands on the measurement accuracy, real-time performance, and adaptability to operating conditions of solid content in process monitoring and control. Electrical Resistance Tomography (ERT) technology, due to its advantages of being non-invasive, visual, having a fast response, and relatively low cost, has become an important means of detecting solid-liquid two-phase flow parameters in pipelines. However, existing ERT-based solid content calculation methods suffer from insufficient utilization of spatiotemporal information in practical applications, leading to the loss of dynamic characteristics. Traditional methods typically use only single-frame measurement data from ERT sensors to calculate solid content, failing to fully utilize the temporal correlation information between frames obtained through continuous sampling along the time axis. This results in the loss of dynamic flow characteristics of the flow field, causing the measurement results to fail to reflect the true evolution of the solid distribution. Under conditions of frequent flow pattern changes or significant flow pulsation, the measurement error increases significantly.

[0003] Furthermore, existing methods for quantitatively calculating solid content often employ the Maxwell-Garnett equation or its improved forms. The theoretical basis of these equations is the assumption that the solid medium is uniformly distributed in the measurement field. However, in actual industrial pipelines, solid-liquid two-phase flow is affected by various factors such as flow velocity, pipe diameter, solid particle size and density, and pipe inclination angle. The medium distribution is often non-uniform, and solid particles may form agglomerates, strips, dunes, or even embolic flows. This significant deviation between the actual distribution and theoretical assumptions leads to substantial systematic errors in the calculation results. Meanwhile, some researchers have attempted to introduce clustering algorithms for image segmentation. However, the commonly used c-means clustering algorithm, based on Euclidean distance, essentially assumes that clusters are spherical or convex, failing to effectively identify irregularly shaped solid agglomerates in actual solid-liquid two-phase flow. It also exhibits poor adaptability to changes in the morphology, density, and fluctuations in operating parameters of the measured object, easily leading to segmentation errors and further amplifying the measurement error of solid content.

[0004] The combination of these problems results in traditional ERT-based solid content calculation methods suffering from common defects such as large measurement errors, narrow applicability, insufficient determinism, and low intelligence, making it difficult to meet the accuracy and real-time requirements of online monitoring of industrial pipelines under complex operating conditions. Therefore, there is an urgent need for a high-precision solid content calculation method that can fully integrate the spatiotemporal information of ERT measurement sequences and adaptively identify non-uniform solid phase distribution patterns, in order to achieve accurate online measurement of solid content in solid-liquid two-phase flows. Summary of the Invention

[0005] This application provides a method and apparatus for calculating solid content based on resistance tomography and Markov clustering, which at least solves the problems of insufficient utilization of spatiotemporal information leading to loss of dynamic features, serious discrepancies between the assumption of solid distribution uniformity and actual working conditions introducing systematic errors, and poor adaptability of clustering algorithms leading to inaccurate image segmentation.

[0006] According to the first aspect of this application, a method for calculating solid content based on resistivity chromatography and Markov clustering is provided, comprising: Time-series measurement data of resistivity tomography sensor in solid-liquid two-phase flow pipeline detection scenario are acquired, and measurement vector containing spatiotemporal dimensions is constructed by combining pre-acquired electrode spatial location information. The spatiotemporal measurement vectors were segmented into solid phase and liquid phase using the calibrated Markov clustering algorithm. The proportion of solid-phase pixels in the overall imaging area is statistically analyzed, and the solid content of the solid-liquid two-phase flow is calculated by combining the pixel weight coefficient.

[0007] In one embodiment, constructing a measurement vector that includes spatiotemporal dimensions includes: Acquire multiple consecutive frames of resistivity tomography reconstructed images; For each pixel location, extract the grayscale value, first-order difference, and second-order difference of the current frame; The grayscale value, first-order difference, and second-order difference of the current frame are grouped into a three-dimensional spatiotemporal feature vector, and the spatiotemporal vectors of all pixels constitute the dataset to be clustered.

[0008] In one embodiment, the excitation mode of the resistance tomography sensor includes any one of single-electrode excitation, adjacent-electrode excitation, or diagonal excitation, and the corresponding measurement mode is adjacent-electrode measurement; the image reconstruction algorithm includes non-iterative algorithms or iterative algorithms.

