Vision-based online monitoring and early warning method and system for galloping of power transmission line of power grid

Abstract

The invention discloses a power grid transmission line galloping online monitoring and early warning method and system based on vision, and belongs to the technical field of power grids, and the method comprises the steps: collecting and preprocessing real-time data of power grid transmission line galloping, and determining the characteristic data of the power grid transmission line galloping; establishing a power grid transmission line galloping on-line monitoring risk prediction model to analyze and predict the power grid transmission line galloping characteristic data, evaluating whether the power grid transmission line galloping is abnormal, and determining a power grid transmission line galloping on-line monitoring risk prediction result; and outputting a power grid transmission line galloping online monitoring risk prediction result, and triggering an early warning alarm when the power grid transmission line galloping is abnormal. According to the invention, the problems that the galloping of the power transmission line of the power grid cannot be monitored online in real time and the abnormal galloping behavior of the power transmission line of the power grid cannot be early warned in time in the prior art are solved. The system can perform real-time on-line monitoring on power grid transmission line galloping, can perform timely early warning management and control on abnormal behaviors of power grid transmission line galloping, and can fully guarantee safe and stable operation of a power grid.

Description

Technical Field

[0001] This invention relates to the field of power grid technology, specifically to a vision-based online monitoring and early warning method and system for power grid transmission line galloping. Background Technology

[0002] Power line galloping is a common problem in power system operation. Especially under extreme weather conditions, power line galloping can lead to mechanical fatigue of transmission lines, damage to insulators, or even line breakage, seriously threatening the safe and stable operation of the power grid.

[0003] Chinese patent CN107563551A discloses a method and system for identifying critical lines in power grid galloping. The method includes: calculating a predicted power grid galloping area; performing spatial overlay analysis between the transmission line and the predicted power grid galloping area to obtain predicted galloping transmission lines; generating several combinations of galloping lines through multiple random samplings of the predicted galloping lines, and calculating the tripping probability of each galloping line in the combination; calculating the load loss of the galloping line combination based on the tripping probability of each galloping line, and calculating the risk level of each galloping line based on the load loss of each galloping line combination to select the critical line in power grid galloping. However, this patent has the following drawbacks:

[0004] Existing technologies cannot perform real-time online monitoring of power grid transmission line galloping, nor can they provide timely early warnings of abnormal galloping behavior, thus failing to provide strong support for the intelligent operation and maintenance of power grid transmission lines and fully guarantee the safe and stable operation of the power grid. Summary of the Invention

[0005] The purpose of this invention is to provide a vision-based online monitoring and early warning method and system for power grid transmission line galloping, which can monitor power grid transmission line galloping in real time and provide timely early warning and control of abnormal galloping behavior, thus fully ensuring the safe and stable operation of the power grid and solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A vision-based online monitoring and early warning method for power transmission line galloping includes:

[0008] Real-time data on power grid transmission line galloping was collected and preprocessed to extract key features of power grid transmission line galloping and determine the characteristic data of power grid transmission line galloping.

[0009] A risk prediction model for online monitoring of power transmission line galloping was established to analyze and predict the characteristic data of power transmission line galloping, assess whether the power transmission line galloping is abnormal, and determine the risk prediction results of online monitoring of power transmission line galloping.

[0010] The system provides visualized output of online monitoring and risk prediction results for power grid transmission line galloping, and automatically triggers early warning alarms when power grid transmission line galloping is abnormal.

[0011] Preferably, real-time data on power grid transmission line galloping is collected, including:

[0012] Non-contact real-time monitoring of power grid transmission line galloping is carried out using high-definition cameras installed on power grid transmission towers, capturing the movement trajectory of power grid transmission lines and collecting video data of power grid transmission line galloping;

[0013] Real-time monitoring of wind speed, wind direction, temperature, and humidity parameters during power grid transmission line galloping is achieved using environmental sensors installed on power grid transmission towers, thereby collecting environmental data related to power grid transmission line galloping.

[0014] Based on video data of power transmission line galloping and environmental data of power transmission line galloping, real-time data of power transmission line galloping is determined.

[0015] Preferably, the real-time data on power grid transmission line galloping is preprocessed, including:

[0016] Clean the real-time data of power grid transmission line galloping to remove noise and identify missing and outlier values.

[0017] The missing and outlier values ​​in the real-time data of power grid transmission line galloping are evaluated to determine whether the missing and outlier values ​​in the real-time data of power grid transmission line galloping are valuable for online monitoring and early warning of power grid transmission line galloping.

[0018] If missing values ​​and outliers in the real-time data of power transmission line galloping are valuable for online monitoring and early warning of power transmission line galloping, then the mean is used to fill in the missing values ​​and the median is used to replace the outliers.

