Heavy oil storage tank water cutting control method, water cutting system, electronic device and storage medium
By employing image segmentation technology based on Markov random fields and convolutional neural networks in heavy oil storage tanks, combined with bidirectional image acquisition, the problem of unstable water cut-off control in heavy oil storage tanks was solved, and precise water cut-off control under high-intensity operation was achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA PETROLEUM & CHEMICAL CORP
- Filing Date
- 2022-01-11
- Publication Date
- 2026-05-05
AI Technical Summary
Traditional mechanical and electronic water cutters in heavy oil storage tanks suffer from malfunctions or sensor failures, leading to unstable and unreliable water cut control.
Image segmentation and recognition are performed using the ICM algorithm based on Markov random fields and convolutional neural networks. Combined with bidirectional image acquisition, the heavy oil ratio is calculated by weighted summation to control the opening and closing of the water-cutting valve.
It achieves precise water cutting control under conditions of similar oil and water densities, improves the stability and reliability of water cutting in heavy oil storage tanks, and reduces maintenance frequency and errors.
Smart Images
Figure CN116468925B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of storage tank technology in the petrochemical industry, and particularly to a water-cutting control method, water-cutting system, electronic equipment, and storage medium for heavy oil storage tanks. Background Technology
[0002] The water cutter is a product designed and developed to meet the needs of dehydration of storage tanks in the petrochemical industry. It is installed at the bottom of the oil tank and is used in the petroleum and chemical industries to replace the manual water cutting method of oil tanks, realize automatic water cutting, greatly reduce the labor intensity of operation and the hydrogen sulfide poisoning phenomenon that occurs during manual water cutting, and reduce oil waste and environmental pollution. At the same time, it meets the standard requirements of the petrochemical system for underground drainage and pipeline discharge.
[0003] Traditional water cutters are divided into two types: mechanical and electronic. Mechanical water cutters automatically open and close the drain using the mechanical force generated by buoyancy, requiring no manual intervention, consuming no energy, and posing no explosion-proof issues. However, most mechanical water cutters used in heavy oil storage tanks commonly experience malfunctions, mainly because the density of heavy oils such as crude oil is similar to that of water, resulting in significant oil-in-water and water-in-oil emulsions. The working principle of traditional mechanical water cutters is based on the buoyancy generated by a float, amplified by levers, and combined with springs and other mechanical structures to control the water cutter valve. When the buoyancy generated by the float in heavy oil and water is similar, it cannot accurately control the water cutter valve. Electronic water cutters identify the medium in the pipeline using sensors and control the valve based on the sensor detection results. For example, using the principle of liquid column resonance, two sensors are installed at different locations in the pipeline. When a mixed medium passes through the pipeline, due to the difference in viscosity and dielectric constant, the sensors send the detection results to the controller, which determines whether to close the water cutter valve based on the oil content of the medium. Electronic water cutters are highly sensitive and can be used in heavy oil storage tanks. However, after a few months of use, the sensor probe will usually become stuck with heavy oil, causing the sensor to fail. The maintenance cycle is too short, making them unsuitable for high-intensity operating storage tanks such as those in refineries.
[0004] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention
[0005] One of the objectives of this invention is to provide a water-cutting control method, water-cutting system, electronic equipment, and storage medium for heavy oil storage tanks, thereby overcoming the problem that traditional mechanical water cutters are not effective for heavy oil storage tanks.
[0006] Another objective of this invention is to provide a water-cutting control method, water-cutting system, electronic equipment, and storage medium for heavy oil storage tanks, thereby improving the stability and reliability of water-cutting in heavy oil storage tanks under high-intensity operation.
[0007] To achieve the above objectives, according to a first aspect of the present invention, the present invention provides a method for controlling water cut-off in heavy oil storage tanks, comprising the following steps:
[0008] S110 acquires top and side view images of the fluid in the water-cutting pipe;
[0009] S151 uses the ICM algorithm based on Markov random fields to segment the top view and side view images;
[0010] Based on the segmentation results, S152 obtains the heavy oil percentage P1 of the top view image and the heavy oil percentage P2 of the side view image.
