Fire valve fault on-line diagnosis method and system

Abstract

The invention relates to the technical field of image processing, in particular to a fire valve fault online diagnosis method and system, and the method comprises the steps: obtaining the gas concentration characteristic quantity of each block according to the gray values of all pixel points in each block, obtaining the airflow vector of each pixel point and the airflow main vector of each block, according to the gas concentration characteristic quantity of each block, the airflow vector of the pixel point in each block and the airflow main vector of each block, obtaining the gas correlation degree of each block; obtaining a gas flow trend vector of each block and a region trend vector of each block according to the gas flow main vector of each block and the gas correlation degree, and obtaining an abnormal degree of each block; and obtaining all abnormal blocks and non-abnormal blocks in the infrared image according to the abnormal degree of each block, carrying out fault diagnosis on the fire valve according to all abnormal blocks and non-abnormal blocks in the infrared image, and giving an alarm when the fire valve has a fault. According to the invention, the infrared image is processed, so that the accuracy of fault diagnosis and detection of the fire valve is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly to an online diagnosis method and system for fire valve faults. Background Art

[0002] With the rapid development of the industrial and engineering fields, fire valves, as key components for controlling fluid flow, their normal operation is crucial for ensuring production safety and resource utilization efficiency. Fire valves are mainly used to regulate gas pressure and flow. Once a fault occurs and leads to gas leakage, it may not only cause serious safety accidents but also result in production stagnation and resource waste. Therefore, the online diagnosis and detection technology for fire valves is of extremely important significance in the industrial and engineering fields.

[0003] In the existing fire valve fault diagnosis technology, infrared thermal imaging cameras are widely used to collect infrared images of fire valves and nearby pipelines. Under normal circumstances, when gas leaks occur somewhere in the pipeline, the gas concentration in the pipeline will change, and this change can be diagnosed through the difference in gray-scale distribution in the infrared image. However, in practical applications, when the fire valve pipelines are very dense, the problem becomes complicated.

[0004] Specifically, when the leakage point of one pipeline is very close to the leakage point of another adjacent pipeline, if gas leaks occur simultaneously at both leakage points, the flow direction and trend of the leaked gas will affect each other. This mutual influence will make the difference in gray-scale distribution of the infrared image become blurred, resulting in the inability to accurately detect the faults of the fire valve when directly analyzing and detecting through the difference in gray-scale distribution of the infrared image.

[0005] To solve this problem, a more accurate and reliable online diagnosis method and system for fire valve faults are needed. This method and system should be able to overcome the interference caused by the mutual influence of leaked gas in the case of dense pipelines, accurately identify the fault state of the fire valve, and thus ensure production safety and resource utilization efficiency in the industrial and engineering fields. Summary of the Invention

[0006] The present invention provides an online diagnosis method and system for fire valve faults to solve the existing problems.

[0007] The online diagnosis method and system for fire valve faults of the present invention adopt the following technical solutions: An embodiment of the present invention provides an online diagnosis method for fire valve faults, and the method includes the following steps: Collect infrared images of the fire valve; Divide the infrared image to obtain several blocks, obtain the gas concentration feature quantity of each block according to the gray values of all pixel points in each block, obtain the air flow vector of each pixel point according to the gradient magnitude and gradient direction of each pixel point, obtain the main air flow vector of each block according to the air flow vectors of all pixel points within each block, and obtain the gas relevance of each block according to the gas concentration feature quantity of each block, the air flow vectors of pixel points within each block, and the main air flow vector of each block; Obtain the air flow trend vector of each block according to the main air flow vector and gas relevance of each block, obtain the adjacent blocks of each block, and obtain the regional trend vector of each block according to the air flow trend vector of each block and the air flow trend vectors of the adjacent blocks of each block. Obtain the degree of abnormality of each block according to the variance of the gas relevance of each block and all its adjacent blocks, the regional trend vector of each block, and the air flow trend vector; Obtain all abnormal blocks and non-abnormal blocks in the infrared image according to the degree of abnormality of each block, diagnose the failure of the fire protection valve according to all abnormal blocks and non-abnormal blocks in the infrared image, and give an alarm when the fire protection valve fails.