[0009] In one embodiment, unsupervised clustering segmentation using a calibrated Markov clustering algorithm includes: The spatiotemporal feature vector set is constructed as a graph structure, with nodes being pixels and edge weights reflecting the similarity between feature vectors. By alternating between expansion and dilation operations, the walk probabilities are gradually concentrated in densely connected regions, eventually dividing the graph into several connected components, each corresponding to a cluster.

[0010] In one embodiment, the calibration method for the calibrated Markov clustering algorithm is as follows: Data was collected under typical operating conditions with known solid content. With the goal of maximizing segmentation accuracy, the combination of expansion and extension parameters was optimized using a grid search method.

[0011] In one embodiment, the Markov clustering algorithm can be replaced by CLIQUE clustering, density-based noise-applied spatial clustering, or spectral clustering.

[0012] According to another aspect of this application, a solid phase content calculation device based on resistivity chromatography and Markov clustering is also provided, comprising: The data acquisition unit is used to acquire time-series measurement data of the resistivity tomography sensor in the solid-liquid two-phase flow pipeline detection scenario, and construct a measurement vector containing spatiotemporal dimensions by combining it with the pre-acquired electrode spatial position information. The clustering segmentation unit is used to perform unsupervised clustering segmentation of spatiotemporal measurement vectors using the calibrated Markov clustering algorithm, dividing the pixels in the reconstructed image into solid phase and liquid phase classes; The solid content calculation unit is used to count the proportion of solid-phase pixels in the overall imaging area and calculate the solid content of the solid-liquid two-phase flow by combining the pixel weight coefficient.

[0013] In one embodiment, the data acquisition unit includes: The image acquisition module is used to acquire multiple consecutive frames of resistivity tomography reconstructed images; The feature extraction module is used to extract the grayscale value, first-order difference, and second-order difference of the current frame for each pixel position. The vector construction module is used to synthesize the grayscale value, first-order difference, and second-order difference group of the current frame into a three-dimensional spatiotemporal feature vector. The spatiotemporal vectors of all pixels constitute the dataset to be clustered.

[0014] In one embodiment, the excitation mode of the resistance tomography sensor includes any one of single-electrode excitation, adjacent-electrode excitation, or diagonal excitation, and the corresponding measurement mode is adjacent-electrode measurement; the image reconstruction algorithm includes non-iterative algorithms or iterative algorithms.

[0015] In one embodiment, the clustering segmentation unit includes: The graph construction module is used to construct a graph structure from the spatiotemporal feature vector set, where nodes are pixels and edge weights reflect the similarity between feature vectors. The clustering execution module is used to gradually concentrate the walk probabilities in densely connected regions by alternating expansion and dilation operations, ultimately dividing the graph into several connected components, each corresponding to a cluster.

[0016] In one embodiment, the device further includes a parameter calibration unit for collecting data under typical operating conditions with known solid content, and optimizing the combination of expansion parameters and extension parameters by a grid search method with the goal of maximizing segmentation accuracy.

[0017] In one embodiment, the Markov clustering algorithm can be replaced by CLIQUE clustering, density-based noise-applied spatial clustering, or spectral clustering.

[0018] This application achieves adaptive identification and accurate segmentation of non-uniformly distributed, arbitrarily shaped solid agglomerates by integrating the spatiotemporal correlation information of resistance tomography measurement sequences with a calibrated Markov clustering algorithm, thereby significantly improving the measurement accuracy and adaptability of solid phase content in solid-liquid two-phase flow under complex working conditions. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart of a solid phase content calculation method based on resistance chromatography and Markov clustering provided in this application.

[0021] Figure 2 This is a flowchart illustrating a method for constructing a measurement vector containing spatiotemporal dimensions in an embodiment of this application.

[0022] Figure 3 This is a flowchart illustrating the method for unsupervised clustering segmentation using a calibrated Markov clustering algorithm in this embodiment of the application.

[0023] Figure 4 A structural block diagram of a solid phase content calculation device based on resistivity chromatography and Markov clustering provided in this application.

