Valve cooling system cooling water temperature estimation method and system based on visual fusion

Abstract

The invention discloses a valve cooling system cooling water temperature estimation method and system based on visual fusion, and relates to the technical field of machine vision, and the method comprises the steps: carrying out the training of an RGB image of a valve cooling system, carrying out the image preprocessing of an input end, carrying out the feature fusion of the preprocessed image, and correspondingly associating a color value corresponding to the heat radiation intensity on the infrared image with a target detected by the RGB image to obtain the heat radiation intensity of pipelines at different positions of the valve cooling system, and performing linear calculation according to the cooling water temperature of the inlet valve and the outlet valve and the heat radiation intensity of the pipelines at different positions. And estimating cooling water temperatures of pipelines at different positions of the valve cooling system according to a linear calculation result, judging whether the estimated temperatures are correct or not according to an estimation result, and matching a strategy according to a judgment result. By combining the visual fusion technology and thermal radiation data analysis, accurate estimation of the cooling water temperature of the valve cooling system is achieved, the operation efficiency of the valve cooling system is improved, potential safety risks are prevented in advance, and long-term stable operation of the system is ensured.

Description

Technical Field

[0001] The present invention relates to the field of machine vision technology, and in particular to a method and system for estimating cooling water temperature of a valve cooling system based on vision fusion. Background Art

[0002] Valve cooling system temperature monitoring technology is a vital part of the power industry, which helps ensure the safe operation and effectiveness of converter valve equipment. In the converter station, the primary equipment valve group generates heat, which in turn causes abnormal temperature of the valve outlet pipeline of the valve cooling system. In order to monitor the temperature of the valve cooling equipment in the converter station, various sensors and monitoring systems are usually used. These sensors can measure parameters such as equipment surface temperature, internal temperature, and efficiency of the cooling system. Some of the commonly used sensors include thermocouples, infrared sensors, fiber optic sensors, etc. These sensors can monitor the temperature of the equipment in real time and transmit the data to the monitoring system for processing and analysis.

[0003] Traditional temperature monitoring technology has some limitations in some aspects. When there is a need to increase the monitoring coverage area of ​​the valve cooling system, more temperature measurement points need to be configured, which often means customized changes in the structure. The manpower, material resources and time costs of adding sensors and shutting down for construction are difficult to achieve for operating converter stations. In order to overcome these problems, simulation modeling is generally used in converter stations to build a high-fidelity model of the valve cooling system, and the cooling water temperature at any position of the valve cooling system is characterized by virtual sensors. However, since this method has no actual operating data support and is only obtained through theoretical calculations, there is a certain deviation between the data of the simulation model and the actual data of the valve cooling system operation.

[0004] With the development of technology, modern monitoring systems usually have intelligent functions, which can analyze data in real time, predict equipment failures and provide alarms and suggestions. In addition, some advanced technologies such as machine learning and artificial intelligence have also been applied to valve cooling system temperature monitoring to improve the accuracy and efficiency of monitoring. Among them, the introduction of machine vision temperature monitoring and analysis has attracted everyone's attention due to its advantages such as non-contact monitoring, high degree of intelligence, good real-time monitoring and analysis, and easy expansion. Summary of the invention

[0005] In view of the above-mentioned existing problems, the present invention provides a cooling water temperature estimation method and system for a valve cooling system based on visual fusion, so as to solve the problems in the prior art of insufficient flexibility in background processing, difficulty in achieving target detection, limitation of the robustness and adaptability of the model, and complexity of the system and difficulty in expansion.

[0006] In order to solve the above technical problems, a cooling water temperature estimation method for valve cooling system based on visual fusion is proposed, including:

[0007] The RGB images of the valve cooling system are trained, and image preprocessing is performed on the input end, and the preprocessed images are feature fused; the color values ​​corresponding to the thermal radiation intensity on the infrared image are associated with the targets detected by the RGB image, and the thermal radiation intensity of the pipelines at different positions of the valve cooling system is obtained. Linear calculation is performed based on the cooling water temperatures of the inlet and outlet valves and the thermal radiation intensity of the pipelines at different positions; the cooling water temperatures of the pipelines at different positions of the valve cooling system are estimated based on the linear calculation results, and whether the estimated temperature is correct is determined based on the estimation results, and the matching strategy is based on the judgment results.