[0008] Further, the obtaining of the gas concentration feature quantity of each block according to the gray values of all pixel points in each block includes the following calculation formula: In the formula, represents the mean value of the gray values of all pixel points in the i-th block, represents the variance of the gray values of all pixel points in the i-th block, represents the gas concentration feature quantity of the i-th block.

[0009] Further, the obtaining of the air flow vector of each pixel point according to the gradient magnitude and gradient direction of each pixel point, and the obtaining of the main air flow vector of each block according to the air flow vectors of all pixel points within each block include the following specific steps: Take the gradient magnitude of each pixel point as the modulus length of the air flow vector of each pixel point, and take the gradient direction of each pixel point as the direction of the air flow vector of each pixel point; Add the air flow vectors of all pixel points in each block for vector data, and record the result as the main air flow vector of each block.

[0010] Further, the obtaining of the gas relevance of each block according to the gas concentration feature quantity of each block, the air flow vectors of pixel points within each block, and the main air flow vector of each block includes the following calculation formula: In the formula, represents the gas concentration feature quantity of the i-th block, represents the air flow vector of the j-th pixel point in the i-th block, represents the main air flow vector of the i-th block, represents the variance of the angle between the airflow vectors of all pixel points in the i-th block and the main airflow vector of the i-th block, n represents the number of all pixel points in the i-th block, represents the magnitude of the vector, represents the gas relevance of the i-th block, represents the cosine value of the angle between the main airflow vector of the i-th block and the airflow vector of the j-th pixel point in the i-th block.

[0011] Furthermore, obtaining the airflow trend vector of each block according to the main airflow vector and gas relevance of each block includes the following specific steps: Taking the gas relevance of each block as the magnitude of the airflow trend vector of each block, and taking the direction of the main airflow vector of each block as the direction of the airflow trend vector of each block, so as to obtain the airflow trend vector of each block.

[0012] Furthermore, obtaining the adjacent blocks of each block, and obtaining the regional trend vector of each block according to the airflow trend vector of each block and the airflow trend vector of the adjacent blocks of each block includes the following specific steps: Obtaining all blocks in the eight-neighborhood of each block, denoted as the adjacent blocks of each block, and adding the vector data according to the airflow trend vectors of each block and the corresponding adjacent blocks to obtain the regional trend vector of each block.

[0013] Furthermore, obtaining the degree of abnormality of each block according to the variance of the gas relevance of each block and all its adjacent blocks, the regional trend vector and the airflow trend vector of each block, and the calculation formula includes: In the formula, represents the regional trend vector of the i-th block, represents the airflow trend vector of the v-th adjacent block of the i-th block, m represents the number of all adjacent blocks of the i-th block, represents the variance of the gas relevance of the i-th block and all its corresponding adjacent blocks, represents the magnitude of the vector, represents the degree of abnormality of the i-th block, represents the cosine value of the angle between the regional trend vector of the i-th block and the airflow trend vector of the v-th adjacent block of the i-th block.

[0014] Furthermore, obtaining all abnormal blocks and non-abnormal blocks in the infrared image according to the degree of abnormality of each block, and diagnosing the fire valve failure according to all abnormal blocks and non-abnormal blocks in the infrared image includes the following specific steps: For any block, when the degree of abnormality of the block is greater than or equal to the preset threshold T1, it is determined that the block is an abnormal block; when the degree of abnormality of the block is less than the preset threshold T1, it is determined that the block is a non-abnormal block; Calculate the ratio of the number of all abnormal blocks in the infrared image to the total number of all blocks in the infrared image, which is denoted as the proportion of abnormal blocks in the infrared image. When the proportion of abnormal blocks in the infrared image is greater than or equal to the preset threshold T2, it is determined that the fire valve is faulty; when the proportion of abnormal blocks in the infrared image is less than the preset threshold T2, it is determined that the fire valve is not faulty.