[0024] Figure 5 This is a specific implementation of an electronic device in the embodiments of this application. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] To address the problems existing in the background technology, this application provides a method for calculating solid content based on resistance chromatography and Markov clustering, such as... Figure 1 As shown, it includes: S101: Acquire time-series measurement data of the resistivity tomography sensor in the solid-liquid two-phase flow pipeline detection scenario, and construct a measurement vector containing spatiotemporal dimensions by combining the pre-acquired electrode spatial position information; Specifically, based on the function of visual reconstruction of the detection target using resistance tomography, a reconstruction algorithm is used to visually reconstruct the detection object in the solid-liquid two-phase flow measurement field of the pipeline. Leveraging the real-time nature of resistance tomography measurements, multiple consecutive reconstructed images are quickly acquired. Based on the grayscale value of each pixel in each frame and its evolution characteristics over time, a three-dimensional spatiotemporal feature vector containing spatial location information and temporal dynamic information is constructed, forming a spatiotemporal feature vector set. The principle of resistance tomography visualization reconstruction refers to using an electrode array arranged at the boundary of the measured field (such as the wall of a pipeline or reactor), applying current excitation and measuring the boundary voltage to obtain projection data, and then, based on a specific mathematical and physical model and image reconstruction algorithm, inverting the acquired boundary measurement data into a spatial distribution image of the conductivity within the measured field. The real-time principle of resistivity tomography (RTG) means that, at the engineering application level, the sampling rate can meet the needs of visual monitoring and control of unsteady, rapidly changing multiphase flow processes. It is sufficient to replace visual observation for process monitoring and can serve as a feedback signal for control loops for process control. It is only limited in extreme physical processes such as transient discharges at extremely high speeds or hypersonic flows. It can fully meet the real-time requirements for more than 95% of chemical, metallurgical, and environmental engineering applications.

[0027] S102: The spatiotemporal measurement vector is segmented into solid phase and liquid phase using the calibrated Markov clustering algorithm to perform unsupervised clustering. The pixels in the reconstructed image are divided into solid phase and liquid phase classes. Unsupervised clustering segmentation refers to the process of automatically dividing a given dataset into several disjoint subsets based on the inherent similarity metrics of data objects, without prior labeling or category guidance. Its aim is to maximize the similarity between data objects within the same class while minimizing the similarity between data objects in different classes. Markov clustering algorithms discover natural clustering structures in data by simulating random walks in graph theory. The algorithm first constructs a graph structure from the spatiotemporal feature vector set, with nodes representing pixels and edge weights reflecting the similarity between feature vectors. Then, by alternating expansion operations (simulating random walks to allow flow to propagate in the graph) and dilation operations (strengthening strong connections and weakening weak connections), the walk probabilities gradually converge in densely connected regions, ultimately segmenting the graph into several connected components, each corresponding to a cluster. In a solid-liquid two-phase flow scenario, pixels within the solid phase particle region exhibit high similarity in the spatiotemporal feature space and naturally cluster into one class; the liquid phase background region clusters into another class. This method does not require a preset number of clusters and can adaptively identify solid aggregates of arbitrary shape and non-uniform distribution, breaking through the traditional method's assumption of uniform solid distribution.

[0028] S103: Calculate the solid content of the solid-liquid two-phase flow by statistically analyzing the proportion of solid-phase pixels in the overall imaging area and combining the pixel weight coefficient.

[0029] The formula for calculating the solid content is:

[0030] in: η Let |C1| be the solid content of the cross section; |C1| be the number of pixels in the solid phase pixel class after clustering in step S2; and |C1|+|C1| be the total number of pixels after discretization within the sensitive field of the ERT sensor (i.e., the sum of the number of pixels in the solid and liquid phase pixel classes within the detection field). This method is based on the pixel counting principle and directly uses the image segmentation results for quantitative calculation, avoiding the errors caused by traditional empirical models such as the Maxwell-Garnett equation that require assumptions about the uniformity of phase distribution.