[0019] If missing or outlier values ​​in the real-time data of power transmission line galloping are of no value for online monitoring and early warning of power transmission line galloping, then the missing or outlier values ​​should be deleted.

[0020] The real-time data of power grid transmission line galloping is normalized to eliminate the dimensional differences in the real-time data of power grid transmission line galloping, thus forming standardized real-time data of power grid transmission line galloping.

[0021] Preferably, preprocessing of real-time data on power grid transmission line galloping also includes:

[0022] Contrast enhancement is performed on real-time data of power grid transmission line galloping to highlight the detailed features in the data.

[0023] Target detection is performed on real-time data of power grid transmission line galloping. The galloping behavior of power grid transmission lines is detected and identified in real time. The galloping behavior of power grid transmission lines is analyzed to extract key features of power grid transmission line galloping and determine the characteristic data of power grid transmission line galloping, including amplitude, frequency and phase.

[0024] Preferably, the real-time data on power grid transmission line galloping is enhanced in contrast, including:

[0025] Extract image data corresponding to real-time data on power grid transmission line galloping;

[0026] Each pair of adjacent image data acquisition times is treated as a single image data group.

[0027] Perform grayscale processing on the two image data contained in the image data group to obtain the grayscale processed image data;

[0028] The similarity between two grayscale processed image data contained in the image data group is compared to obtain the similarity value between the two grayscale processed image data.

[0029] The similarity value is compared with a preset similarity threshold;

[0030] When the similarity value is not lower than the preset similarity threshold, the real-time data of the power grid transmission line galloping is enhanced according to the preset contrast enhancement range, wherein the preset contrast enhancement range is 1%-6%.

[0031] When the similarity value is lower than the preset similarity threshold, the image data group with the similarity value lower than the preset similarity threshold is taken as the target image data group.

[0032] For a target image dataset, retrieve the similarity value between two grayscale processed image data points contained within it;

[0033] Based on the similarity value between the two grayscale processed image data contained in the target image data group, the grayscale values ​​of the two image data contained in the target image data group are adjusted respectively.

[0034] Preferably, adjusting the grayscale values ​​of two image data points contained in the target image data group based on the similarity value between the two grayscale processed image data points retrieved from the target image data group includes:

[0035] Retrieve the two grayscale processed image data contained in the target image data group;

[0036] For each grayscale processed image data, obtain its corresponding average grayscale value.

[0037] The difference between the average gray values ​​of the two gray-scale processed image data contained in the target image data group is processed, and the absolute value of the difference between the average gray values ​​of the two gray-scale processed image data contained in the target image data group is obtained.

[0038] The absolute value of the difference is compared with a preset difference threshold;

[0039] When the absolute value of the difference does not exceed the preset difference threshold, the first contrast enhancement model is used to perform contrast enhancement processing on the two image data contained in the target image data group.

[0040] When the absolute value of the difference exceeds the preset difference threshold, the second contrast enhancement model is used to perform contrast enhancement processing on the two image data contained in the target image data group.

[0041] Preferably, assessing whether power grid transmission line galloping is abnormal includes:

[0042] Based on the vision-based online monitoring and early warning requirements for power grid transmission line galloping, an online monitoring risk prediction model for power grid transmission line galloping is established, and the online monitoring risk prediction model for power grid transmission line galloping is deployed in a real online monitoring and early warning environment for power grid transmission lines galloping.

[0043] The characteristic data of power transmission line galloping are input into the online monitoring risk prediction model for power transmission line galloping. The model is then used to analyze and predict the characteristic data of power transmission line galloping to assess whether the galloping is abnormal, and finally determine the online monitoring risk prediction result for power transmission line galloping.

[0044] Preferably, an online monitoring risk prediction model for power grid transmission line galloping is established, and the following operations are performed:

[0045] Collect historical data on power grid transmission line galloping, and divide the collected historical data to determine the training set and test set;

[0046] Based on machine learning technology, a training set is used to train the machine learning model, enabling the machine learning model to autonomously learn the risk prediction behavior of online monitoring of power grid transmission line galloping from the training set and evaluate whether the power grid transmission line galloping is abnormal, thereby determining the online monitoring risk prediction model of power grid transmission line galloping based on machine learning.

[0047] The machine learning-based online monitoring risk prediction model for power grid transmission line galloping was tested using a test set to evaluate its performance and determine whether it can achieve the expected effect of predicting the risk of power grid transmission line galloping.

[0048] When the machine learning-based online monitoring risk prediction model for power grid transmission line galloping fails to achieve the expected prediction effect, the parameters of the machine learning-based online monitoring risk prediction model for power grid transmission line galloping are adjusted, and the model is continuously iterated and optimized until it can achieve the expected prediction effect, thereby determining the optimal online monitoring risk prediction model for power grid transmission line galloping.