[0015] Furthermore, the fire valve fault is a gas leakage fault.

[0016] The present invention also provides an on-line diagnosis system for fire valve faults, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the above-mentioned any one of the steps is implemented.

[0017] The beneficial effects of the technical solution of the present invention are as follows: the gas concentration feature quantity of each block is obtained according to the gray values of all pixel points in each block, the air flow vector of each pixel point is obtained according to the gradient amplitude and gradient direction of each pixel point, the main air flow vector of each block is obtained according to the air flow vectors of all pixel points in each block, and the gas correlation degree of each block is obtained according to the gas concentration feature quantity of each block, the air flow vectors of pixel points in each block, and the main air flow vector of each block, excluding the mutual influence of the flow direction and trend of the leaked gas at multiple leakage points, and initially determining the blocks with abnormalities; the degree of abnormality of each block is obtained according to the variance of the gas correlation degree between each block and all adjacent blocks, the regional trend vector and the air flow trend vector of each block; all abnormal blocks and non-abnormal blocks in the infrared image are obtained according to the degree of abnormality of each block, and the fire valve fault is diagnosed according to all abnormal blocks and non-abnormal blocks in the infrared image, and an alarm is given when the fire valve fails, improving the accuracy of the diagnosis and detection of fire valve faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is a step flow chart of an on-line diagnosis method for fire valve faults of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] To further illustrate the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following specifically describes, in conjunction with the accompanying drawings and preferred embodiments, a method and system for online diagnosis of fire-fighting valve failures proposed according to the present invention, including its specific implementation manners, structures, features and effects. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0022] The following specifically describes the specific solution of a method and system for online diagnosis of fire-fighting valve failures provided by the present invention with reference to the accompanying drawings.

[0023] Please refer to Figure 1 , which shows a flowchart of the steps of a method for online diagnosis of fire-fighting valve failures provided by an embodiment of the present invention. The method includes the following steps: Step S001: Collect infrared images of the fire-fighting valve.

[0024] There are many types of fire-fighting valves, which can be divided into various types according to their functions, structures, driving methods, etc. Common fire-fighting valves include, but are not limited to, alarm valves, signal valves, check valves, safety valves, pressure reducing valves, test valves / drain valves. After long-term use, the valve seals will lose their sealing performance due to aging, wear or corrosion, resulting in gas leakage. Or due to cracks in the valve body or valve cover, that is, cracks appear in the valve body or valve cover due to manufacturing defects, poor welding quality, water hammer impact, freezing, etc., resulting in gas leakage.

[0025] It should be noted that in order to analyze whether there is an abnormal situation of gas leakage at the fire-fighting valve, it is necessary to collect infrared images of the fire-fighting valve for analysis.

[0026] Specifically, use an infrared thermal imaging camera to take infrared thermal imaging images of the fire-fighting valve, and perform gray-scale preprocessing on the infrared thermal imaging images of the fire-fighting valve to obtain the infrared images of the preprocessed fire-fighting valve.

[0027] So far, the infrared images of the fire-fighting valve are obtained.

[0028] Step S002: Divide the infrared image into several blocks. Obtain the gas concentration feature quantity of each block based on the gray values of all pixel points in each block. Obtain the air flow vector of each pixel point based on the gradient amplitude and gradient direction of each pixel point. Obtain the main air flow vector of each block based on the air flow vectors of all pixel points within each block. Obtain the gas relevance of each block based on the gas concentration feature quantity of each block, the air flow vectors of pixel points within each block, and the main air flow vector of each block.