[0031] In one embodiment, such as Figure 2 The construction of a measurement vector containing spatiotemporal dimensions includes: S201: Acquire multiple consecutive frames of resistivity tomography reconstructed images; Specifically, when a solid-liquid two-phase flow occurs in a pipe, the motion of solid particles exhibits spatiotemporal continuity; the pixel state at a given location at a current moment is correlated with the states at previous times and in its neighborhood. To fully exploit this characteristic, three consecutive frames of resistivity tomography reconstructed images are acquired, and based on the... t Three frames along the flow direction at each moment G ( t -2),G ( t -1), G ( t (Then proceed with further processing.)

[0032] S202: For each pixel location, extract the grayscale value, first-order difference, and second-order difference of the current frame; For the first in the detection field Ω k Each pixel, based on grayscale value g k ( t Calculate the following first-order difference Δ 1 g k ( t ) and second-order difference Δ 2 g k ( t ):

[0033] here: A ( t )={ A 1( t ), A 2( t ),…, A n ( t )}, A k ( t )=( g k ( t ), △ 1 g k ( t ), △ 2 g k ( t )) T , k =1, 2,…, n .

[0034] in, t For frame index, k=1,2,...,n,n This represents the number of pixels in a single frame of an image.

[0035] S203: Combine the grayscale value, first-order difference, and second-order difference of the current frame into a three-dimensional spatiotemporal feature vector. The spatiotemporal vectors of all pixels constitute the dataset to be clustered.

[0036] The form of the three-dimensional spatiotemporal feature vector is A k (t )=( g k ( t ), △ 1 g k ( t ), △ 2 g k ( t )) T， in gk(t) For the detection field Ω, the first k The grayscale values ​​of each pixel. The spatiotemporal vectors of all pixels constitute the dataset to be clustered. A(t) = {A1(t), A2(t),...,An(t)} This method integrates the spatial distribution and temporal evolution of image sequences into a single feature space, laying the foundation for subsequent accurate segmentation. From the 1st to the... t Continuous sampling at each time point T Each sample constitutes a spatiotemporal vector set. {A (t)} The total number of image frames is T .

[0037] In one embodiment, the excitation mode of the resistance tomography sensor includes any one of single-electrode excitation, adjacent-electrode excitation, or diagonal excitation, and the corresponding measurement mode is adjacent-electrode measurement. Taking a 16-electrode resistance tomography sensor using a single-electrode excitation-single-electrode measurement method as an example, 240 measurement values ​​can be obtained, which constitute a detection vector. The image reconstruction algorithm can be a non-iterative algorithm (such as linear back-projection algorithm, Tikhonov regularization algorithm) or an iterative algorithm (such as Landweber iterative algorithm, conjugate gradient algorithm, Newton-Raphson algorithm, etc.). When high time resolution is required, non-iterative algorithms are used; when high time resolution is required, iterative algorithms are used. Taking the most common back-projection algorithm as an example, the target reconstruction algorithm uses the following form: G=S T U in, S yes m×n A 3D sensitivity matrix reflects the changes in sensitivity from... m A boundary measurement to Ω middle n Constitutive relations between pixels G In the measurement field n The set is composed of the gray levels of individual pixels. U This is the measurement vector.

[0038] In one embodiment, the excitation mode of the resistance tomography sensor includes any one of single-electrode excitation, adjacent-electrode excitation, or diagonal excitation, and the corresponding measurement mode is adjacent-electrode measurement; the image reconstruction algorithm includes non-iterative algorithms or iterative algorithms.

[0039] In one embodiment, such as Figure 3 As shown, unsupervised clustering segmentation using the calibrated Markov clustering algorithm includes: S301: Construct the spatiotemporal feature vector set into a graph structure, with nodes being pixels and edge weights reflecting the similarity between feature vectors; Specifically, the set of spatiotemporal vectors {A(t)} Each vector is used as a node to create a non-fully connected network graph. G Calculate the distance from each node in the graph to it. s Construct a distance matrix using the distances of the nearest neighbors as non-zero values. W And normalization is performed along the line.

[0040] S302: By alternating between expansion and dilation operations, the walk probability is gradually concentrated in densely connected regions, eventually dividing the graph into several connected components, each corresponding to a cluster.

