Method and device for monitoring heat leakage in chemical process

Abstract

The invention relates to the technical field of heat leakage detection, in particular to a method and a device for monitoring heat leakage in a chemical process. According to the method, spatial distance matrix information is introduced into a covariance matrix of thermal imaging image pixels, so that the spatial relationship between pixel points can be better reflected; a feature matrix is calculated according to the covariance matrix, so that data features are extracted to realize data dimension reduction; and determining a T2 contribution index and an SPE contribution index of each pixel point of the to-be-tested thermal image data based on the feature matrix and the to-be-tested thermal image data. However, whether abnormity occurs or not is judged only according to the contribution degree index of a single pixel, and a serious false alarm problem is easy to generate, so that T2 contribution degree indexes and SPE contribution degree indexes of all pixels in an adjacent region of each pixel need to be comprehensively considered, a regional detection index needs to be constructed, and an abnormal region needs to be determined. Therefore, the position of thermal leakage in the chemical process can be accurately detected.

Description

Technical Field

[0001] The invention relates to the technical field of heat leakage detection, and in particular to a heat leakage monitoring method and device for a chemical process. Background Art

[0002] The chemical industry is the most important manufacturing component of the process industry, and is a pillar industry for economic and social development and an important cornerstone of the real economy. However, there are also safety hazards of heat leakage in the chemical process. Once the high-temperature leakage flows into water, it will explode, causing huge casualties and equipment property losses. In addition, the workshop where the accident occurred has a slow repair speed and a long recovery period, which will also have a great impact on the normal production of the factory. At present, the response and treatment of high-temperature leakage accidents mainly focuses on strengthening the preliminary safety management work and continuously improving the production process, such as strengthening the preparation and drills of accident emergency plans, strengthening the safety production responsibility system, and improving the production process and process parameters. However, there is a lack of intelligent and automated forecasting and warning devices for heavy and extremely large high-temperature leakage accidents. Once a heat leakage explosion accident occurs, it will cause irreparable huge losses. Therefore, it is of great significance to carry out research on high-temperature heat leakage detection.

[0003] Infrared thermal imaging technology uses photoelectric technology to detect infrared signals of specific wavelength bands of thermal radiation from objects, converts the signals into images and graphics that can be distinguished by human vision, and can further calculate the temperature value. Compared with the point temperature measurement method of traditional temperature sensors, infrared thermal imagers have the characteristics of wide monitoring range, high monitoring density, and non-contact detection. They can effectively reflect the surface temperature distribution characteristics of the monitored object, thereby quickly and effectively detecting temperature anomalies and heat leaks.

[0004] However, most of the existing technical solutions focus on a single specific device and are heavily dependent on the mechanism and digital model of the monitored object. They are methods based on quantitative and qualitative models and are not suitable for chemical processes.

[0005] Based on this, the present invention proposes a heat leak monitoring method and device for a chemical process to solve the above technical problems. Summary of the invention

[0006] The present invention describes a method and device for monitoring heat leakage in a chemical process, which can accurately detect the location of heat leakage in the chemical process.

[0007] According to a first aspect, the present invention provides a method for monitoring heat leaks in a chemical process, the method comprising:

[0008] Acquire a thermal image data set and thermal image data to be tested; wherein the thermal image data set includes thermal imaging samples at multiple different times;

[0009] Performing region segmentation on each of the thermal imaging image samples to obtain segmented thermal imaging image samples;

[0010] Determining regional feature influencing factors based on the segmented thermal imaging image samples;

[0011] Based on the regional feature influencing factor, determining the spatial distance between every two pixel points in the segmented thermal imaging image sample;

[0012] The thermal image data set is converted to obtain a two-dimensional matrix of the thermal image data set, and a spatial distance matrix of the thermal image data set is determined according to an arrangement relationship of the two-dimensional matrix and a spatial distance between each two pixel points;

[0013] Determining a covariance matrix based on the spatial distance matrix and the two-dimensional matrix;

[0014] Based on the covariance matrix, determining a feature matrix;

[0015] Based on the characteristic matrix and the thermal image data to be measured, determine the T of each pixel of the thermal image data to be measured. 2 Contribution index and SPE contribution index;

[0016] According to the T 2 The contribution index and the SPE contribution index determine the T of the neighboring area of ​​each pixel of the thermal image data to be tested. 2 The contribution score of and the contribution score of SPE;

[0017] Based on the T 2 The contribution score of and the contribution score of the SPE are used to determine the abnormal area.