[0029] It should be noted that when a failure occurs at the fire protection valve, gas leakage occurs at the failure point, causing the gas concentration in the pipeline to change. Therefore, when the fire protection valve fails, the greater the difference in the gray values of the pixel points within the block closer to the fire protection valve in the image. That is, it is possible to analyze whether each block is abnormal based on the gray difference between all pixel points in each block.

[0030] Furthermore, it should be noted that since the collected image is an infrared image, the gas has a relatively high gray value after grayscale conversion of the infrared image. Therefore, the higher the average value of the gray values of all pixel points in each block, the greater the possibility that the block is a normal block, and vice versa. Also, because when the fire protection valve has no defects or leaks, the gas concentration distribution within each block is relatively uniform. Therefore, it is possible to analyze whether each block is a normal block based on the variance of the gray values of all pixel points in each block.

[0031] Specifically, a threshold T is preset. In this embodiment, T = 100 is taken as an example for description, and this embodiment does not make specific limitations. T can be determined according to specific implementation situations. The infrared image is evenly divided into T blocks.

[0032] Obtain the gas concentration feature quantity of each block based on the gray values of all pixel points in each block, which is expressed by the formula: In the formula, represents the average value of the gray values of all pixel points in the i-th block, represents the variance of the gray values of all pixel points in the i-th block, represents the gas concentration feature quantity of the i-th block.

[0033] Among them, the larger the average value of the grayscale values of all pixel points in each block, the greater the possibility that the block is a normal block; the smaller the average value of the grayscale values of all pixel points in each block, the smaller the possibility that the block is a normal block. When the variance of the grayscale values of all pixel points in each block is smaller, it indicates that the grayscale values of the pixel points in the block are relatively close, that is, the grayscale difference of the block is smaller, then the possibility that the block is a normal block is greater, that is, the corresponding gas concentration characteristic quantity is larger. When the variance of the grayscale values of all pixel points in each block is larger, it indicates that the grayscale values of the pixel points in the block are quite different, then the possibility that the block is a normal block is smaller, that is, the corresponding gas concentration characteristic quantity is smaller.

[0034] Thus, the gas concentration characteristic quantity of each block is obtained.

[0035] Obtain the gradient magnitude and gradient direction of all pixel points. Take the gradient magnitude of each pixel point as the modulus length of the air flow vector, and take the gradient direction as the direction of the air flow vector. Thus, the air flow vector of each pixel point is obtained.

[0036] Add the air flow vectors of all pixel points in each block for vector data, and record the result as the main air flow vector of each block.

[0037] It should be noted that when the fire valve does not malfunction, the gas flow direction in the pipeline is relatively the same, that is, the included angle between the directions of the air flow vectors of the pixel points in each block should be small. When the fire valve malfunctions, due to the influence of external air and pressure, the gas flow direction in the pipeline becomes more inconsistent, that is, the gas in the pipeline is relatively chaotic. Therefore, the difference between the air flow vectors of all pixel points in each block and the main air flow vector of each block can be analyzed.

[0038] Specifically, the gas relevance of each block is obtained according to the gas concentration characteristic quantity of each block, the air flow vectors of the pixel points in each block, and the main air flow vector of each block. It is expressed by the formula: In the formula, represents the gas concentration characteristic quantity of the i-th block, represents the air flow vector of the j-th pixel point in the i-th block, represents the main air flow vector of the i-th block, represents the variance of the included angle between the air flow vectors of all pixel points in the i-th block and the main air flow vector of the i-th block. n represents the number of all pixel points in the i-th block, represents the modulus length of the vector, represents the gas relevance of the i-th block, represents the cosine value of the included angle between the main air flow vector of the i-th block and the air flow vector of the j-th pixel point in the i-th block.