[0008] As a preferred solution of the method for estimating the cooling water temperature of a valve cooling system based on visual fusion described in the present invention, the image preprocessing includes training the RGB image of the valve cooling system using the YOLOv5 deep learning algorithm, and performing Mosaic data enhancement and Rectangular black edge filling at the input end;

[0009] YOLOv5 deep learning algorithm training includes collecting and labeling RGB image datasets related to the valve cooling system, defining the location and category of each pipeline, configuring the learning rate, batch size and number of iterations for model training, extracting features from the YOLOv5 model using the labeled image dataset, and optimizing the target detection capability through multiple convolutional layers and activation functions. After training, the performance of the model is tested using an independent validation dataset, and adjusted and optimized based on the detection accuracy of the algorithm. The YOLOv5 network consists of an input end, a backbone network, a neck network and a detection layer.

[0010] As a preferred solution of the method for estimating the cooling water temperature of a valve cooling system based on visual fusion described in the present invention, the feature fusion includes sending the image preprocessed at the input end of the YOLOv5 network to the backbone network for feature extraction, and then performing feature fusion through the neck network;

[0011] The feature fusion includes introducing a weighted bidirectional feature pyramid network to perform multi-scale feature fusion, and introducing an SKAttention mechanism to perform convolution operations to obtain different feature maps;

[0012] The weighted bidirectional feature pyramid network includes initializing a weighted bidirectional feature pyramid network structure, calculating learnable weights, and performing top-down and bottom-up multi-scale fusion; the top-down path fuses high-level semantic information into low-level features, the bottom-up path fuses low-level high-resolution features into high-level features, and repeatedly applies the multi-scale fusion process to complete multi-scale fusion;

[0013] The formula for calculating the learnable weight is expressed as:

[0014]

[0015] Among them, w i is the learnable weight, w j is the normalized sum of learnable weights, ε is a bias to avoid numerical instability, and i and j are variable indices;

[0016] The SKAttention mechanism includes a row segmentation operation, which uses different convolution kernels for convolution to obtain multiple feature maps, and uses the SE structure to process the features obtained by each convolution kernel, obtain the channel attention matrix, and add the processed feature maps.

[0017] As a preferred solution of the method for estimating the cooling water temperature of a valve cooling system based on visual fusion described in the present invention, the corresponding association includes: performing denoising and image correction on the infrared image, converting the pixel value of the infrared image into thermal radiation intensity, mapping the pipeline position coordinates in the RGB image to the infrared image, and extracting the color value of the thermal radiation intensity at the corresponding position;

[0018] The denoising includes removing noise in the image using a Gaussian filter; the image correction includes obtaining a corrected image using a flat field correction; and the converting of the pixel values ​​of the infrared image into thermal radiation intensity includes mapping the pixel values ​​of the corrected infrared image to thermal radiation intensity using a calibration curve.

[0019] As a preferred solution of the method for estimating the cooling water temperature of a valve cooling system based on visual fusion described in the present invention, the linear calculation includes calculating the thermal radiation intensity value of each pipeline position according to the mapping relationship, collecting the cooling water temperature data of the inlet and outlet valves, and establishing a linear model for linear calculation;

[0020] The heat formula of the valve cooling system pipeline is expressed as:

[0021]

[0022] Where Q is the heat generated by the pipeline, t p is the cooling water temperature in the pipeline, t k is the ambient air temperature, α n is the heat transfer coefficient of the inner surface of the pipe, α w is the heat transfer coefficient of the outer surface of the pipe, d n is the inner diameter of the pipe, d w is the outer diameter of the pipe;

[0023] The relationship between pipeline heat and infrared image color is:

[0024] Q = k Grey ×S+C grey

[0025] Among them, Q is the heat of the pipeline, S is the gray level of the image, and k Grey is the grayscale thermal radiation coefficient, C grey is the grayscale thermal radiation constant.