[0049] Preferably, an early warning alarm is automatically triggered when abnormal power grid transmission line galloping occurs, including:

[0050] The system displays the online monitoring and risk prediction results of power grid transmission line galloping in a visual format to managers in real time. When abnormal behavior of power grid transmission line galloping is detected, an early warning alarm is automatically triggered, prompting managers to carry out timely maintenance and management of the power grid transmission lines.

[0051] According to another aspect of the present invention, a vision-based online monitoring and early warning system for power grid transmission line galloping is also provided, for implementing the vision-based online monitoring and early warning method for power grid transmission line galloping as described above, comprising:

[0052] The front-end acquisition module is used to collect the motion trajectory and environmental parameters of power grid transmission lines during galloping, and to determine the real-time data of power grid transmission line galloping.

[0053] The data processing module is used to preprocess real-time data on power grid transmission line galloping and determine the characteristic data of power grid transmission line galloping.

[0054] The risk assessment module is used to analyze and predict the characteristic data of power grid transmission line galloping, and to determine the risk prediction results of online monitoring of power grid transmission line galloping.

[0055] The early warning output module is used to visualize and output the risk prediction results of online monitoring of power grid transmission line galloping, and automatically triggers an early warning alarm when the power grid transmission line galloping is abnormal.

[0056] Compared with the prior art, the beneficial effects of the present invention are:

[0057] This invention collects the motion trajectory and environmental parameters of power transmission lines during galloping to determine real-time data. By preprocessing this data, key features of the galloping are extracted, and characteristic data are identified. An online monitoring risk prediction model is established to analyze and predict the characteristic data, assessing whether the galloping is abnormal and determining the online monitoring risk prediction results. These results are then visualized and output. Furthermore, an early warning alarm is automatically triggered when abnormal galloping occurs, prompting management personnel to perform timely maintenance and management of the power transmission lines. This invention enables real-time online monitoring of power transmission line galloping and timely early warning and control of abnormal galloping behavior, thus fully ensuring the safe and stable operation of the power grid. Attached Figure Description

[0058] Figure 1 This is a block diagram of the vision-based online monitoring and early warning system for power transmission line galloping according to the present invention;

[0059] Figure 2 This is a flowchart of the vision-based online monitoring and early warning method for power transmission line galloping according to the present invention. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] To address the current limitations of real-time online monitoring of power transmission line galloping and the inability to provide timely early warnings of abnormal galloping behavior, thus failing to adequately guarantee the safe and stable operation of the power grid, please refer to [link to relevant documentation]. Figures 1-2 This embodiment provides the following technical solution:

[0062] The vision-based online monitoring and early warning system for power grid transmission line galloping includes: a front-end acquisition module, a data processing module, a risk assessment module, and an early warning output module.

[0063] Specifically, through the interactive communication between the front-end acquisition module, data processing module, risk assessment module, and early warning output module, the galloping of power grid transmission lines can be monitored online in real time. When abnormal galloping of power grid transmission lines is detected online, an early warning alarm is automatically triggered, prompting managers to carry out timely maintenance and management of power grid transmission lines, which can fully ensure the safe and stable operation of the power grid.

[0064] The system comprises the following modules: a front-end acquisition module for collecting the motion trajectory and environmental parameters of power transmission lines during galloping, and a data processing module for preprocessing the real-time data to determine the characteristic data of power transmission line galloping; a risk assessment module for analyzing and predicting the characteristic data of power transmission line galloping, and a risk prediction result for online monitoring of power transmission line galloping; and an early warning output module for visually outputting the risk prediction result of online monitoring of power transmission line galloping, and automatically triggering an early warning alarm when power transmission line galloping is abnormal, which can fully ensure the safe and stable operation of the power grid.

[0065] To better illustrate the vision-based online monitoring and early warning process for power transmission line galloping, this embodiment provides a vision-based online monitoring and early warning method for power transmission line galloping, implemented based on the aforementioned vision-based online monitoring and early warning system for power transmission line galloping, including:

[0066] Real-time data on power grid transmission line galloping was collected and preprocessed to extract key features of power grid transmission line galloping and determine the characteristic data of power grid transmission line galloping.

[0067] In this embodiment, real-time data on power grid transmission line galloping is collected, including:

[0068] Non-contact real-time monitoring of power grid transmission line galloping is carried out using high-definition cameras installed on power grid transmission towers, capturing the movement trajectory of power grid transmission lines and collecting video data of power grid transmission line galloping;

[0069] Real-time monitoring of wind speed, wind direction, temperature, and humidity parameters during power grid transmission line galloping is achieved using environmental sensors installed on power grid transmission towers, thereby collecting environmental data related to power grid transmission line galloping.