[0039] Among them, It represents the average value of the lengths of the air flow vectors of all pixel points within the i-th block mapped onto the main air flow vector of the i-th block. It represents the ratio of the average value of the lengths of the air flow vectors of all pixel points within the i-th block mapped onto the main air flow vector of the i-th block to the modulus length of the main air flow vector of the i-th block. When this ratio is larger, that is, closer to 1, it indicates that the gas flow direction within the i-th block is more uniform, and the greater the possibility that the i-th block is a normal block. When this ratio is smaller, that is, closer to 0, it indicates that the gas flow direction within the i-th block is more non-uniform, that is, more chaotic, and the greater the possibility that the i-th block is an abnormal block, that is, the greater the possibility that there is a fault in the fire protection valve.

[0040] Thus, the gas correlation degree of each block is obtained.

[0041] Step S003: Obtain the air flow trend vector of each block based on the main air flow vector and gas correlation degree of each block, obtain the adjacent blocks of each block, obtain the regional trend vector of each block according to the air flow trend vector of each block and the air flow trend vectors of the adjacent blocks of each block, and obtain the abnormality degree of each block according to the variance of the gas correlation degrees of each block and all adjacent blocks, the regional trend vector of each block, and the air flow trend vector.

[0042] It should be noted that in the above steps, only the data of all pixel points in each block are analyzed. However, when all pixel points in a block are located at the defect, the difference between the air flow vectors of all pixel points in this block is relatively small at this time, and then this block can also be considered as a normal block; therefore, blocks in the neighborhood of a block can be selected for joint analysis, and the difference between the vectors of all blocks in the neighborhood is used to further analyze whether there is a fault.

[0043] Furthermore, it should be noted that when a fault occurs at the fire protection valve, due to the pressure difference inside and outside the pipeline, the pressure of the gas at the notch changes, that is, there is a difference in the trend between the vectors of pixel points in the image, and the difference is relatively large, while the difference in the trend between the vectors of pixel points in the normal area is relatively small.

[0044] Specifically, take the gas correlation degree of each block as the modulus length of the air flow trend vector of each block, and take the direction of the main air flow vector of each block as the direction of the air flow trend vector of each block, so as to obtain the air flow trend vector of each block.

[0045] Obtain all blocks in the eight-neighborhood of each block, denoted as the adjacent blocks of each block, and perform vector addition of the vector data according to the air flow trend vectors of each block and the corresponding adjacent blocks to obtain the regional trend vector of each block.

[0046] Obtain the abnormality degree of each block according to the variance of the gas correlation degrees of each block and all adjacent blocks, the regional trend vector of each block, and the air flow trend vector. It is specifically expressed by the formula as: In the formula, represents the regional trend vector of the i-th block, represents the airflow trend vector of the v-th adjacent block of the i-th block, and m represents the number of all adjacent blocks of the i-th block. represents the variance of the gas correlation degree between the i-th block and all its corresponding adjacent blocks, represents the modulus of the vector, represents the degree of abnormality of the i-th block, represents the cosine value of the angle between the regional trend vector of the i-th block and the airflow trend vector of the v-th adjacent block of the i-th block.

[0047] Among them, represents the difference between the modulus of the regional trend vector of the i-th block and the modulus of the airflow trend vector of the v-th adjacent block of the i-th block, represents the average value of the differences between the modulus of the regional trend vector of the i-th block and the moduli of the airflow trend vectors of all its adjacent blocks. When the average value of the differences between the moduli is larger, the block is more abnormal, that is, the possibility of a fire protection valve failure is greater. When the average value of the differences between the moduli is smaller, the block is more normal, that is, the possibility of a fire protection valve failure is smaller; represents the average value of the cosine values of the angles between the regional trend vector of the i-th block and the airflow trend vectors of all its adjacent blocks. When the average value of the cosine values of the angles is smaller, the block is more normal, that is, the possibility of a fire protection valve failure is smaller. When the average value of the cosine values of the angles is larger, the block is more abnormal, that is, the possibility of a fire protection valve failure is greater. When the variance of the gas correlation degree between each block and all its corresponding adjacent blocks is larger, it indicates that the gas correlation degrees in this block and its adjacent blocks are more inconsistent, that is, the possibility of a fire protection valve failure is greater. When the variance of the gas correlation degree between each block and all its corresponding adjacent blocks is smaller, it indicates that the gas correlation degrees in this block and its adjacent blocks are more consistent, that is, the possibility of a fire protection valve failure is smaller.