[0026] As a preferred solution of the method for estimating the cooling water temperature of the valve cooling system based on visual fusion described in the present invention, wherein: the estimating the cooling water temperature of the pipelines at different positions of the valve cooling system includes using the established linear model and combining statistical methods to estimate the model parameters and estimate the cooling water temperature of the pipelines at different positions of the valve cooling system;

[0027] The formula for calculating cooling water temperature is:

[0028]

[0029] Among them, T i is the cooling water temperature, T 进阀 is the cooling water temperature of the inlet valve pipeline, T 出阀 is the cooling water temperature of the valve outlet pipeline, S 进阀 is the gray level of the inlet valve pipeline in the infrared image, S 出阀 is the gray level of the valve outlet pipeline in the infrared image, S i is the grayscale level of the pipeline infrared image.

[0030] As a preferred solution of the method for estimating the cooling water temperature of a valve cooling system based on visual fusion described in the present invention, the matching strategy includes setting a threshold to determine whether the temperature is abnormal, comparing the estimated temperature with the actual measured temperature, dividing the abnormality level according to the difference between the estimated temperature and the measured temperature, and matching different strategies;

[0031] The abnormal level classification includes: when the difference is between 0 and 2°C, it means that the estimated data is within the normal error range, and the current equipment parameters and operating status continue to be maintained, and regular inspections are carried out; when the difference is between 2 and 5°C, it means that the estimated data is within the slight error range, and the cause of the error is analyzed, the effectiveness of the model and strategy is evaluated, the early warning system is started, and the operation and maintenance personnel are notified to check; when the difference is greater than 5°C, it means that a serious error has occurred in the estimation, the system operation is stopped immediately, and emergency repairs are carried out. According to the analysis results, the model parameters and strategies are iteratively optimized.

[0032] Another object of the present invention is to provide a valve cooling system cooling water temperature estimation system based on visual fusion. The present invention associates RGB images with infrared images to facilitate image observation and analysis. The system of the present invention combines the temperature monitoring and analysis technology of image fusion with the temperature physical measurement points originally configured in the valve cooling system to provide a more reliable and intelligent solution for the cooling water temperature analysis of the valve cooling system.

[0033] As a preferred solution of the valve cooling system cooling water temperature estimation system based on visual fusion described in the present invention, it is characterized by including an image preprocessing and feature fusion module, an image association module, a pipeline cooling water temperature estimation module and an anomaly detection and parameter optimization module.

[0034] The image preprocessing and feature fusion module is used to train RGB images using the YOLOv5 deep learning algorithm, and apply Mosaic data enhancement and Rectangular black edge filling technology to enhance the generalization ability of the model, and perform feature fusion on the preprocessed data.

[0035] The image association module is used to perform denoising and image correction on the infrared image, convert the pixel value of the infrared image into thermal radiation intensity, map the pipeline position coordinates in the RGB image to the infrared image, and extract the thermal radiation intensity color value of the corresponding position.

[0036] The pipeline cooling water temperature estimation module is used to establish a relationship model between pipeline heat and infrared image color, collect cooling water temperature data of the inlet and outlet valves, calculate the thermal radiation intensity value of each pipeline position through a linear model, and estimate the cooling water temperature.

[0037] The anomaly detection and parameter optimization module is used to set a threshold to determine whether the temperature is abnormal, compare the estimated temperature with the actual measured temperature, divide the anomaly level according to the difference, and match different strategies according to different anomaly levels.

[0038] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the method described in the method for estimating the cooling water temperature of a valve cooling system based on visual fusion are implemented.

[0039] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the method described in the method for estimating the cooling water temperature of a valve cooling system based on visual fusion are implemented.

[0040] Beneficial effects of the present invention: The present invention significantly improves the image quality by training and preprocessing the RGB image of the valve cooling system, lays an accurate data foundation for subsequent feature extraction and model training, helps to reduce interference factors in the model training process, improves the model's recognition accuracy of pipeline features, and ensures the accuracy of subsequent temperature estimation; through the feature fusion step, deeply explores the color and texture features of cooling water in the image, enhances the model's ability to understand complex scenes, and makes temperature estimation more accurate in a changing working environment; associates the thermal radiation intensity of the infrared image with the target detected by the RGB image, and accurately obtains the thermal radiation intensity of pipelines at different positions in the valve cooling system, It ensures the precise match between thermodynamic data and visual data, provides reliable thermal radiation data for temperature estimation, and further improves the accuracy of estimation. Through linear calculation steps, combined with the cooling water temperature at the inlet and outlet of the valve, the temperature estimation process is simplified, the calculation efficiency is improved, and the stability of the estimation results is guaranteed, which is convenient for real-time monitoring and early warning. By setting a threshold to determine whether the temperature is abnormal and matching the corresponding strategy according to the judgment result, real-time monitoring and intelligent response to the operating status of the valve cooling system are realized, which can timely detect and handle temperature abnormalities, prevent equipment damage, and ensure the safe operation of the system. At the same time, through the matching of different strategies, the operation and maintenance efficiency is improved and the maintenance cost is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work, among which:

[0042] Figure 1 An overall flow chart of a method for estimating cooling water temperature of a valve cooling system based on visual fusion provided in accordance with an embodiment of the present invention.

[0043] Figure 2 A schematic diagram of the bidirectional feature pyramid network BiFPN structure of a valve cooling system cooling water temperature estimation method based on visual fusion provided by an embodiment of the present invention.

[0044] Figure 3 A schematic diagram of the SKAttention mechanism structure of a method for estimating cooling water temperature in a valve cooling system based on visual fusion provided in accordance with an embodiment of the present invention.

[0045] Figure 4 A schematic diagram of the association of RGB image and infrared image data of a valve cooling system cooling water temperature estimation method based on visual fusion provided by an embodiment of the present invention.

[0046] Figure 5 A system solution flow chart of a valve cooling system cooling water temperature estimation system based on visual fusion provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0047] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0048] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is mutually exclusive with other embodiments, either individually or selectively.

[0050] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0051] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0052] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0053] Example 1, reference Figure 1-Figure 4 , which is the first embodiment of the present invention, provides a method for estimating cooling water temperature of a valve cooling system based on visual fusion, comprising:

[0054] S1: Train the RGB image of the valve cooling system, perform image preprocessing at the input end, and perform feature fusion on the preprocessed image.

[0055] The image preprocessing includes training the RGB image of the valve cooling system using the YOLOv5 deep learning algorithm, and performing Mosaic data enhancement and Rectangular black edge filling at the input end;

[0056] YOLOv5 deep learning algorithm training includes collecting and labeling RGB image datasets related to the valve cooling system, defining the location and category of each pipeline, configuring the learning rate, batch size and number of iterations for model training, extracting features from the YOLOv5 model using the labeled image dataset, and optimizing the target detection capability through multiple convolutional layers and activation functions. After training, the performance of the model is tested using an independent validation dataset, and adjusted and optimized based on the detection accuracy of the algorithm. The YOLOv5 network consists of an input end, a backbone network, a neck network and a detection layer.

[0057] The Mosaic data enhancement includes merging four different images into a new image, specifically determining the center points of the four images, constructing a style of the combined image, calculating the coordinate points of the four images, and transforming the annotation box according to the synthesized image, and cropping the unannotated border part;

[0058] The Rectangular black padding method includes calculating the new size of the target image, adding black padding to the edge of the image, and retaining the center position of the target object.

[0059] It should be noted that the feature fusion includes sending the image preprocessed at the input end of the YOLOv5 network to the backbone network for feature extraction, and then performing feature fusion through the neck network;

[0060] The feature extraction includes extracting the color features of the cooling water from the preprocessed image, extracting the texture features of the cooling water using a gray level co-occurrence matrix, and calculating the area, perimeter and circularity of the cooling water region;

[0061] The feature fusion includes introducing a weighted bidirectional feature pyramid network to perform multi-scale feature fusion, and introducing the SKAttention mechanism to perform convolution operations to obtain different feature maps;

[0062] like Figure 2 The weighted bidirectional feature pyramid network includes initializing a weighted bidirectional feature pyramid network structure, calculating learnable weights, and performing top-down and bottom-up multi-scale fusion; the top-down path fuses high-level semantic information into low-level features, the bottom-up path fuses low-level high-resolution features into high-level features, and repeatedly applies the multi-scale fusion process to complete multi-scale fusion;

[0063] The formula for calculating the learnable weight is expressed as:

[0064]

[0065] Among them, w i is the learnable weight, w j is the normalized sum of learnable weights, ε is a bias to avoid numerical instability, and i and j are variable indices;

[0066] like Figure 3 The SKAttention mechanism includes a row segmentation operation, using different convolution kernels for convolution to obtain multiple feature maps, and using the SE structure to process the features obtained by each convolution kernel, obtain the channel attention matrix, and add the processed feature maps.