[0018] According to a second aspect, the present invention provides a thermal leak detection device for a chemical process, comprising:

[0019] An acquisition unit is configured to acquire a thermal image data set and thermal image data to be tested; wherein the thermal image data set includes a plurality of thermal imaging image samples at different times;

[0020] A first data processing unit is configured to perform region segmentation on each of the thermal imaging image samples to obtain segmented thermal imaging image samples;

[0021] A second data processing unit is configured to determine a regional feature influencing factor based on the segmented thermal imaging image sample;

[0022] A third data processing unit is configured to determine the spatial distance between every two pixels in the segmented thermal imaging image sample based on the regional feature influencing factor;

[0023] a fourth data processing unit configured to convert the thermal image data set to obtain a two-dimensional matrix of the thermal image data set, and determine a spatial distance matrix of the thermal image data set according to an arrangement relationship of the two-dimensional matrix and a spatial distance between each two pixel points;

[0024] a fifth data processing unit, configured to determine a covariance matrix based on the spatial distance matrix and the two-dimensional matrix;

[0025] a sixth data processing unit, configured to determine a feature matrix based on the covariance matrix;

[0026] The seventh data processing unit is configured to determine the T of each pixel of the thermal image data to be tested based on the feature matrix and the thermal image data to be tested. 2 Contribution index and SPE contribution index;

[0027] An eighth data processing unit is configured to 2 The contribution index and the SPE contribution index determine the T of the neighboring area of ​​each pixel of the thermal image data to be tested. 2 The contribution score of and the contribution score of SPE;

[0028] A ninth data processing unit is configured to 2 The contribution score of and the contribution score of the SPE are used to determine the abnormal area.

[0029] In a third aspect, an embodiment of this specification further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in any embodiment of this specification is implemented.

[0030] In a fourth aspect, an embodiment of the present specification further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed in a computer, the computer is caused to execute the method described in any embodiment of the present specification.

[0031] According to the method and device for monitoring heat leakage in a chemical process provided by the present invention, first, a thermal image data set and thermal image data to be tested are obtained, and then each thermal image sample is segmented into regions to obtain segmented thermal image samples. In the thermal image samples, since the pixels at the edges of two different objects are adjacent but their spatial distance is very far, it is necessary to determine the regional feature influencing factor based on the segmented thermal image samples. The regional feature influencing factor is used to characterize the influence of the spatial distance. Then, the spatial distance between each two pixels in the segmented thermal image sample is determined based on the regional feature influencing factor. The thermal image data set is segmented into regions. The two-dimensional matrix of the thermal image data set is converted, and the spatial distance matrix of the thermal image data set is determined according to the arrangement relationship of the two-dimensional matrix and the spatial distance between every two pixels. The arrangement relationship includes horizontal expansion and vertical expansion. The spatial distance matrix must be consistent with the arrangement relationship of the two-dimensional matrix. Then, the covariance matrix is ​​determined according to the spatial distance matrix and the two-dimensional matrix. The spatial distance matrix is ​​added to the covariance matrix, which can better reflect the spatial relationship between each pixel. Then, the feature matrix is ​​determined according to the covariance matrix. Based on the feature matrix and the thermal image data to be tested, the T of each pixel of the thermal image data to be tested is determined. 2 Contribution index and SPE contribution index, when T 2 When the contribution index and SPE contribution index exceed the preset threshold, it can only indicate that the corresponding variable (pixel position) may have a problem. However, a high contribution rate does not mean a high probability of failure, so some normal pixels may be misdiagnosed as the source of failure, while some faulty pixels may be diagnosed as normal. Therefore, it is also necessary to consider the T 2 Contribution index and SPE contribution index determine the T of the neighboring area of ​​each pixel of the thermal image data to be tested. 2 The contribution score of T 2 The contribution fraction of the chemical process and the contribution fraction of the SPE are used to determine the abnormal area. This can accurately detect the location of heat leaks in chemical processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0033] Figure 1 A schematic flow chart of a method for monitoring heat leaks in a chemical process according to an embodiment is shown;

[0034] Figure 2 A schematic block diagram of a heat leak detection device for a chemical process according to an embodiment is shown;

[0035] Figure 3 A schematic diagram showing a region segmentation of a thermal imaging image according to an embodiment;

[0036] Figure 4 A schematic diagram of a simulated industrial storage tank according to one embodiment is shown;

[0037] Figure 5 A fault monitoring diagram according to an embodiment is shown.