[0048] Thus, the degree of abnormality of each block is obtained.

[0049] Step S004: Obtain all abnormal blocks and non-abnormal blocks in the infrared image according to the degree of abnormality of each block, and perform diagnosis of fire protection valve failures based on all abnormal blocks and non-abnormal blocks in the infrared image.

[0050] Preset a threshold T1. In this embodiment, T1 = 0.8 is used as an example for description, and this embodiment does not make specific limitations. T1 can be determined according to specific implementation situations. For any block, when the degree of abnormality of the block is greater than or equal to the preset threshold T, it is determined that the block is an abnormal block; when the degree of abnormality of the block is less than the preset threshold T, it is determined that the block is a non-abnormal block.

[0051] So far, all abnormal blocks and non-abnormal blocks in the infrared image are obtained.

[0052] Preset a threshold T2. In this embodiment, T2 = 0.5 is taken as an example for description. This embodiment does not make specific limitations, and T2 can be determined according to specific implementation situations.

[0053] Calculate the ratio of the number of all abnormal blocks in the infrared image to the total number of all blocks in the infrared image, which is recorded as the proportion of abnormal blocks in the infrared image. When the proportion of abnormal blocks in the infrared image is greater than or equal to the preset threshold T2, it is determined that the fire valve is faulty; when the proportion of abnormal blocks in the infrared image is less than the preset threshold T2, it is determined that the fire valve is not faulty.

[0054] Alarm when the fire valve fails.

[0055] This embodiment provides an on-line diagnosis system for fire valve faults, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, steps S001 to S004 are implemented.

[0056] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An online diagnostic method for fire valve failures, characterized in that, The method includes the following steps: Collect the infrared image of the fire valve; Divide the infrared image to obtain several blocks, obtain the gas concentration feature quantity of each block according to the gray values of all pixel points in each block, obtain the air flow vector of each pixel point according to the gradient amplitude and gradient direction of each pixel point, obtain the main air flow vector of each block according to the air flow vectors of all pixel points in each block, and obtain the gas relevance of each block according to the gas concentration feature quantity of each block, the air flow vectors of pixel points in each block, and the main air flow vector of each block; Obtain the air flow trend vector of each block according to the main air flow vector and gas relevance of each block, obtain the adjacent blocks of each block, obtain the regional trend vector of each block according to the air flow trend vector of each block and the air flow trend vectors of the adjacent blocks of each block, and obtain the abnormality degree of each block according to the variance of the gas relevance of each block and all its adjacent blocks, the regional trend vector of each block, and the air flow trend vector; Obtain all abnormal blocks and non-abnormal blocks in the infrared image according to the abnormality degree of each block, diagnose the faults of the fire valve according to all abnormal blocks and non-abnormal blocks in the infrared image, and give an alarm when the fire valve fails.

2. The online diagnosis method for fire valve faults according to claim 1, characterized in that, The obtaining of the gas concentration feature quantity of each block according to the gray values of all pixel points in each block includes the following calculation formula: Wherein, represents the mean value of the gray values of all pixel points in the i-th block, represents the variance of the gray values of all pixel points in the i-th block, represents the gas concentration characteristic quantity of the i-th block.

3. The on-line fault diagnosis method for a fire protection valve according to claim 1, characterized in that The obtaining of the air flow vector of each pixel point according to the gradient amplitude and gradient direction of each pixel point, and the obtaining of the main air flow vector of each block according to the air flow vectors of all pixel points in each block include the following specific steps: Take the gradient amplitude of each pixel point as the modulus length of the air flow vector of each pixel point, and take the gradient direction of each pixel point as the direction of the air flow vector of each pixel point; Add the air flow vectors of all pixel points in each block for vector data, and record the result as the main air flow vector of each block.