[0067] S2: Correlate the color values ​​corresponding to the thermal radiation intensity on the infrared image with the targets detected by the RGB image, obtain the thermal radiation intensity of the pipelines at different positions of the valve cooling system, and perform linear calculation based on the cooling water temperature at the inlet and outlet valves and the thermal radiation intensity of the pipelines at different positions.

[0068] Further, such as Figure 4 , the corresponding association includes, performing denoising and image correction on the infrared image, converting the pixel value of the infrared image into thermal radiation intensity, mapping the pipeline position coordinates in the RGB image to the infrared image, and extracting the thermal radiation intensity color value of the corresponding position;

[0069] The denoising includes removing noise in the image using a Gaussian filter; the image correction includes obtaining a corrected image using a flat field correction; and the converting of the pixel values ​​of the infrared image into thermal radiation intensity includes mapping the pixel values ​​of the corrected infrared image to thermal radiation intensity using a calibration curve.

[0070] Furthermore, the linear calculation includes calculating the thermal radiation intensity value of each pipeline position according to the mapping relationship, collecting the cooling water temperature data of the inlet valve and the outlet valve, and establishing a linear model for linear calculation;

[0071] The heat formula of the valve cooling system pipeline is expressed as:

[0072]

[0073] Where Q is the heat generated by the pipeline, t p is the cooling water temperature in the pipeline, t k is the ambient air temperature, α n is the heat transfer coefficient of the inner surface of the pipe, α w is the heat transfer coefficient of the outer surface of the pipe, d n is the inner diameter of the pipe, d w is the outer diameter of the pipe;

[0074] The relationship between pipeline heat and infrared image color is:

[0075] Q = k Grey ×S+C grey

[0076] Among them, Q is the heat of the pipeline, S is the gray level of the image, and k Grey is the grayscale thermal radiation coefficient, C grey is the grayscale thermal radiation constant.

[0077] S3: Estimate the cooling water temperature of the pipelines at different positions of the valve cooling system according to the linear calculation results, determine whether the estimated temperature is correct according to the estimation results, and match the strategy according to the determination results.

[0078] Further, the estimating the cooling water temperature of the pipelines at different positions of the valve cooling system includes using the established linear model and estimating the model parameters in combination with statistical methods and estimating the cooling water temperature of the pipelines at different positions of the valve cooling system;

[0079] The formula for calculating cooling water temperature is:

[0080]

[0081] Among them, T i is the cooling water temperature, T 进阀 is the cooling water temperature of the inlet valve pipeline, T 出阀 is the cooling water temperature of the valve outlet pipeline, S进阀 is the gray level of the inlet valve pipeline in the infrared image, S 出阀 is the gray level of the valve outlet pipeline in the infrared image, S i is the grayscale level of the pipeline infrared image.

[0082] Furthermore, the matching strategy includes setting a threshold to determine whether the temperature is abnormal, comparing the estimated temperature with the actual measured temperature, dividing the abnormality level according to the difference between the estimated temperature and the measured temperature, and matching different strategies;

[0083] The abnormal level classification includes: when the difference is between 0 and 2°C, it means that the estimated data is within the normal error range, and the current equipment parameters and operating status continue to be maintained, and regular inspections are carried out; when the difference is between 2 and 5°C, it means that the estimated data is within the slight error range, and the cause of the error is analyzed, the effectiveness of the model and strategy is evaluated, the early warning system is started, and the operation and maintenance personnel are notified to check; when the difference is greater than 5°C, it means that a serious error has occurred in the estimation, the system operation is stopped immediately, and emergency repairs are carried out. According to the analysis results, the model parameters and strategies are iteratively optimized.

[0084] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0085] Example 2, reference Figure 5 , which is the second embodiment of the present invention, provides a valve cooling system cooling water temperature estimation system based on visual fusion, including an image preprocessing and feature fusion module 100, an image association module 200, a pipeline cooling water temperature estimation module 300 and an anomaly detection and parameter optimization module 400.