[0038] Figure 6 A fault monitoring diagram according to one embodiment is shown;

[0039] Figure 7 A fault monitoring diagram according to one embodiment is shown;

[0040] Figure 8 A schematic diagram of fault location according to an embodiment is shown. DETAILED DESCRIPTION

[0041] The solution provided by the present invention is described below in conjunction with the accompanying drawings.

[0042] Figure 1 A schematic diagram of a process flow diagram of a method for monitoring heat leaks in a chemical process according to an embodiment is shown. It is understood that the method can be executed by any device, equipment, platform, or equipment cluster having computing and processing capabilities. Figure 1 As shown, the method includes:

[0043] Step 100: Acquire a thermal image data set and thermal image data to be tested; wherein the thermal image data set includes thermal image samples at multiple different times;

[0044] Step 102, performing region segmentation on each thermal imaging sample to obtain a segmented thermal imaging sample;

[0045] Step 104: determining the regional feature influencing factor based on the segmented thermal imaging image sample;

[0046] Step 106: Determine the spatial distance between every two pixels in the segmented thermal imaging sample based on the regional feature influencing factor;

[0047] Step 108: convert the thermal image data set to obtain a two-dimensional matrix of the thermal image data set, and determine the spatial distance matrix of the thermal image data set according to the arrangement relationship of the two-dimensional matrix and the spatial distance between every two pixels;

[0048] Step 110: determining a covariance matrix based on the spatial distance matrix and the two-dimensional matrix;

[0049] Step 112: Determine a feature matrix based on the covariance matrix;

[0050] Step 114: Determine the T of each pixel of the thermal image data to be tested based on the feature matrix and the thermal image data to be tested. 2 Contribution index and SPE contribution index;

[0051] Step 116: According to T 2 Contribution index and SPE contribution index determine the T of the neighboring area of ​​each pixel of the thermal image data to be tested. 2 The contribution score of and the contribution score of SPE;

[0052] Step 118: Based on T 2 The contribution score of and the contribution score of SPE are used to determine the abnormal area.

[0053] In this embodiment, first, a thermal image data set and thermal image data to be tested are obtained, and then each thermal image sample is segmented to obtain a segmented thermal image sample. In the thermal image sample, since the pixels at the edges of two different objects are adjacent but their spatial distance is very far, a regional feature influencing factor is determined based on the segmented thermal image sample. The regional feature influencing factor is used to characterize the influence of the spatial distance. Then, based on the regional feature influencing factor, the spatial distance between every two pixels in the segmented thermal image sample is determined. The thermal image data set is converted to obtain a two-dimensional matrix of the thermal image data set. Then, based on the arrangement relationship of the two-dimensional matrix and the spatial distance between every two pixels, a spatial distance matrix of the thermal image data set is determined. The arrangement relationship includes horizontal expansion and vertical expansion. The spatial distance matrix must be consistent with the arrangement relationship of the two-dimensional matrix. Then, based on the spatial distance matrix and the two-dimensional matrix, a covariance matrix is ​​determined. The spatial distance matrix is ​​added to the covariance matrix, which can better reflect the spatial relationship between each pixel. Then, based on the covariance matrix, a feature matrix is ​​determined. Based on the feature matrix and the thermal image data to be tested, the T of each pixel of the thermal image data to be tested is determined. 2 Contribution index and SPE contribution index, when T 2 When the contribution index and SPE contribution index exceed the preset threshold, it can only indicate that the corresponding variable (pixel position) may have a problem. However, a high contribution rate does not mean a high probability of failure, so some normal pixels may be misdiagnosed as the source of failure, while some faulty pixels may be diagnosed as normal. Therefore, it is also necessary to consider the T 2 Contribution index and SPE contribution index determine the T of the neighboring area of ​​each pixel of the thermal image data to be tested. 2 The contribution score of T 2The contribution fraction of the chemical process and the contribution fraction of the SPE are used to determine the abnormal area. This can accurately detect the location of heat leaks in chemical processes.