[0041] Specifically, repeat the following two steps until convergence: expansion operation, matrix computation. W of k Power of 1 W k Simulated random walks allow flow to propagate through the graph; the dilation operation, based on parameters... r right W Sharpen the image to strengthen strong connections and weaken weak connections. Check. W Does it meet the convergence condition? W k+1 - W k ||≤ ε ( ε If the condition is a sufficiently small integer, then output the clustering results of all vectors (the class labels of the points); otherwise, continue repeating the expansion and dilation operations. The final implementation is for... {A(t)} All vectors are clustered and segmented to obtain two classes: one corresponding to the solid-phase pixel and the other to the solid-phase pixel. C1 and the corresponding liquid phase pixel class C2 .

[0042] In one embodiment, the calibration method for the calibrated Markov clustering algorithm is as follows: data is collected under typical operating conditions with known solid content, and the parameter combination is optimized using a grid search method with the goal of maximizing segmentation accuracy. Key parameters of the Markov clustering algorithm include the dilation parameter. r(Typically, a value between 2.0 and 3.0 is used, as it affects the granularity of clustering) and expansion parameters e (Usually 2 is chosen, which affects the step size of the random walk).

[0043] In one embodiment, the Markov clustering algorithm can be replaced by CLIQUE (CLustering In QUEst) clustering, density-based spatial clustering of applications with noise (DBSCAN), or spectral clustering, etc. These alternative algorithms can also achieve unsupervised clustering segmentation of spatiotemporal feature vector sets, automatically classifying pixels in an image into solid-phase pixel classes and liquid-phase pixel classes.

[0044] The specific implementation process of the present invention will be described below with reference to the accompanying drawings: 1. Determine the excitation and measurement methods within a given solid-liquid two-phase flow detection field Ω. S 1, and obtained the first t Moment m 240 measurement values ​​can be obtained by using a 16-electrode ERT sensor with a single-electrode excitation-single-electric measurement method, which constitute a detection vector.

[0045] 2. Determine the division of the testing area. S 2, Ω is divided into n A set of pixels { g 1, g 2, …, g n}

[0046] 3. Determine the clustering algorithm S 3. Reconstruct the target in Ω. When high time resolution is required, use non-iterative algorithms; when high time resolution is required, use iterative algorithms.

[0047] Taking the most common back-projection algorithm as an example, the target reconstruction algorithm uses the following form: G=S T U here S yes m × n A 3D sensitivity matrix reflects the changes in sensitivity from... m The boundary is measured to Ω. n constitutive relations between pixels G In the measurement field n The set is composed of the gray levels of individual pixels.

[0048] 4. In the first tAt each sampling time, the ERT sensor is used to measure and input the detection vector. U ( t ) S 4.

[0049] 5. In the first t Each sampling time, using U ( t The reconstruction algorithm reconstructs the detected target in Ω. G ( t ) S 5, here G ( t )={ g 1( t ), g 2( t ),…, g n ( t )} here g 1( t ), g 2( t ),…, g n ( t ) is the first t In the reconstructed image at each sampling time n The grayscale value of each pixel.

[0050] 6. In the t Reconstruction at each sampling time G ( t ), G ( t -1) and G ( t -1) S 6.

[0051] 7. In the t Construct a spatiotemporal vector at each sampling time. A ( t ) S 7.

[0052] According to the t Three frames along the flow direction at each moment G ( t -2), G ( t -1), G ( t ); For the first in the detection field Ω k Each pixel, based on grayscale value g k ( t Calculate the following first-order difference Δ1 g k ( t ) and second-order difference Δ 2 g k ( t ):

[0053] here: A ( t )={ A 1( t ), A 2( t ),…, A n ( t )}, A k ( t )=( g k ( t ), △ 1 g k ( t ), △ 2 g k ( t )) T , k =1, 2,…, n . 8. From number 1 to number 2 t Continuous sampling at each time point T 10 samples constitute the spatiotemporal vector set { A ( t )} S 8.

[0054] 9. Select a set of spatiotemporal vectors labeled with solid content to calibrate the MCL algorithm. S expansion factor of 9 e and expansion coefficient r .

[0055] 10. Clustering using the MCL algorithm S 10. Implement for { A ( t All vectors are clustered and divided into two classes: one for each solid pixel. C 1 and the corresponding liquid phase pixel class C 2; In addition to calibrating extended parameters e and expansion parameters r Apart from that, the other steps are the same as the standard MCL algorithm.

