Overhead heating power pipeline leakage monitoring method based on surface temperature high-resolution inversion

Abstract

The invention discloses an overhead heat pipeline leakage monitoring method based on surface temperature high-resolution inversion, and the method comprises the steps: collecting overhead heat pipeline data, and obtaining thermal infrared historical data; constructing an initial downscaling model, and training the initial downscaling model by using the thermal infrared historical data to obtain a downscaling model; obtaining an MODIS thermal infrared detection image, inputting the MODIS thermal infrared detection image into the downscaling model, and obtaining a thermal infrared image with a high spatial resolution; performing surface temperature inversion on the thermal infrared image with the high spatial resolution to obtain surface temperature; and performing anomaly detection on the surface temperature based on the overhead heat pipeline data, and completing overhead heat pipeline leakage monitoring based on a detection result. According to the invention, low-cost, high-efficiency and large-range monitoring of leakage of the overhead heat pipeline can be realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of thermal systems, and in particular relates to an overhead thermal pipeline leakage monitoring method based on high-resolution inversion of ground surface temperature. Background Art

[0002] As the infrastructure of the city, overhead thermal pipelines play an important role in ensuring heating for residents and industrial production. With the development of urbanization and industrialization, overhead thermal pipelines have become an important channel for energy transmission, and leakage incidents may lead to serious safety accidents and environmental problems. Therefore, an effective and economical method for monitoring leakage of overhead thermal pipelines is crucial to protecting the safety of people's lives and property and maintaining environmental sustainability. The existing methods for monitoring leakage of overhead thermal pipelines are mainly based on establishing ground monitoring points, with a limited monitoring range and high detection costs, and large-scale monitoring cannot be carried out. Remote sensing technology has the advantages of large-scale synchronous observation, low cost, and high accuracy, and has great potential in monitoring overhead thermal pipelines. Therefore, in response to the above needs, the present invention proposes a method for monitoring leakage of overhead thermal pipelines based on high-resolution inversion of surface temperature. Summary of the invention

[0003] In order to solve the above technical problems, the present invention proposes an overhead thermal pipeline leakage monitoring method based on high-resolution inversion of surface temperature, which can realize low-cost and high-precision monitoring of overhead thermal pipe leakage problems.

[0004] To achieve the above object, the present invention provides an overhead thermal pipeline leakage monitoring method based on high-resolution inversion of ground surface temperature, comprising:

[0005] Collect overhead thermal pipeline data and obtain thermal infrared historical data;

[0006] Constructing an initial downscaling model, and using the thermal infrared historical data to train the initial downscaling model to obtain a downscaling model;

[0007] Acquire a MODIS thermal infrared detection image, and input the MODIS thermal infrared detection image into the downscaling model to acquire a thermal infrared image with high spatial resolution;

[0008] Performing surface temperature inversion on the thermal infrared image with high spatial resolution to obtain the surface temperature;

[0009] Based on the overhead thermal pipeline data, the surface temperature is detected abnormally, and based on the detection results, the overhead thermal pipeline leakage monitoring is completed.

[0010] Optionally, the overhead thermal pipeline data includes: temperature data, height data, geographic coordinate data, and pipeline material data;

[0011] The thermal infrared historical data include: MODIS thermal infrared historical image data, Landsat thermal infrared historical image data and UAV thermal infrared historical image data.

[0012] Optionally, the initial downscaling model includes: a shallow feature extraction layer, a deep feature extraction layer and an image reconstruction layer;

[0013] The shallow feature extraction layer is used to extract all the ground features of the thermal infrared historical data;

[0014] The deep feature extraction layer is used to extract the thermal pipeline features of the thermal infrared historical data;

[0015] The image reconstruction layer is used to enhance the thermal pipeline features based on the overall terrain features.

[0016] Optionally, the shallow feature extraction layer consists of a convolutional layer;

[0017] The deep feature extraction layer is composed of a plurality of residual blocks, wherein the residual block includes a convolutional layer and a skip connection layer;

[0018] The image reconstruction layer is composed of a feature splicing layer, an image upsampling layer and a convolution layer in sequence.

[0019] Optionally, obtaining the ground surface temperature includes: using a single window algorithm to perform ground surface temperature inversion on the thermal infrared image with high spatial resolution to obtain the ground surface temperature.