[0054] In one embodiment of the present invention, each thermal imaging image sample is segmented into regions to obtain segmented thermal imaging image samples, including:

[0055] Grayscale each thermal imaging sample to obtain a grayscale image of each thermal imaging sample;

[0056] The grayscale image of each thermal imaging sample is segmented into regions to obtain a segmented thermal imaging sample.

[0057] like Figure 3 As shown, in this embodiment, for the thermal image dataset {G t}, where G t is a sample of thermal images at time t=1,2,…,n. The first step is to convert the 3-channel RGB image into a 1-channel grayscale image using the following formula:

[0058] Gray=0.3*R+0.59*G+0.11*B

[0059] Here Gray, R, G, and B are the values ​​in grayscale images and RGB images respectively. t}∈R n×m×l , where m and l are the number of rows and columns in each grayscale image.

[0060] Dataset {G t}First, it must be converted into a two-dimensional matrix, that is, each grayscale image data should be converted into a vector. It should be noted that thermal distribution is closely related to spatial distribution, that is, pixels with closer spatial distances have closer thermal relationships. However, during traditional column or row expansion operations, when a two-dimensional matrix grayscale image is converted into a one-dimensional vector, spatial information is usually lost. Therefore, before converting to a one-dimensional vector, the spatial position information of the original pixel points needs to be retained. The present invention uses the K-means algorithm to perform preliminary clustering of image pixels according to pixel values, and uses the canny edge detection algorithm to extract edge information, and finally obtains image segmentation results (segmented thermal imaging sample) representing different spatial regions.

[0061] In one embodiment of the present invention, the spatial distance between every two pixels is determined by the following formula:

[0062]

[0063] In the formula, σ is a weight factor with a positive value, η is the regional characteristic influencing factor, t represents the time, G t (i1, j1) and Gt (i2, j2) represents different pixel points, i represents the row, and j represents the column.

[0064] like Figure 3 As shown, in this embodiment, when G t (i1,j1) and G t When (i2, j2) belong to the same region, η=1; when they do not belong to the same region, η takes a larger value, such as 100.

[0065] In this embodiment, the data set {G t}First, it must be converted into a two-dimensional matrix, that is, each grayscale image data G t should be converted into a vector as follows:

[0066] X(t)=[G t (1,1),…,G t (1,l),…,G t (m,1),…,G t (m,l)]

[0067] At the same time, the distance matrix is ​​calculated as follows

[0068]

[0069] In this embodiment, the covariance matrix is ​​determined by the following formula:

[0070]

[0071] Where S′ is the covariance matrix, X is a two-dimensional matrix, and D is the spatial distance matrix.

[0072] Then, by performing SVD decomposition on S′, we can obtain

[0073] S′=P T ΛP

[0074] Where P is a unit orthogonal matrix; Λ is a diagonal matrix, and the elements λ on the diagonal matrix are i Arrange from large to small λ1≥λ2≥…λ s ≥ 0. In order to achieve data dimension reduction of X(t), the present invention extracts the main components in X(t) according to the cumulative percentage variance method (CPV):

[0075]

[0076] In the formula, ε is a parameter, which is usually set to 85%. Through the above formula, the present invention can determine the number of principal components k, and extract the first k columns in P as the load matrix.

[0077] In one embodiment of the present invention, the T of each pixel of the thermal image data to be tested is determined based on the feature matrix and the thermal image data to be tested. 2 Contribution indicators and SPE contribution indicators, including:

[0078] The thermal image data to be tested is converted to obtain a one-dimensional vector;

[0079] Based on the one-dimensional vector and the characteristic matrix, determine the T of each pixel of the thermal image data to be tested. 2 Contribution index and SPE contribution index.

[0080] In this embodiment, a one-dimensional vector is represented by x(t).

[0081] In one embodiment of the present invention, T 2 The contribution index and SPE contribution index are determined by the following formula:

[0082]

[0083] In the formula, ConT j 2 For the T 2 Contribution index, ConSPE j is the SPE contribution index, P is the feature matrix, P i is the i-th column of P, P j,i YesP i The jth row of is x and The jth column of , k is the number of principal components.