[0011] S153 calculates the proportion of heavy oil in the fluid in the water-cutting pipe by weighted summation, n = μ1P1 + μ2P2, where μ1 is the weight of the top view image, μ2 is the weight of the side view image, and μ1 + μ2 = 1.
[0012] S154 compares the calculated heavy oil percentage n in the water-cutting pipe with the preset shut-off threshold n0. When n≥n0, the water-cutting pipe is shut off; when n<n0, the water-cutting pipe is kept in its water-cutting state and the process returns to step S110.
[0013] Furthermore, in the above technical solution, the weights μ1 and μ2 of the top view image are determined based on the density of heavy oil.
[0014] Furthermore, in the above technical solution, the weight of the top view image is μ1 = ρ0 - 0.45, where ρ0 is the ratio of the density of heavy oil to the density of water.
[0015] Furthermore, in the above technical solution, the preset shutdown threshold n0 is 5% to 8%.
[0016] Furthermore, in the above technical solution, the method for controlling water removal from heavy oil storage tanks also includes the following step before step S151:
[0017] S120 uses a convolutional neural network to perform image recognition on top-view and side-view images to obtain recognition results;
[0018] Based on the identification result, S130 shuts off the water cutting pipe, returns to step 110, or proceeds to step S151.
[0019] Furthermore, in the above technical solution, step S120 includes:
[0020] S121 acquires historical top-view and side-view images of the fluid in the water-cutting pipe to establish a training sample set;
[0021] S122 assigns labels to each historical top-view image and historical side-view image in the training sample set, denoted as "cutting water" and "closing";
[0022] S123 inputs the training sample set and the corresponding set of labels into the convolutional neural network for training, and obtains the trained convolutional neural network model.
[0023] S124 uses a trained convolutional neural network model to identify the acquired top-view and side-view images and obtain the identification results.
[0024] Furthermore, in the above technical solution, step S130 includes: when the recognition results of the top view image and the side view image at the same time are both "closed", closing the water cutting pipe; when the recognition results of the top view image and the side view image at the same time are both "water cutting", maintaining the water cutting state of the water cutting pipe and returning to step S110; and when the recognition results of the top view image and the side view image at the same time are different, proceeding to step S151.
[0025] Furthermore, in the above technical solution, step S122 includes: using the ICM algorithm based on Markov random fields to segment the historical top-view image and the historical side-view image; obtaining the heavy oil percentage Q1 of the historical top-view image and the heavy oil percentage Q2 of the historical side-view image based on the segmentation results; when Q1≥5%, the historical top-view image is labeled "closed"; when Q1<5%, the historical top-view image is labeled "water cut"; when Q2≥5%, the historical side-view image is labeled "closed"; when Q2<5%, the historical side-view image is labeled "water cut".
[0026] According to a second aspect of the present invention, a heavy oil storage tank water-cutting system is provided, comprising: an image acquisition unit for acquiring top-view and side-view images of fluid in a water-cutting pipe; an identification unit for performing image recognition on the top-view and side-view images using a convolutional neural network to obtain identification results; a calculation unit for performing image segmentation on the top-view and side-view images using an ICM algorithm based on Markov random fields, and obtaining the heavy oil percentage P1 of the top-view image and the heavy oil percentage P2 of the side-view image based on the segmentation results; calculating the heavy oil percentage n of the fluid in the water-cutting pipe by weighted summation n = μ1P1 + μ2P2, where μ1 is the weight of the top-view image, μ2 is the weight of the side-view image, and μ1 + μ2 = 1; and a control unit for controlling the water-cutting pipe according to the identification results obtained by the identification unit and the heavy oil percentage of the fluid in the water-cutting pipe calculated by the calculation unit.
[0027] Furthermore, in the above technical solution, the identification unit includes: a model acquisition submodule, which is used to establish a training sample set based on historical top-view and historical side-view images of the fluid in the water-cutting pipe; to assign labels to each historical top-view and historical side-view image in the training sample set, denoted as "water-cutting" and "closed"; to input the training sample set and the corresponding assigned labels into a convolutional neural network for training, thereby obtaining a trained convolutional neural network model; and an image recognition submodule, which uses the trained convolutional neural network model to identify the acquired top-view and side-view images, thereby obtaining the recognition result.