[0020] Optionally, using a single window algorithm to perform surface temperature inversion on the high spatial resolution thermal infrared image includes:

[0021] The built-in parameters are obtained by calculating the on-board brightness temperature value, surface emissivity, atmospheric transmittance and atmospheric average temperature;

[0022] The surface temperature is inverted using the built-in parameters.

[0023] Optionally, the method for calculating the brightness temperature value on the star is:

[0024] L=MQ dn +A

[0025] T=K1 / ln(1+K2 / L)

[0026] Among them, T is the brightness temperature of the star, L is the thermal radiation of the star, and Q dn is the pixel grayscale value, M is the slope of the band response function, A is the intercept of the band response function, K1 and K2 are sensor-specific calibration constants;

[0027] The method for calculating the emissivity of the table is:

[0028] ε=M1(T,λ) / M2(T,λ)

[0029] Among them, ε is the surface emissivity, M1(T,λ) is the radiation emittance of the object at temperature T and wavelength λ; M2(T,λ) is the radiation emittance of the black body at the same temperature and wavelength;

[0030] The method for calculating the atmospheric transmittance is:

[0031] τ=m-nw

[0032] Among them, τ is the atmospheric transmittance, m and n are constants, and w is the atmospheric water vapor content;

[0033] Calculate the atmospheric average operating temperature Ta;

[0034] The method for calculating the built-in parameters is:

[0035] C=τε

[0036] D = (1-τ)[1+(1-ε])τ]

[0037] Among them, C and D are built-in parameters;

[0038] The method for calculating the surface temperature is:

[0039] Ts=[a(1-CD)+(b(1-CD)+C+D)T-DT a ] / C

[0040] Among them, T S is the surface temperature, a and b are constants built into the formula.

[0041] Optionally, based on the overhead thermal pipeline data, the ground surface temperature is detected abnormally, and based on the detection result, the overhead thermal pipeline leakage monitoring is completed, including:

[0042] The monitoring area is determined according to the overhead thermal pipeline data, and a buffer zone is established based on the monitoring area with the overhead thermal pipeline as the center. The remaining area is used as a non-detection zone, and the temperature difference between the surface temperature of each point in the buffer zone and the average temperature of the non-detection zone is calculated. If the temperature difference exceeds a threshold, the early warning system is triggered.

[0043] Compared with the prior art, the present invention has the following advantages and technical effects:

[0044] The present invention improves the spatial resolution of the MODIS image to be detected by constructing a thermal infrared image downscaling model, adopts a single-window algorithm to invert the surface temperature of the monitoring area, performs surface temperature anomaly detection, and realizes large-scale overhead thermal pipeline leakage monitoring. The present invention uses remote sensing technology to perform large-scale synchronous monitoring, which has the advantages of high efficiency, low cost, wide range, and short cycle. It can quickly determine the leakage location of the overhead thermal pipeline, which is of great significance for improving the operating efficiency of the heating system and reducing energy waste and environmental pollution. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0046] Figure 1 is a flow chart of an overhead thermal pipeline leakage monitoring method based on high-resolution inversion of ground surface temperature according to an embodiment of the present invention;

[0047] Figure 2 is a convolutional neural network structure diagram of an embodiment of the present invention;

[0048] Figure 3 4 is a schematic diagram of a downscaling model according to an embodiment of the present invention. DETAILED DESCRIPTION

[0049] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0050] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0051] The present invention proposes a method for monitoring leakage of overhead thermal pipelines based on high-resolution inversion of ground surface temperature. Figure 1 As shown, the specific steps include:

[0052] Collect overhead thermal pipeline data and obtain thermal infrared historical data;

[0053] Specifically, thermal infrared historical data include: MODIS thermal infrared historical image data, Landsat thermal infrared historical image data and UAV thermal infrared historical image data;

[0054] The MODIS thermal infrared historical images are the 31st and 32nd band data of the MOD021KM L1B dataset, with a spatial resolution of 1000 meters and a single satellite revisit period of less than 1 day;

[0055] Landsat thermal infrared historical images are thermal infrared sensor band data from Landsat 8 and Landsat 9, with a spatial resolution of 100 meters and a binary satellite revisit period of 8 days;

[0056] The drone thermal infrared image is collected by the WIRIS Pro thermal infrared camera. The size of a single image is 640*512 pixels, and the spatial resolution is 0.2 meters.