[0084] once or Con SPE j If it exceeds the threshold, it means that there may be a problem with the corresponding variable (pixel position). Although the above method can select pixels that contribute most to the fault. However, a high contribution rate does not mean a high probability of fault, so some normal pixels may be misdiagnosed as fault sources, while some faulty pixels may be diagnosed as normal. When a thermal fault occurs, all pixels in a specific area are abnormal. Based on the above principles, the present invention proposes the following fault location strategy.

[0085] Transfer the contribution score vector to a matrix according to the spatial position of each pixel:

[0086]

[0087] SCon SPE i,j =Con SPE i*l+j

[0088] Use the convolution window function to update the contribution score:

[0089]

[0090] Where β is the size of the convolution window function. Therefore, fault location is based on pixels in a specific area rather than a single pixel, avoiding the problem of incorrect fault location.

[0091] In one embodiment of the present invention, based on T 2 The contribution scores of and SPE are used to determine abnormal areas, including:

[0092] When the pixel T 2 When the contribution score of the pixel is greater than the first preset threshold or the contribution score of the SPE of the pixel is greater than the second preset threshold, the pixel is determined to be an abnormal pixel;

[0093] Count all abnormal pixels in the thermal image data to be tested to determine the abnormal area.

[0094] In this embodiment, when the pixel T 2 When the contribution score of the pixel is greater than the first preset threshold or the contribution score of the SPE of the pixel is greater than the second preset threshold, the pixel is determined to be an abnormal pixel, and all abnormal pixels in the thermal image data to be tested are counted as abnormal areas. Those skilled in the art can customize the first preset threshold and the second preset threshold according to actual usage.

[0095] like Figure 4 As shown, in this embodiment, an experimental tank system is used to test the proposed method, which consists of 2 main tanks. The water inlet of the system is located in the right water tank, and the water outlet is located in the left water tank. The two water tanks are connected by a fully insulated pipe, and water will flow from the right water tank to the left water tank. Both water tanks are equipped with bulkheads, which divide the water tanks into two chambers, and the two chambers are connected by a bottom channel. The following equipment is used in this experiment: a thermal imager for acquiring thermal images (Guide sensmartMobIR Air, 25Hz, 0.06℃ accuracy); a heating module (CHB401, <0.5% full scale); a temperature detection module, including 6 temperature sensors (SXPT100, accuracy 0.15℃).

[0096] The normal working status of the tank system is as follows:

[0097] 1) The heating module is installed near the water inlet. (a) Once the temperature reaches the set upper limit (46°C), the heater will be turned off immediately; (b) until the detection value drops to the lower limit (44°C), the heater will be turned on again. In this way, the temperature of the entire system will be limited to a certain range. And the temperature decreases from the right to the left.

[0098] 2) The water level in all chambers is maintained at a certain height, for example 80%.

[0099] We collected 2000 normal data samples for model training. The sampling period is 1 second. In addition, 2000 data samples were generated for testing. Each data set contains 2000 samples, and the failure occurs at the 501st sample point. There are three types of failures:

[0100] 1) Insulation layer peeling simulates heat leakage fault;

[0101] 2) Add a code water inlet to the right chamber of the right box to simulate abnormal stable working conditions;

[0102] 3) Add an additional code water inlet in the left cavity of the left water tank to simulate the situation where the conditions of one cavity change while the conditions of the other three cavities remain unchanged.

[0103] For comparison, six temperature sensors were installed in the experimental tank system, and their positions are as follows: Figure 4 As shown. The data obtained by the temperature sensor is monitored using traditional PCA. Applicable to temperature sensors and infrared thermal imaging technology.