[0028] Furthermore, in the above technical solution, the image acquisition unit includes: a transparent pipe section, which is a section of the water cutting pipe; two cameras, which respectively acquire top view and side view images of the fluid in the water cutting pipe through the transparent pipe section; and a heat insulation component, which maintains the working temperature of the transparent pipe section at 45-55℃.
[0029] Furthermore, in the above technical solution, the heavy oil storage tank water cutting system also includes a flushing unit, which is used to flush the water cutting pipe after it is closed.
[0030] According to a third aspect of the present invention, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the at least one processor to perform a heavy oil storage tank water-cutting control method as described in any of the above technical solutions.
[0031] According to a fourth aspect of the present invention, the present invention provides a non-transitory computer-readable storage medium storing computer-executable instructions for causing a computer to execute the heavy oil storage tank water-cutting control method of any one of the above-described technical solutions.
[0032] Compared with the prior art, the present invention has one or more of the following beneficial effects:
[0033] 1. The water-cutting control method for heavy oil storage tanks of the present invention is based on machine vision technology. Through image segmentation processing, it can achieve precise water-cutting control even when applied to scenarios where oil and water densities are similar. At the same time, the present invention reduces the error caused by the superposition of oil and water in a single-direction image by acquiring images in both directions.
[0034] 2. The water-cutting control method and water-cutting system for heavy oil storage tanks of the present invention are stable in operation, easy to maintain, and can maintain good stability and reliability even under high-intensity operation.
[0035] 3. A convolutional neural network model is established using historical images. The model is then used to pre-judge water-cutting control in real-time images. Only when the judgment results of the two images are inconsistent are the real-time images segmented and weighted to make a judgment, which improves the response efficiency of water-cutting control.
[0036] 4. The heavy oil storage tank water-cutting system of the present invention is equipped with a heat-insulating component to prevent crude oil from adhering due to low temperature, thereby affecting the identification effect.
[0037] 5. By setting up a flushing unit, the attached heavy oil is washed away to avoid affecting the next water cutting control and to ensure the accuracy of water cutting control.
[0038] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it according to the contents of the specification, and to make the above and other objects, technical features and advantages of the present invention easier to understand, one or more preferred embodiments are listed below and described in detail with reference to the accompanying drawings. Attached Figure Description
[0039] Figure 1 This is a flowchart of a water-cut control method for heavy oil storage tanks according to an embodiment of the present invention.
[0040] Figure 2 This is a schematic diagram of a heavy oil storage tank water-cutting system according to an embodiment of the present invention.
[0041] Figure 3 This is a schematic diagram of the hardware structure of an electronic device for implementing a water-cut control method for heavy oil storage tanks according to an embodiment of the present invention. Detailed Implementation
[0042] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings, but it should be understood that the scope of protection of the present invention is not limited to the specific embodiments.
[0043] Unless otherwise expressly stated, throughout the specification and claims, the term "comprising" or its variations such as "including" or "comprises" shall be understood to include the stated elements or components without excluding other elements or other components.
[0044] In this document, for ease of description, spatial relative terms such as “below,” “under,” “down,” “above,” “above,” “up,” etc., are used to describe the relationship of one element or feature to another element or feature in the accompanying drawings. It should be understood that spatial relative terms are intended to encompass different orientations of an object in use or operation, in addition to those depicted in the figures. For example, if an object in the figure is flipped, an element described as “below” or “under” another element or feature would be oriented “above” that element or feature. Thus, the exemplary term “below” can encompass both the downward and upward orientations. An object may also have other orientations (rotated 90 degrees or other orientations), and the spatial relative terms used herein should be interpreted accordingly.
[0045] In this document, the terms "first," "second," etc., are used to distinguish two different elements or parts, and are not used to define specific positions or relative relationships. In other words, in some embodiments, the terms "first," "second," etc., can also be used interchangeably.