[0057] Construct an initial downscaling model and use thermal infrared historical data to train the initial downscaling model to obtain a downscaling model. The schematic diagram is as follows: Figure 3 As shown;

[0058] Specifically, the principle of the downscaling model is to establish a one-to-many mapping relationship between low-resolution images and high-resolution images pixel by pixel, such as Figure 3 As shown in the figure, under the same field size, the original low-resolution thermal infrared image is 5*5 pixels in size, and the high-resolution thermal infrared image is 20*20 pixels in size. In the downscaling model training stage, the model establishes a one-to-many mapping of pixels, that is, 1 original thermal infrared image pixel corresponds to 4*4 high-resolution thermal infrared image pixels. After the training, the mapping relationship is stored as a downscaling model. In the model prediction stage, the low-resolution thermal infrared image is input, and the high-resolution thermal infrared image is obtained through the pixel-to-pixel mapping relationship.

[0059] Specifically, the MODIS, Landsat, and UAV simultaneous thermal infrared historical images are made into sample labels of uniform range size;

[0060] Using convolutional neural network ( Figure 2 ) Build a downscaling model, input sample labels, and iterate the training of the downscaling model;

[0061] Recent thermal infrared images are input to train and maintain the downscaling model to maintain the accuracy of the downscaling model.

[0062] Obtain MODIS thermal infrared detection images, and input the MODIS thermal infrared detection images into the downscaling model to obtain thermal infrared images with high spatial resolution;

[0063] Specifically, a thermal infrared image with a high spatial resolution refers to a thermal infrared image with a spatial resolution of 0.2 meters obtained by downscaling a MODIS thermal infrared image with a spatial resolution of 1000 meters.

[0064] Perform surface temperature inversion on high spatial resolution thermal infrared images to obtain surface temperature;

[0065] Based on the overhead thermal pipeline data, the surface temperature is detected abnormally, and based on the detection results, the overhead thermal pipeline leakage monitoring is completed.

[0066] Furthermore, the overhead thermal pipeline data includes: temperature data, height data, geographic coordinate data, and pipeline material data. The data comes from ground surveys and is used to determine the size of the monitoring area, temperature inversion parameters, and temperature anomaly detection thresholds.

[0067] Specifically, the monitoring area includes a 5-kilometer radius centered on the overhead thermal pipeline (the pipeline location is determined by geographic coordinate data).

[0068] The temperature inversion parameters include the elevation parameters and coordinate parameters used in the satellite data preprocessing process, which are obtained through the height data and geographic coordinate data obtained from the field survey.

[0069] The temperature anomaly detection threshold is the average temperature of the non-detection area plus 5°C. The average temperature of the non-detection area is obtained by averaging the surface temperature obtained by inversion in the previous step.

[0070] Further, the initial downscaling model includes: a shallow feature extraction layer, a deep feature extraction layer and an image reconstruction layer;

[0071] The shallow feature extraction layer is used to extract all the ground features of thermal infrared historical data;

[0072] Specifically, the shallow feature extraction module consists of a 3*3 convolutional layer, which inputs a thermal infrared image (single channel) and obtains a multi-channel shallow feature map of the same size (width and height are the same as the input image, and the number of channels is 64), with the purpose of extracting all the ground features of the input image. This module has two outputs, one as the input of the deep feature extraction module, and the other as the input of the image reconstruction module.

[0073] The deep feature extraction layer is used to extract thermal pipeline features from thermal infrared historical data;

[0074] Specifically, the deep feature extraction module is an efficient residual group composed of multiple residual blocks connected in series. It inputs the shallow feature map obtained in the previous step to obtain the deep feature map, and its purpose is to extract the thermal pipeline features in the input image. This module has only one output, which is used as the input of the image reconstruction module.

[0075] The image reconstruction layer is used to enhance the thermal pipeline features based on all terrain features.