[0086] The image preprocessing and feature fusion module 100 is used to train RGB images using the YOLOv5 deep learning algorithm, and apply Mosaic data enhancement and Rectangular black edge filling technology to enhance the generalization ability of the model, and perform feature fusion on the preprocessed data.

[0087] The image association module 200 is used to perform denoising and image correction on the infrared image, convert the pixel value of the infrared image into thermal radiation intensity, map the pipeline position coordinates in the RGB image to the infrared image, and extract the thermal radiation intensity color value of the corresponding position.

[0088] The pipeline cooling water temperature estimation module 300 is used to establish a relationship model between pipeline heat and infrared image color, collect cooling water temperature data of the inlet and outlet valves, calculate the thermal radiation intensity value of each pipeline position through a linear model, and estimate the cooling water temperature.

[0089] The anomaly detection and parameter optimization module 400 is used to set a threshold to determine whether the temperature is abnormal, compare the estimated temperature with the actual measured temperature, classify the anomaly level according to the difference, and match different strategies according to different anomaly levels.

[0090] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0091] Embodiment 3, the third embodiment of the present invention, is different from the first two embodiments in that:

[0092] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

[0093] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0094] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0095] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

Claims

1. A method for estimating cooling water temperature of a valve cooling system based on visual fusion, characterized in that: include, The RGB images of the valve cooling system are trained, the image is preprocessed at the input, and the preprocessed images are feature fused; The color value corresponding to the thermal radiation intensity on the infrared image is associated with the target detected by the RGB image, and the thermal radiation intensity of the pipelines at different positions of the valve cooling system is obtained. A linear calculation is performed based on the cooling water temperature at the inlet and outlet valves and the thermal radiation intensity of the pipelines at different positions. The cooling water temperature of the pipelines at different positions of the valve cooling system is estimated according to the linear calculation results, whether the estimated temperature is correct is judged according to the estimation results, and the matching strategy is determined according to the judgment results.

2. The method for estimating cooling water temperature of a valve cooling system based on visual fusion according to claim 1, characterized in that: The image preprocessing includes training the RGB image of the valve cooling system using the YOLOv5 deep learning algorithm, and performing Mosaic data enhancement and Rectangular black edge filling at the input end; YOLOv5 deep learning algorithm training includes collecting and labeling RGB image datasets related to the valve cooling system, defining the location and category of each pipeline, configuring the learning rate, batch size and number of iterations for model training, extracting features from the YOLOv5 model using the labeled image dataset, and optimizing the target detection capability through multiple convolutional layers and activation functions. After training, the performance of the model is tested using an independent validation dataset, and adjusted and optimized based on the detection accuracy of the algorithm. The YOLOv5 network consists of an input end, a backbone network, a neck network and a detection layer.

3. The method for estimating cooling water temperature of a valve cooling system based on visual fusion according to claim 2, characterized in that: The feature fusion includes sending the image preprocessed at the input end of the YOLOv5 network to the backbone network for feature extraction, and then performing feature fusion through the neck network; The feature fusion includes introducing a weighted bidirectional feature pyramid network to perform multi-scale feature fusion, and introducing the SKAttention mechanism to perform convolution operations to obtain different feature maps; The weighted bidirectional feature pyramid network includes initializing a weighted bidirectional feature pyramid network structure, calculating learnable weights, and performing top-down and bottom-up multi-scale fusion; the top-down path fuses high-level semantic information into low-level features, the bottom-up path fuses low-level high-resolution features into high-level features, and repeatedly applies the multi-scale fusion process to complete multi-scale fusion; The SKAttention mechanism includes a row segmentation operation, which uses different convolution kernels for convolution to obtain multiple feature maps, and uses the SE structure to process the features obtained by each convolution kernel, obtain the channel attention matrix, and add the processed feature maps.

4. The method for estimating cooling water temperature of a valve cooling system based on visual fusion according to claim 3, characterized in that: The corresponding association includes de-noising and image correction of the infrared image, converting the pixel value of the infrared image into thermal radiation intensity, mapping the pipeline position coordinates in the RGB image to the infrared image, and extracting the thermal radiation intensity color value of the corresponding position; The denoising includes removing noise in the image using a Gaussian filter; the image correction includes obtaining a corrected image using a flat field correction; and the converting of the pixel values ​​of the infrared image into thermal radiation intensity includes mapping the pixel values ​​of the corrected infrared image to thermal radiation intensity using a calibration curve.