[0046] like Figure 1 As shown, a water-cut control method for a heavy oil storage tank according to a specific embodiment of the present invention controls the water-cut pipeline of the heavy oil storage tank. It includes the following steps:
[0047] The S110 acquires top and side views of the fluid in the water-cutting pipe.
[0048] S120 uses a convolutional neural network to perform image recognition on top-view and side-view images to obtain recognition results.
[0049] Based on the identification result, S130 shuts off the water cutting pipe, returns to step 110, or proceeds to step S151.
[0050] S151 uses the ICM algorithm based on Markov random fields to segment the top and side views.
[0051] Because heavy oil and water have similar densities, heavy oil-water pipes often contain numerous oil-in-water and water-in-oil emulsions, resulting in unclear boundaries between the heavy oil and water in images captured by cameras inside the pipes. This invention employs an ICM algorithm based on Markov random fields for image segmentation to address the difficulty in distinguishing oil-water boundaries. For example, the image segmentation process is as follows:
[0052] Initialize the image, calculate its size, and obtain the image parameter set P(F / ω), where P is the conditional probability of the observation field F with respect to the label value ω, corresponding to different classifications λ∈Λ. Calculate the mean μ. λ and variance σ λβ is the coupling coefficient of the model, which yields the parameter set and the initial state θ0=(μ1,μ2,...,μ) of the label field ω in the oil-water image segmentation. L ,σ1,σ2,...,σ L , β). Iterate through each pixel of the image and take Initial segmentation complete. Set the total number of iterations to 500, k to the current iteration number, and the marker field for each pixel s. Let S be the set of all pixels s. Determine the convergence condition and calculate the current global capability value E. k :
[0053]
[0054] If |E k -E k If -1|≤Δ, then the global energy change is considered small, and the final segmentation result is obtained, where Δ is the threshold constant, representing the energy variable; if |E k -E k If -1|>Δ, then k+1, and re-traverse the pixels.
[0055] Based on the segmentation results, S152 obtains the heavy oil percentage P1 of the top view image and the heavy oil percentage P2 of the side view image.
[0056] S153 calculates the proportion of heavy oil in the fluid in the water-cutting pipe by weighted summation, n = μ1P1 + μ2P2, where μ1 is the weight of the top view image, μ2 is the weight of the side view image, and μ1 + μ2 = 1.
[0057] Furthermore, in one or more exemplary embodiments of the present invention, the weight μ1 of the top view image and the weight μ2 of the side view image can be determined based on the density of heavy oil. The smaller the density difference between heavy oil and water, the higher the weight of the top view image; the larger the density difference between heavy oil and water, the higher the weight of the side view image. Generally, the ratio of the density of heavy oil to the density of water is between 0.75 and 0.95. The larger the ratio, the smaller the density difference between heavy oil and water, the more uniform the oil-water mixing in the water-cutting pipe, and the larger the error of the top view image. Therefore, the weight of the top view image should be reduced. For example, the weight μ1 of the top view image is ρ0 - 0.45, where ρ0 is the ratio of the density of heavy oil to the density of water. Furthermore, in one or more exemplary embodiments of the present invention, the preset shut-off threshold n0 can be set to 5% to 8%. If more stringent environmental protection requirements need to be met, the preset shut-off threshold n0 is set to 5%.
[0058] S154 compares the calculated heavy oil percentage n in the water-cutting pipe with the preset shut-off threshold n0. When n≥n0, the water-cutting pipe is shut off; when n<n0, the water-cutting pipe is kept in its water-cutting state and the process returns to step S110.
[0059] Furthermore, in one or more exemplary embodiments of the present invention, in order to improve the efficiency of real-time analysis, a convolutional neural network is first used to perform image recognition on the top view image and the side view image. Step S120 includes:
[0060] S121 acquires historical top-view and side-view images of the fluid in the water-cutting pipe to establish a training sample set;
[0061] S122 assigns labels to each historical top-view image and historical side-view image in the training sample set, denoted as "cutting water" and "closing";
[0062] S123 inputs the training sample set and the corresponding set of labels into the convolutional neural network for training, and obtains the trained convolutional neural network model.
[0063] S124 uses a trained convolutional neural network model to identify the acquired top-view and side-view images and obtain the identification results.