[0056] The calibrated MCL algorithm clustering steps are as follows:

[0057] 11. The number of solid-phase pixels used is in Ω ( t The percentage of total pixels is used to calculate the solid content. S 11:

[0058] Here, |·| represents the number of elements in the set.

[0059] Based on the same inventive concept, this application also provides a solid phase content calculation device based on resistance tomography and Markov clustering, which can be used to implement the method described in the above embodiments, as described in the following embodiments. Since the principle of solving the problem using this solid phase content calculation device based on resistance tomography and Markov clustering is similar to that of the solid phase content calculation method based on resistance tomography and Markov clustering, the implementation of the solid phase content calculation device based on resistance tomography and Markov clustering can refer to the implementation of the solid phase content calculation method based on resistance tomography and Markov clustering, and will not be repeated. As used below, the term "unit" or "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the system described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0060] According to another aspect of this application, a solid content calculation device based on resistivity chromatography and Markov clustering is also provided, such as... Figure 4 As shown, it includes: The data acquisition unit 401 is used to acquire time-series measurement data of the resistivity tomography sensor in the solid-liquid two-phase flow pipeline detection scenario, and construct a measurement vector containing spatiotemporal dimensions by combining the pre-acquired electrode spatial position information. Clustering segmentation unit 402 is used to perform unsupervised clustering segmentation of spatiotemporal measurement vectors using the calibrated Markov clustering algorithm, and to divide the pixels in the reconstructed image into solid phase class and liquid phase class; The content calculation unit 403 is used to count the proportion of solid-phase pixels in the overall imaging area and calculate the solid content of the solid-liquid two-phase flow by combining the pixel weight coefficient.

[0061] In one embodiment, the data acquisition unit includes: The image acquisition module is used to acquire multiple consecutive frames of resistivity tomography reconstructed images; The feature extraction module is used to extract the grayscale value, first-order difference, and second-order difference of the current frame for each pixel position. The vector construction module is used to synthesize the grayscale value, first-order difference, and second-order difference group of the current frame into a three-dimensional spatiotemporal feature vector. The spatiotemporal vectors of all pixels constitute the dataset to be clustered.

[0062] In one embodiment, the excitation mode of the resistance tomography sensor includes any one of single-electrode excitation, adjacent-electrode excitation, or diagonal excitation, and the corresponding measurement mode is adjacent-electrode measurement; the image reconstruction algorithm includes non-iterative algorithms or iterative algorithms.

[0063] In one embodiment, the clustering segmentation unit includes: The graph construction module is used to construct a graph structure from the spatiotemporal feature vector set, where nodes are pixels and edge weights reflect the similarity between feature vectors. The clustering execution module is used to gradually concentrate the walk probabilities in densely connected regions by alternating expansion and dilation operations, ultimately dividing the graph into several connected components, each corresponding to a cluster.

[0064] In one embodiment, the device further includes a parameter calibration unit for collecting data under typical operating conditions with known solid content, and optimizing the combination of expansion parameters and extension parameters by a grid search method with the goal of maximizing segmentation accuracy.

[0065] In one embodiment, the Markov clustering algorithm can be replaced by CLIQUE clustering, density-based noise-applied spatial clustering, or spectral clustering.

[0066] This application achieves adaptive identification and accurate segmentation of non-uniformly distributed, arbitrarily shaped solid agglomerates by integrating the spatiotemporal correlation information of resistance tomography measurement sequences with a calibrated Markov clustering algorithm, thereby significantly improving the measurement accuracy and adaptability of solid phase content in solid-liquid two-phase flow under complex working conditions.

[0067] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0068] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0069] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0070] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0071] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

[0072] This application also provides a specific implementation of an electronic device capable of implementing all the steps in the methods described above. See [link to implementation details]. Figure 5 The electronic device specifically includes the following: Figure 5 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of this application. The electronic device includes: a processor 501, a memory 502, and a bus 503.

[0073] The processor 501 and the memory 502 communicate with each other via the bus 503.