[0076] Specifically, the image reconstruction module consists of an upsampling layer and a convolution layer. It first concatenates the input shallow feature map with the deep feature map. The purpose is to use the shallow feature map to enhance the thermal pipeline feature information in the deep feature map to compensate for the feature information loss caused by the deep feature extraction process. Then the image is enlarged through the upsampling layer to increase the number of rows and columns of the image. Finally, the convolution layer establishes the mapping relationship between the low-resolution thermal infrared image and the high-resolution thermal infrared image to improve the image resolution and output a high-resolution thermal infrared image.

[0077] Furthermore, the shallow feature extraction layer consists of convolutional layers;

[0078] The deep feature extraction layer is composed of multiple residual blocks, where the residual block includes a convolutional layer and a skip connection layer;

[0079] The image reconstruction layer consists of feature concatenation, image upsampling and convolution layers in sequence.

[0080] Furthermore, obtaining the ground surface temperature includes: using a single window algorithm to perform ground surface temperature inversion on a thermal infrared image with a high spatial resolution to obtain the ground surface temperature.

[0081] Furthermore, the single window algorithm is used to invert the surface temperature of high spatial resolution thermal infrared images, including:

[0082] The built-in parameters are obtained by calculating the on-board brightness temperature value, surface emissivity, atmospheric transmittance and atmospheric average temperature;

[0083] Invert the surface temperature using built-in parameters.

[0084] Furthermore, the method for calculating the brightness temperature value on the star is:

[0085] L=MQ dn +A

[0086] T=K1 / ln(1+K2 / L)

[0087] Among them, T is the brightness temperature of the star, L is the thermal radiation of the star, and Q dn is the pixel grayscale value, M is the slope of the band response function, A is the intercept of the band response function, K1 and K2 are sensor-specific calibration constants;

[0088] The method for calculating the emissivity of the table is:

[0089] ε=M1(T,λ) / M2(T,λ)

[0090] Among them, ε is the surface emissivity, M1(T,λ) is the radiation emittance of the object at temperature T and wavelength λ; M2(T,λ) is the radiation emittance of the black body at the same temperature and wavelength;

[0091] The method for calculating atmospheric transmittance is:

[0092] τ=m-nw

[0093] Among them, τ is the atmospheric transmittance, m and n are constants, and w is the atmospheric water vapor content;

[0094] Calculate the average atmospheric temperature Ta;

[0095] Tropical average atmosphere Ta = 17.9769 + 0.91715T0

[0096] The calculation formula for mid-latitude summer is: Ta = 16.0110 + 0.92621T0;

[0097] The calculation formula for mid-latitude winter is: Ta = 19.2704 + 0.91118T0;

[0098] T0 is the near-surface temperature, obtained from field measurements or meteorological stations.

[0099] The method to calculate the built-in parameters is:

[0100] C=τε

[0101] D = (1-τ)[1+(1-ε])τ]

[0102] Among them, C and D are built-in parameters;

[0103] The method for calculating the surface temperature is:

[0104] Ts=[a(1-CD)+(b(1-CD)+C+D)T-DT a ] / C

[0105] Among them, T S is the surface temperature, a and b are constants built into the formula.

[0106] Furthermore, the abnormal surface temperature is detected, and based on the detection results, the overhead thermal pipeline leakage monitoring is completed, including:

[0107] A buffer zone is established with the overhead thermal pipeline as the center, and the rest of the area is used as a non-detection zone. The temperature difference between the surface temperature at each point in the buffer zone and the average temperature of the non-detection zone is calculated. If the temperature difference exceeds the threshold, the early warning system is triggered.

[0108] Specifically, the buffer zone radius is set at 100 meters, the non-detection area is the range between 100 meters and 5 kilometers around the overhead thermal pipeline, and the temperature difference threshold is that the average temperature of the monitoring area exceeds the average temperature of the non-detection area by 5°C.

[0109] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for monitoring leakage of overhead thermal pipelines based on high-resolution inversion of ground surface temperature, characterized in that: include: Collect overhead thermal pipeline data and obtain thermal infrared historical data; Constructing an initial downscaling model, and using the thermal infrared historical data to train the initial downscaling model to obtain a downscaling model; Acquire a MODIS thermal infrared detection image, and input the MODIS thermal infrared detection image into the downscaling model to acquire a thermal infrared image with high spatial resolution; Performing surface temperature inversion on the thermal infrared image with high spatial resolution to obtain the surface temperature; Based on the overhead thermal pipeline data, the surface temperature is detected for abnormality, and according to the detection results, the overhead thermal pipeline leakage monitoring is completed.