[0064] Furthermore, in one or more exemplary embodiments of the present invention, step S130 includes: closing the water-cutting pipe when the recognition results of the top view image and the side view image at the same time are both "closed"; maintaining the water-cutting state of the water-cutting pipe and returning to step S110 when the recognition results of the top view image and the side view image at the same time are both "cutting water"; and proceeding to step S151 when the recognition results of the top view image and the side view image at the same time are different.
[0065] Further, in one or more exemplary embodiments of the present invention, step S122 includes: performing image segmentation on the historical top-view image and the historical side-view image using an ICM algorithm based on Markov random fields; obtaining the heavy oil percentage Q1 of the historical top-view image and the heavy oil percentage Q2 of the historical side-view image based on the segmentation results; when Q1 ≥ 5%, the historical top-view image is labeled "closed"; when Q1 < 5%, the historical top-view image is labeled "water cut"; when Q2 ≥ 5%, the historical side-view image is labeled "closed"; when Q2 < 5%, the historical side-view image is labeled "water cut". The processing method for historical images can refer to the above image segmentation method, and will not be repeated here.
[0066] The following describes in more detail the water-cutting control method, water-cutting system, electronic equipment, and storage medium for heavy oil storage tanks according to specific embodiments. It should be understood that the embodiments are merely exemplary and the present invention is not limited thereto.
[0067] Example 1
[0068] refer to Figure 2As shown, the water-cutting pipe 10 is connected to the bottom of the heavy oil storage tank (not shown in the figure). When the operator opens the water-cutting pipe 10 to cut the water, the heavy oil storage tank water-cutting system of this embodiment monitors the water-cutting situation and controls the closure of the water-cutting pipe. The heavy oil storage tank water-cutting system of this embodiment includes: an image acquisition unit 20, which is used to acquire top view and side view images of the fluid in the water-cutting pipe 10; an identification unit 30, which uses a convolutional neural network to perform image recognition on the top view and side view images to obtain identification results; a calculation unit 40, which uses an ICM algorithm based on Markov random fields to segment the top view and side view images, and obtains the heavy oil ratio P1 of the top view image and the heavy oil ratio P2 of the side view image according to the segmentation results; calculates the heavy oil ratio n of the fluid in the water-cutting pipe by weighted summation n = μ1P1 + μ2P2, where μ1 is the weight of the top view image, μ2 is the weight of the side view image, and μ1 + μ2 = 1; and a control unit 50, which is used to control the water-cutting valve of the water-cutting pipe 10 according to the identification results obtained by the identification unit 30 and the heavy oil ratio of the fluid in the water-cutting pipe calculated by the calculation unit 40.
[0069] In this embodiment, the identification unit 30 includes: a model acquisition submodule, which is used to establish a training sample set based on historical top-view and historical side-view images of fluid in the water-cutting pipe; to assign labels to each historical top-view and historical side-view image in the training sample set, denoted as "water-cutting" and "closed"; to input the training sample set and the corresponding assigned labels into a convolutional neural network for training, thereby obtaining a trained convolutional neural network model; and an image recognition submodule, which uses the trained convolutional neural network model to identify the acquired top-view and side-view images, thereby obtaining the recognition result.
[0070] In this embodiment, the image acquisition unit 20 may include: a transparent pipe section, which is a section of the water-cutting pipe; two cameras, which respectively acquire top-view and side-view images of the fluid in the water-cutting pipe through the transparent pipe section; and a heat insulation component, which maintains the operating temperature of the transparent pipe section at 45-55°C. For example, the heat insulation component may be a heating wire, a heat insulation layer, etc., laid on the outside of the water-cutting pipe wall.
[0071] In the heavy oil storage tank water-cutting system of this embodiment, a return oil pipeline and a flushing unit are provided. When the water-cutting valve is closed, the liquid in the water-cutting pipeline can be transported back to the heavy oil storage tank through the return oil pipeline, and then the flushing unit is used to flush the closed water-cutting pipeline with hot water.
[0072] Example 2
[0073] This embodiment provides a non-transitory (non-volatile) computer storage medium that stores computer-executable instructions that can execute the methods in any of the above method embodiments and achieve the same technical effect.