5. The method for estimating cooling water temperature of a valve cooling system based on visual fusion according to claim 4, characterized in that: The linear calculation includes calculating the thermal radiation intensity value of each pipeline position according to the mapping relationship, collecting the cooling water temperature data of the inlet valve and the outlet valve, and establishing a linear model for linear calculation; The heat formula of the valve cooling system pipeline is expressed as: Where Q is the heat generated by the pipeline, t p is the cooling water temperature in the pipeline, t k is the ambient air temperature, α n is the heat transfer coefficient of the inner surface of the pipe, α w is the heat transfer coefficient of the outer surface of the pipe, d n is the inner diameter of the pipe, d w is the outer diameter of the pipe; The relationship between pipeline heat and infrared image color is: Q=k Grey ×S+C grey Among them, Q is the heat of the pipeline, S is the gray level of the image, and k Grey is the grayscale thermal radiation coefficient, C grey is the grayscale thermal radiation constant.

6. The method for estimating cooling water temperature of a valve cooling system based on visual fusion according to claim 5, characterized in that: The estimating the cooling water temperature of the pipelines at different positions of the valve cooling system includes using the established linear model and combining statistical methods to estimate the model parameters and estimate the cooling water temperature of the pipelines at different positions of the valve cooling system; The formula for calculating cooling water temperature is: Among them, T i is the cooling water temperature, T 进阀 is the cooling water temperature of the inlet valve pipeline, T 出阀 is the cooling water temperature of the valve outlet pipeline, S 进阀 is the gray level of the inlet valve pipeline in the infrared image, S 出阀 is the gray level of the valve outlet pipeline in the infrared image, S i is the grayscale level of the pipeline infrared image.

7. The method for estimating cooling water temperature of a valve cooling system based on visual fusion according to claim 6, characterized in that: The matching strategy includes setting a threshold to determine whether the temperature is abnormal, comparing the estimated temperature with the actual measured temperature, dividing the abnormality level according to the difference between the estimated temperature and the measured temperature, and matching different strategies; The abnormal level classification includes: when the difference is between 0 and 2°C, it means that the estimated data is within the normal error range, and the current equipment parameters and operating status continue to be maintained, and regular inspections are carried out; when the difference is between 2 and 5°C, it means that the estimated data is within the slight error range, and the cause of the error is analyzed, the effectiveness of the model and strategy is evaluated, the early warning system is started, and the operation and maintenance personnel are notified to check; when the difference is greater than 5°C, it means that a serious error has occurred in the estimation, the system operation is stopped immediately, and emergency repairs are carried out. According to the analysis results, the model parameters and strategies are iteratively optimized.

8. A system using the method for estimating the cooling water temperature of a valve cooling system based on visual fusion as claimed in any one of claims 1 to 7, characterized in that: It includes image preprocessing and feature fusion module, image association module, pipeline cooling water temperature estimation module, and anomaly detection and parameter optimization module; The image preprocessing and feature fusion module is used to train RGB images using the YOLOv5 deep learning algorithm, and apply Mosaic data enhancement and Rectangular black edge filling technology to enhance the generalization ability of the model, and perform feature fusion on the preprocessed data; The image association module is used to perform denoising and image correction on the infrared image, convert the pixel value of the infrared image into thermal radiation intensity, map the pipeline position coordinates in the RGB image to the infrared image, and extract the thermal radiation intensity color value of the corresponding position; The pipeline cooling water temperature estimation module is used to establish a relationship model between pipeline heat and infrared image color, collect cooling water temperature data of the inlet and outlet valves, calculate the thermal radiation intensity value of each pipeline position through a linear model, and estimate the cooling water temperature; The anomaly detection and parameter optimization module is used to set a threshold to determine whether the temperature is abnormal, compare the estimated temperature with the actual measured temperature, divide the anomaly level according to the difference, and match different strategies according to different anomaly levels.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for estimating the cooling water temperature of a valve cooling system based on visual fusion according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for estimating the cooling water temperature of a valve cooling system based on visual fusion according to any one of claims 1 to 7 are implemented.

Images