[0104] like Figure 5 , 6 As shown in FIG. 7 , in this embodiment, the fault detection results of the method proposed by the present invention and the traditional PCA method are shown. Figure 5 As shown in Figure 1, PCA cannot detect Fault 1 because all temperature sensors are only installed inside the experimental tank system, so PCA cannot detect temperature anomalies on the surface of the equipment. Unlike traditional temperature sensors, thermal images can successfully capture temperature changes on the surface of an object, thereby detecting faults earlier. Figure 6-7 From the faults 2-3 in the example, it can be seen that SIPCA (the method of the present invention) detects these faults earlier than PCA: SIPCA detects fault 2 in the 517th sample, and PCA detects it in the 528th sample; SIPCA detects fault 3 in the 510th sample, and PCA detects it in the 558th sample. The reason for this phenomenon is that PCA is based on the data of the temperature sensor, and when the sensor is far away from the fault point, the detection delay is large; unlike traditional temperature sensors, infrared thermal imagers monitor the temperature of the process at pixel-level granularity, so this kind of fault can be discovered earlier. In addition, SIPCA uses spatial information for model description, so it is sensitive to temperature changes between adjacent spaces, so it can detect faults earlier.

[0105] like Figure 8 As shown, in this embodiment, the technical solution of the present invention is applied to the fault location result of fault 1. Based on the fault location strategy of spatial information, the technical solution of the present invention can successfully locate the fault source position (red part) in the image and filter out false alarm pixels.

[0106] According to another aspect of the embodiment, the present invention provides a thermal leak detection device for a chemical process. Figure 2 A schematic block diagram of a heat leak detection device for a chemical process according to an embodiment is shown. It is understood that the device can be implemented by any device, equipment, platform and equipment cluster with computing and processing capabilities. Figure 2 As shown, the device includes: an acquisition unit 200, a first data processing unit 202, a second data processing unit 204, a third data processing unit 206, a fourth data processing unit 208, a fifth data processing unit 210, a sixth data processing unit 212, a seventh data processing unit 214, an eighth data processing unit 216, and a ninth data processing unit 218. The main functions of each component unit are as follows:

[0107] The acquisition unit 200 is configured to acquire a thermal image data set and thermal image data to be tested; wherein the thermal image data set includes a plurality of thermal imaging image samples at different times;

[0108] The first data processing unit 202 is configured to perform region segmentation on each of the thermal imaging image samples to obtain segmented thermal imaging image samples;

[0109] A second data processing unit 204 is configured to determine a regional feature influencing factor based on the segmented thermal imaging image sample;

[0110] The third data processing unit 206 is configured to determine the spatial distance between every two pixels in the segmented thermal imaging image sample based on the regional feature influencing factor;

[0111] The fourth data processing unit 208 is configured to convert the thermal image data set to obtain a two-dimensional matrix of the thermal image data set, and determine a spatial distance matrix of the thermal image data set according to an arrangement relationship of the two-dimensional matrix and a spatial distance between each two pixel points;

[0112] A fifth data processing unit 210 is configured to determine a covariance matrix based on the spatial distance matrix and the two-dimensional matrix;

[0113] A sixth data processing unit 212 is configured to determine a feature matrix based on the covariance matrix;

[0114] The seventh data processing unit 214 is configured to determine the T of each pixel of the thermal image data to be measured based on the characteristic matrix and the thermal image data to be measured. 2 Contribution index and SPE contribution index;

[0115] The eighth data processing unit 216 is configured to process data according to the T 2 The contribution index and the SPE contribution index determine the T of each pixel of the thermal image data to be tested. 2 The contribution score of and the contribution score of SPE;

[0116] The ninth data processing unit 218 is configured to 2 The contribution score of and the contribution score of the SPE are used to determine the abnormal area.

[0117] According to another embodiment, there is also provided a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute a combination of Figure 1 The method described.

[0118] According to another embodiment of the present invention, there is also provided an electronic device, comprising a memory and a processor, wherein the memory stores an executable code, and when the processor executes the executable code, the Figure 1 The method described.

[0119] The various embodiments of the present invention are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0120] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the present invention can be implemented by hardware, software, firmware or any combination thereof. When implemented by software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium.