[0074] Example 3
[0075] This embodiment provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, which, when executed by a computer, cause the computer to perform the methods described above and achieve the same technical effects.
[0076] Example 4
[0077] Figure 3 This is a schematic diagram of the hardware structure of the electronic device for implementing the water-cut control method for heavy oil storage tanks according to this embodiment. The device includes one or more processors 610 and a memory 620. Taking one processor 610 as an example, the device may also include an input device 630 and an output device 640.
[0078] The processor 610, memory 620, input device 630, and output device 640 can be connected via a bus or other means. Figure 3 Taking the example of a connection between China and Israel via a bus.
[0079] The memory 620, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules. The processor 610 executes various functional applications and data processing of the electronic device by running the non-transitory software programs, instructions, and modules stored in the memory 620, thereby implementing the processing method of the above-described method embodiments.
[0080] The memory 620 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and applications required for at least one function; the data storage area may store data, etc. Furthermore, the memory 620 may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 620 may optionally include memory remotely located relative to the processor 610, and these remote memories may be connected to the processing device via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.
[0081] Input device 630 can receive input digital or character information and generate signal input. Output device 640 may include display devices such as a display screen.
[0082] One or more modules are stored in memory 620 and, when executed by one or more processors 610, execute:
[0083] S110 acquires top and side view images of the fluid in the water-cutting pipe;
[0084] S151 uses the ICM algorithm based on Markov random fields to segment the top view and side view images;
[0085] Based on the segmentation results, S152 obtains the heavy oil percentage P1 of the top view image and the heavy oil percentage P2 of the side view image.
[0086] S153 calculates the proportion of heavy oil in the fluid in the water-cutting pipe by weighted summation, n = μ1P1 + μ2P2, where μ1 is the weight of the top view image, μ2 is the weight of the side view image, and μ1 + μ2 = 1.
[0087] S154 compares the calculated heavy oil percentage n in the water-cutting pipe with the preset shut-off threshold n0. When n≥n0, the water-cutting pipe is shut off; when n<n0, the water-cutting pipe is kept in its water-cutting state and the process returns to step S110.
[0088] The above-described product can execute the methods provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the methods. Technical details not described in detail in this embodiment can be found in the methods provided in other embodiments of the present invention.
[0089] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.
[0090] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a general-purpose hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.
[0091] The foregoing description of specific exemplary embodiments of the present invention is for illustrative and explanatory purposes. These descriptions are not intended to limit the invention to the precise forms disclosed, and it will be apparent that many changes and variations can be made in accordance with the foregoing teachings. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application, thereby enabling those skilled in the art to implement and utilize various different exemplary embodiments of the invention, as well as various different choices and variations. Any simple modifications, equivalent changes, and alterations made to the foregoing exemplary embodiments should fall within the scope of protection of the present invention.
Claims
1. A method for controlling water shedding in a heavy oil storage tank, characterized in that, Includes the following steps: S110 acquires top and side view images of the fluid in the water-cutting pipe; S151 uses the ICM algorithm based on Markov random fields to segment the top view and side view images; Based on the segmentation results, S152 obtains the heavy oil percentage of the top-view image. The proportion of heavy oil in the side view image ; S153 calculates the heavy oil content of the fluid in the water-cutting pipe using a weighted summation method. In the formula, The weights of the top-view image, The weights of the side view image. Weights of the top-view image and the weights of the side view image The weights of the top-view image are determined based on the density of heavy oil. In the formula This is the ratio of the density of heavy oil to the density of water. S154 compares the calculated heavy oil content of the fluid in the water-cutting pipe. and preset shutdown threshold ,when When, shut off the water supply pipe; when At this time, maintain the water-cutting state of the water-cutting pipe and return to step S110.
2. The method for controlling water removal in heavy oil storage tanks according to claim 1, characterized in that, The preset shutdown threshold It ranges from 5% to 8%.