[0074] The processor 501 is used to call the computer program in the memory 502. When the processor executes the computer program, it implements all the steps in the method in the above embodiments. For example, when the processor executes the computer program, it implements the following steps: S101: Acquire time-series measurement data of the resistivity tomography sensor in the solid-liquid two-phase flow pipeline detection scenario, and construct a measurement vector containing spatiotemporal dimensions by combining the pre-acquired electrode spatial position information; S102: The spatiotemporal measurement vector is segmented into solid phase and liquid phase using the calibrated Markov clustering algorithm to perform unsupervised clustering. The pixels in the reconstructed image are divided into solid phase and liquid phase classes. S103: Calculate the solid content of the solid-liquid two-phase flow by statistically analyzing the proportion of solid-phase pixels in the overall imaging area and combining the pixel weight coefficient.

[0075] Embodiments of this application also provide a computer-readable storage medium capable of implementing all steps of the methods in the above embodiments. The computer-readable storage medium stores a computer program that, when executed by a processor, implements all steps of the methods in the above embodiments. For example, when the processor executes the computer program, it implements the following steps: S101: Acquire time-series measurement data of the resistivity tomography sensor in the solid-liquid two-phase flow pipeline detection scenario, and construct a measurement vector containing spatiotemporal dimensions by combining the pre-acquired electrode spatial position information; S102: The spatiotemporal measurement vector is segmented into solid phase and liquid phase using the calibrated Markov clustering algorithm to perform unsupervised clustering. The pixels in the reconstructed image are divided into solid phase and liquid phase classes. S103: Calculate the solid content of the solid-liquid two-phase flow by statistically analyzing the proportion of solid-phase pixels in the overall imaging area and combining the pixel weight coefficient.

[0076] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on its differences from other embodiments. In particular, for hardware + program embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. Although the embodiments in this specification provide the method operation steps as shown in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive means. The order of steps listed in the embodiments is merely one possible execution order among many steps and does not represent the only execution order. In actual device or terminal product execution, the methods can be executed in the order shown in the embodiments or drawings or in parallel (e.g., in a parallel processor or multi-threaded processing environment, or even a distributed data processing environment). The terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, product, or apparatus that includes a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, product, or apparatus. Without further limitations, the presence of other identical or equivalent elements in the process, method, product, or apparatus that includes said elements is not excluded. For ease of description, the above devices are described in terms of function, divided into various modules. Of course, in implementing the embodiments of this specification, the functions of each module can be implemented in one or more software and / or hardware, or the module implementing the same function can be implemented by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings or direct couplings or communication connections may be through some interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms. This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions, which are executable by the processor of the computer or other programmable data processing device, produce instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0077] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, the embodiments of this specification can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The various embodiments in this specification are described in a progressive manner, and similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for system embodiments, since they are basically similar to method embodiments, the description is relatively simple, and relevant parts can be referred to the description of the method embodiments. In the description of this specification, the reference to the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., means that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the embodiments of this specification.

[0078] In this specification, the illustrative expressions of the terms used do not necessarily refer to the same embodiments or examples. Furthermore, those skilled in the art can combine and integrate different embodiments or examples described in this specification, as well as features of different embodiments or examples, without contradiction. The above descriptions are merely embodiments of this specification and are not intended to limit the embodiments of this specification. Various modifications and variations can be made to the embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the embodiments of this specification should be included within the scope of the claims of the embodiments of this specification.

Claims

1. A method for calculating solid content based on resistivity tomography and Markov clustering, characterized in that, include: Time-series measurement data of resistivity tomography sensor in solid-liquid two-phase flow pipeline detection scenario are acquired, and measurement vector containing spatiotemporal dimensions is constructed by combining pre-acquired electrode spatial location information. The spatiotemporal measurement vector was segmented into solid phase and liquid phase using the calibrated Markov clustering algorithm. The proportion of solid-phase pixels in the overall imaging area is statistically analyzed, and the solid content of the solid-liquid two-phase flow is calculated by combining the pixel weight coefficient.