4. The on-line fault diagnosis method of a fire control valve according to claim 1, characterized in that, The obtaining of the gas relevance of each block according to the gas concentration feature quantity of each block, the air flow vectors of pixel points in each block, and the main air flow vector of each block includes the following calculation formula: Wherein, represents the gas concentration characteristic quantity of the i-th block, represents the airflow vector of the j-th pixel point in the i-th block, represents the main airflow vector of the i-th block, represents the variance of the angle between the airflow vectors of all pixel points in the i-th block and the main airflow vector of the i-th block, and n represents the number of all pixel points in the i-th block, represents the modulus of the vector, represents the gas correlation degree of the i-th block, represents the cosine value of the angle between the main airflow vector of the i-th block and the airflow vector of the j-th pixel point in the i-th block.

5. The online diagnosis method for fire valve faults according to claim 1, characterized in that, The obtaining of the air flow trend vector of each block according to the main air flow vector and gas relevance of each block includes the following specific steps: Take the gas relevance of each block as the modulus length of the air flow trend vector of each block, and take the direction of the main air flow vector of each block as the direction of the air flow trend vector of each block, so as to obtain the air flow trend vector of each block.

6. The on-line fault diagnosis method for a fire protection valve according to claim 1, characterized in that The obtaining of the adjacent blocks of each block, and the obtaining of the regional trend vector of each block according to the air flow trend vector of each block and the air flow trend vectors of the adjacent blocks of each block include the following specific steps: Obtain all blocks in the eight-neighborhood of each block, record them as the adjacent blocks of each block, and add the air flow trend vectors of each block and the corresponding adjacent blocks for vector data to obtain the regional trend vector of each block.

7. The online fault diagnosis method for a fire protection valve according to claim 1, wherein The obtaining of the abnormality degree of each block according to the variance of the gas relevance of each block and all its adjacent blocks, the regional trend vector of each block, and the air flow trend vector includes the following calculation formula: In the formula, represents the regional trend vector of the i-th block, represents the air flow trend vector of the v-th adjacent block of the i-th block, and m represents the number of all adjacent blocks of the i-th block. represents the variance of the gas correlation degree between the i-th block and all its corresponding adjacent blocks, represents the modulus of the vector, represents the degree of abnormality of the i-th block, represents the cosine value of the angle between the regional trend vector of the i-th block and the air flow trend vector of the v-th adjacent block of the i-th block.

8. The on-line fault diagnosis method for a fire control valve according to claim 1, characterized in that The obtaining of all abnormal blocks and non-abnormal blocks in the infrared image according to the abnormality degree of each block, and the diagnosing of the faults of the fire valve according to all abnormal blocks and non-abnormal blocks in the infrared image include the following specific steps: For any block, if the abnormality degree of the block is greater than or equal to the preset threshold T1, then the block is determined to be an abnormal block; if the abnormality degree of the block is less than the preset threshold T1, then the block is determined to be a non-abnormal block. Calculate the ratio of the number of all abnormal blocks in the infrared image to the total number of all blocks in the infrared image, which is denoted as the proportion of abnormal blocks in the infrared image. When the proportion of abnormal blocks in the infrared image is greater than or equal to the preset threshold T2, it is determined that the fire protection valve has a fault; when the proportion of abnormal blocks in the infrared image is less than the preset threshold T2, it is determined that the fire protection valve has no fault.

9. The online fault diagnosis method for a fire protection valve according to claim 1, characterized in that, The fault of the fire protection valve is a gas leakage fault.

10. An on-line fault diagnosis system for a fire protection valve, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the online diagnosis method for the fire protection valve fault as described in any one of claims 1-8.

Images