[0121] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for monitoring heat leaks in a chemical process, characterized in that: The method comprises: Acquire a thermal image data set and thermal image data to be tested; wherein the thermal image data set includes thermal imaging samples at multiple different times; Performing region segmentation on each of the thermal imaging image samples to obtain segmented thermal imaging image samples; Determining regional feature influencing factors based on the segmented thermal imaging image samples; Based on the regional feature influencing factor, determining the spatial distance between every two pixel points in the segmented thermal imaging image sample; The thermal image data set is converted to obtain a two-dimensional matrix of the thermal image data set, and a spatial distance matrix of the thermal image data set is determined according to an arrangement relationship of the two-dimensional matrix and a spatial distance between each two pixel points; Determining a covariance matrix based on the spatial distance matrix and the two-dimensional matrix; Based on the covariance matrix, determining a feature matrix; Based on the characteristic matrix and the thermal image data to be measured, determine the T of each pixel of the thermal image data to be measured. 2 Contribution index and SPE contribution index; According to the T 2 The contribution index and the SPE contribution index determine the T of the neighboring area of ​​each pixel of the thermal image data to be tested. 2 The contribution score of and the contribution score of SPE; Based on the T 2 The contribution score of and the contribution score of the SPE are used to determine the abnormal area.

2. The method according to claim 1, characterized in that The spatial distance between every two pixels is determined by the following formula: In the formula, σ is a weight factor with a positive value, η is the regional characteristic influencing factor, t represents the time, G t (i1, j1) and G t (i2, j2) represents different pixel points, i represents the row, and j represents the column.

3. The method according to claim 2, characterized in that The covariance matrix is ​​determined by the following formula: Wherein, S′ is the covariance matrix, X is the two-dimensional matrix, and D is the spatial distance matrix.

4. The method according to claim 3, characterized in that: The step of determining the T of each pixel of the thermal image data to be measured based on the feature matrix and the thermal image data to be measured 2 Contribution indicators and SPE contribution indicators, including: Converting the thermal image data to be measured to obtain a one-dimensional vector; Based on the one-dimensional vector and the characteristic matrix, determine the T of each pixel of the thermal image data to be tested. 2 Contribution index and SPE contribution index.

5. The method according to claim 4, characterized in that The T 2 The contribution index and SPE contribution index are determined by the following formula: Where ConT j 2 For the T 2 Contribution index, ConSPE j is the SPE contribution index, P is the feature matrix, P i is the i-th column of P, P j,i YesP i The jth row of is x and The jth column of , k is the number of principal components.

6. The method according to claim 5, characterized in that The step of performing region segmentation on each of the thermal imaging image samples to obtain segmented thermal imaging image samples includes: Performing grayscale processing on each of the thermal imaging image samples to obtain a grayscale image of each thermal imaging image sample; The grayscale image of each thermal imaging image sample is segmented into regions to obtain segmented thermal imaging image samples.

7. The method according to claim 6, characterized in that Based on the T 2 The contribution score of and the contribution score of the SPE are used to determine the abnormal area, including: When the pixel T 2 When the contribution score of the pixel point is greater than a first preset threshold or the contribution score of the SPE of the pixel point is greater than a second preset threshold, the pixel point is determined to be an abnormal pixel point; All abnormal pixel points in the thermal image data to be tested are counted to determine the abnormal area.

8. A heat leak detection device for a chemical process, characterized in that: include: An acquisition unit is configured to acquire a thermal image data set and thermal image data to be tested; wherein the thermal image data set includes a plurality of thermal imaging image samples at different times; A first data processing unit is configured to perform region segmentation on each of the thermal imaging image samples to obtain segmented thermal imaging image samples; A second data processing unit is configured to determine a regional feature influencing factor based on the segmented thermal imaging image sample; A third data processing unit is configured to determine the spatial distance between every two pixels in the segmented thermal imaging image sample based on the regional feature influencing factor; a fourth data processing unit configured to convert the thermal image data set to obtain a two-dimensional matrix of the thermal image data set, and determine a spatial distance matrix of the thermal image data set according to an arrangement relationship of the two-dimensional matrix and a spatial distance between each two pixel points; a fifth data processing unit, configured to determine a covariance matrix based on the spatial distance matrix and the two-dimensional matrix; a sixth data processing unit, configured to determine a feature matrix based on the covariance matrix; The seventh data processing unit is configured to determine the T of each pixel of the thermal image data to be tested based on the feature matrix and the thermal image data to be tested. 2 Contribution index and SPE contribution index; An eighth data processing unit is configured to process data according to the T 2 The contribution index and the SPE contribution index determine the T of the neighboring area of ​​each pixel of the thermal image data to be tested. 2 The contribution score of and the contribution score of SPE; A ninth data processing unit is configured to 2 The contribution score of and the contribution score of the SPE are used to determine the abnormal area.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 7.

Images