2. The method for monitoring leakage of overhead thermal pipelines based on high-resolution inversion of ground surface temperature according to claim 1 is characterized in that: The overhead thermal pipeline data includes: temperature data, height data, geographic coordinate data, and pipeline material data; The thermal infrared historical data include: MODIS thermal infrared historical image data, Landsat thermal infrared historical image data and UAV thermal infrared historical image data.

3. The method for monitoring leakage of overhead thermal pipelines based on high-resolution inversion of ground surface temperature according to claim 1 is characterized in that: The initial downscaling model includes: a shallow feature extraction layer, a deep feature extraction layer and an image reconstruction layer; The shallow feature extraction layer is used to extract all the ground features of the thermal infrared historical data; The deep feature extraction layer is used to extract the thermal pipeline features of the thermal infrared historical data; The image reconstruction layer is used to enhance the thermal pipeline features based on the overall terrain features.

4. The method for monitoring leakage of overhead thermal pipelines based on high-resolution inversion of ground surface temperature according to claim 3 is characterized in that: The shallow feature extraction layer is composed of a convolutional layer; The deep feature extraction layer is composed of a plurality of residual blocks, wherein the residual block includes a convolutional layer and a skip connection layer; The image reconstruction layer is composed of a feature splicing layer, an image upsampling layer and a convolution layer in sequence.

5. The method for monitoring leakage of overhead thermal pipelines based on high-resolution inversion of ground surface temperature according to claim 1, characterized in that: Acquiring the ground surface temperature includes: using a single window algorithm to perform ground surface temperature inversion on the thermal infrared image with high spatial resolution to acquire the ground surface temperature.

6. The method for monitoring leakage of overhead thermal pipelines based on high-resolution inversion of ground surface temperature according to claim 5, characterized in that: The single window algorithm is used to invert the surface temperature of the thermal infrared image with high spatial resolution, including: The built-in parameters are obtained by calculating the on-board brightness temperature value, surface emissivity, atmospheric transmittance and atmospheric average temperature; The surface temperature is inverted using the built-in parameters.

7. The method for monitoring leakage of overhead thermal pipelines based on high-resolution inversion of ground surface temperature according to claim 6, characterized in that: The method for calculating the brightness temperature value on the star is: L=MQ dn +A T=K1 / ln(1+K2 / L) Among them, T is the brightness temperature of the star, L is the thermal radiation of the star, and Q dn is the pixel grayscale value, M is the slope of the band response function, A is the intercept of the band response function, K1 and K2 are sensor-specific calibration constants; The method for calculating the emissivity of the table is: ε=M1(T,λ) / M2(T,λ) Among them, ε is the surface emissivity, M1(T,λ) is the radiation emittance of the object at temperature T and wavelength λ; M2(T,λ) is the radiation emittance of the black body at the same temperature and wavelength; The method for calculating the atmospheric transmittance is: τ=m-nw Among them, τ is the atmospheric transmittance, m and n are constants, and w is the atmospheric water vapor content; The method for calculating the built-in parameters is: C=τε D = (1-τ)[1+(1-ε])τ] Among them, C and D are built-in parameters; The method for calculating the surface temperature is: Ts=[a(1-C-D)+(b(1-C-D)+C+D)T-DT a ] / C Among them, T S is the surface temperature, a and b are constants built into the formula.

8. The method for monitoring leakage of overhead thermal pipelines based on high-resolution inversion of ground surface temperature according to claim 1, characterized in that: Based on the overhead thermal pipeline data, the surface temperature is detected abnormally, and based on the detection result, the overhead thermal pipeline leakage monitoring is completed, including: The monitoring area is determined according to the overhead thermal pipeline data, and a buffer zone is established based on the monitoring area with the overhead thermal pipeline as the center. The remaining area is used as a non-detection zone, and the temperature difference between the surface temperature of each point in the buffer zone and the average temperature of the non-detection zone is calculated. If the temperature difference exceeds a threshold, the early warning system is triggered.

Images