2. The method according to claim 1, characterized in that, The constructed measurement vector containing spatiotemporal dimensions includes: Acquire multiple consecutive frames of resistivity tomography reconstructed images; For each pixel location, extract the grayscale value, first-order difference, and second-order difference of the current frame; The grayscale value, first-order difference, and second-order difference of the current frame are grouped into a three-dimensional spatiotemporal feature vector, and the spatiotemporal vectors of all pixels constitute the dataset to be clustered.

3. The method according to claim 1, characterized in that, The excitation mode of the resistance tomography sensor includes any one of single-electrode excitation, adjacent-electrode excitation, or diagonal excitation, and the corresponding measurement mode is adjacent-electrode measurement; the image reconstruction algorithm includes non-iterative algorithms or iterative algorithms.

4. The method according to claim 1, characterized in that, The unsupervised clustering segmentation using the calibrated Markov clustering algorithm includes: The spatiotemporal feature vector set is constructed as a graph structure, with nodes being pixels and edge weights reflecting the similarity between feature vectors. By alternating between expansion and dilation operations, the walk probabilities are gradually concentrated in densely connected regions, eventually dividing the graph into several connected components, each corresponding to a cluster.

5. The method according to claim 1, characterized in that, The calibration method for the calibrated Markov clustering algorithm is as follows: Data was collected under typical operating conditions with known solid content. With the goal of maximizing segmentation accuracy, the combination of expansion and extension parameters was optimized using a grid search method.

6. The method according to claim 1, characterized in that, The Markov clustering algorithm can be replaced by CLIQUE clustering, density-based noise-applied spatial clustering, or spectral clustering.

7. A device for calculating solid content based on resistivity chromatography and Markov clustering, characterized in that, include: The data acquisition unit is used to acquire time-series measurement data of the resistivity tomography sensor in the solid-liquid two-phase flow pipeline detection scenario, and construct a measurement vector containing spatiotemporal dimensions by combining it with the pre-acquired electrode spatial position information. The clustering segmentation unit is used to perform unsupervised clustering segmentation on the spatiotemporal measurement vector using the calibrated Markov clustering algorithm, dividing the pixels in the reconstructed image into solid phase and liquid phase classes. The solid content calculation unit is used to count the proportion of solid-phase pixels in the overall imaging area and calculate the solid content of the solid-liquid two-phase flow by combining the pixel weight coefficient.

8. The apparatus according to claim 7, characterized in that, The data acquisition unit includes: The image acquisition module is used to acquire multiple consecutive frames of resistivity tomography reconstructed images; The feature extraction module is used to extract the grayscale value, first-order difference, and second-order difference of the current frame for each pixel position. The vector construction module is used to synthesize the grayscale value, first-order difference, and second-order difference group of the current frame into a three-dimensional spatiotemporal feature vector. The spatiotemporal vectors of all pixels constitute the dataset to be clustered.

9. The apparatus according to claim 7, characterized in that, The excitation mode of the resistance tomography sensor includes any one of single-electrode excitation, adjacent-electrode excitation, or diagonal excitation, and the corresponding measurement mode is adjacent-electrode measurement; the image reconstruction algorithm includes non-iterative algorithms or iterative algorithms.

10. The apparatus according to claim 7, characterized in that, The clustering segmentation unit includes: The graph construction module is used to construct a graph structure from the spatiotemporal feature vector set, where nodes are pixels and edge weights reflect the similarity between feature vectors. The clustering execution module is used to gradually concentrate the walk probabilities in densely connected regions by alternating expansion and dilation operations, ultimately dividing the graph into several connected components, each corresponding to a cluster.

11. The apparatus according to claim 7, characterized in that, It also includes a parameter calibration unit, which is used to collect data under typical working conditions with known solid content, and optimize the combination of expansion parameters and extension parameters by means of grid search method with the goal of maximizing the segmentation accuracy.

12. The apparatus according to claim 7, characterized in that, The Markov clustering algorithm can be replaced by CLIQUE clustering, density-based noise-applied spatial clustering, or spectral clustering.

13. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the solid phase content calculation method based on resistivity tomography and Markov clustering as described in any one of claims 1 to 6.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the solid phase content calculation method based on resistivity tomography and Markov clustering as described in any one of claims 1 to 6.

Images