3. The method for controlling water cut-off in heavy oil storage tanks according to claim 1, characterized in that, Step S151 also includes: S120 uses a convolutional neural network to perform image recognition on top-view and side-view images to obtain recognition results; Based on the identification result, S130 shuts off the water cutting pipe, returns to step 110, or proceeds to step S151.
4. The method for controlling water cut-off in heavy oil storage tanks according to claim 3, characterized in that, Step S120 includes: S121 acquires historical top-view and side-view images of the fluid in the water-cutting pipe to establish a training sample set; S122 assigns labels to each historical top-view image and historical side-view image in the training sample set, denoted as "cutting water" and "closing"; S123 inputs the training sample set and the corresponding set of labels into the convolutional neural network for training, and obtains the trained convolutional neural network model. S124 uses a trained convolutional neural network model to identify the acquired top-view and side-view images and obtain the identification results.
5. The method for controlling water removal in heavy oil storage tanks according to claim 4, characterized in that, Step S130 includes: When the recognition results of both the top view and the side view at the same moment are "closed", the water cutting pipe is closed. When the recognition results of both the top view and the side view at the same moment are "water cutting", maintain the water cutting state of the water cutting pipe and return to step S110; and When the recognition results of the top view image and the side view image at the same time are different, proceed to step S151.
6. The method for controlling water cut-off in heavy oil storage tanks according to claim 4, characterized in that, Step S122 includes: The ICM algorithm based on Markov random fields is used to segment historical top-view and historical side-view images. Based on the segmentation results, the proportion of heavy oil in the historical top-view image was obtained. The proportion of heavy oil in historical side view images ; when At that time, the historical overhead view image is labeled "Closed"; when The historical aerial view image is labeled "water-cutting"; when When the historical side view image is displayed, the label is "Closed"; when... The historical side view image is labeled "cutting water".
7. A water-cutting system for a heavy oil storage tank, characterized in that, include: Image acquisition unit, which is used to acquire top view and side view images of the fluid in the water cutting pipe; The recognition unit uses a convolutional neural network to perform image recognition on top-view and side-view images to obtain recognition results; The computing unit uses the ICM algorithm based on Markov random fields to segment the top-view and side-view images. Based on the segmentation results, it obtains the heavy oil percentage of the top-view image. The proportion of heavy oil in the side view image The proportion of heavy oil in the fluid of the water-cutting pipe is calculated by weighted summation. In the formula, The weights of the top-view image, The weights of the side view image. Weights of the top-view image and the weights of the side view image The weights of the top-view image are determined based on the density of heavy oil. In the formula This is the ratio of the density of heavy oil to the density of water. A control unit is used to control the water-cutting pipe based on the identification result obtained by the identification unit and the heavy oil ratio of the fluid in the water-cutting pipe calculated by the calculation unit.
8. The heavy oil storage tank water-cutting system according to claim 7, characterized in that, The identification unit includes: The model acquisition submodule is used to establish a training sample set based on historical top-view and side-view images of the fluid in the water-cutting pipe; to assign labels to each historical top-view and side-view image in the training sample set, denoted as "water-cutting" and "closed"; and to input the training sample set and the corresponding labels into a convolutional neural network for training, thereby obtaining the trained convolutional neural network model. The image recognition submodule uses a trained convolutional neural network model to recognize the acquired top-view and side-view images and obtain the recognition results.
9. The heavy oil storage tank water-cutting system according to claim 7, characterized in that, The image acquisition unit includes: A transparent pipe section, which is a section of the water-cutting pipe; Two cameras, through the transparent pipe section, respectively capture top-view and side-view images of the fluid in the water-cutting pipe; and The insulation component maintains the working temperature of the transparent tube section at 45~55℃.
10. The heavy oil storage tank water-cutting system according to claim 7, characterized in that, Also includes: The flushing unit is used to flush the shut-off water pipes.
11. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, causes the at least one processor to perform the heavy oil storage tank water cut-off control method as described in any one of claims 1 to 6.
12. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer-executable instructions for causing the computer to perform the heavy oil storage tank water-cutting control method as described in any one of claims 1 to 6.
Citation Information
Patent Citations
Coal digestion tubular reaction device
CN110394143A
Time and location-based linking of captured medical information with medical records
US20210313051A1