[0070] Based on video data of power transmission line galloping and environmental data of power transmission line galloping, real-time data of power transmission line galloping is determined.

[0071] In this embodiment, the real-time data on power grid transmission line galloping is preprocessed, including:

[0072] Clean the real-time data of power grid transmission line galloping to remove noise and identify missing and outlier values.

[0073] The missing and outlier values ​​in the real-time data of power grid transmission line galloping are evaluated to determine whether the missing and outlier values ​​in the real-time data of power grid transmission line galloping are valuable for online monitoring and early warning of power grid transmission line galloping.

[0074] If missing values ​​and outliers in the real-time data of power transmission line galloping are valuable for online monitoring and early warning of power transmission line galloping, then the mean is used to fill in the missing values ​​and the median is used to replace the outliers.

[0075] If missing or outlier values ​​in the real-time data of power transmission line galloping are of no value for online monitoring and early warning of power transmission line galloping, then the missing or outlier values ​​should be deleted.

[0076] The real-time data of power grid transmission line galloping is normalized to eliminate the dimensional differences in the real-time data of power grid transmission line galloping, thus forming standardized real-time data of power grid transmission line galloping.

[0077] In this embodiment, preprocessing of real-time data on power grid transmission line galloping also includes:

[0078] Contrast enhancement is performed on real-time data of power grid transmission line galloping to highlight the detailed features in the data.

[0079] Target detection is performed on real-time data of power grid transmission line galloping. The galloping behavior of power grid transmission lines is detected and identified in real time. The galloping behavior of power grid transmission lines is analyzed to extract key features of power grid transmission line galloping and determine the characteristic data of power grid transmission line galloping, including amplitude, frequency and phase.

[0080] A risk prediction model for online monitoring of power transmission line galloping was established to analyze and predict the characteristic data of power transmission line galloping, assess whether the power transmission line galloping is abnormal, and determine the risk prediction results of online monitoring of power transmission line galloping.

[0081] Specifically, contrast enhancement is performed on real-time data of power grid transmission line galloping, including:

[0082] Extract image data corresponding to real-time data on power grid transmission line galloping;

[0083] Each pair of adjacent image data acquisition times is treated as a single image data group.

[0084] Perform grayscale processing on the two image data contained in the image data group to obtain the grayscale processed image data;

[0085] The similarity between two grayscale processed image data contained in the image data group is compared to obtain the similarity value between the two grayscale processed image data.

[0086] The similarity value is compared with a preset similarity threshold;

[0087] When the similarity value is not lower than the preset similarity threshold, the real-time data of the power grid transmission line galloping is enhanced according to the preset contrast enhancement range, wherein the preset contrast enhancement range is 1%-6%.

[0088] When the similarity value is lower than the preset similarity threshold, the image data group with the similarity value lower than the preset similarity threshold is taken as the target image data group.

[0089] For a target image dataset, retrieve the similarity value between two grayscale processed image data points contained within it;

[0090] Based on the similarity value between the two grayscale processed image data contained in the target image data group, the grayscale values ​​of the two image data contained in the target image data group are adjusted respectively.

[0091] The technical effect of the above solution is as follows: Image data corresponding to real-time data of power grid transmission line galloping is extracted. Image data corresponding to each two adjacent image data acquisition times are grouped into one image data set. The two image data sets in each set are processed into grayscale images. The similarity value between these two grayscale images is calculated. The calculated similarity value is compared with a preset similarity threshold. When the similarity value is not lower than the preset threshold, the two image data sets are considered to have high similarity. Due to environmental or equipment factors, the overall brightness or contrast change is not significant. In this case, the contrast of the real-time power grid transmission line galloping data is enhanced according to a preset contrast enhancement range (1%-6%) to improve image clarity and recognizability. When the similarity value is lower than the preset threshold, the two image data sets are considered to have significant differences, possibly due to large galloping amplitude or environmental changes. In this case, this image data set is designated as the target image data set. For the target image data set, the grayscale values ​​of the two grayscale processed image data sets within it are adjusted according to the similarity value between them.

[0092] By comparing similarity data, the technical solution can dynamically adapt to changes in power grid transmission line galloping. When the galloping amplitude is small, a preset contrast enhancement range is used to avoid over-enhancement that could lead to image distortion. When the galloping amplitude is large, grayscale adjustments are made to highlight differences, facilitating observation and analysis, and ensuring that the contrast enhancement processing matches the actual situation of the image data. Contrast enhancement improves image clarity and recognizability, making galloping details more apparent and helping maintenance personnel to more accurately judge the galloping situation. By setting a similarity threshold, the technical solution can filter out minor changes caused by environmental or equipment factors, reducing the possibility of misjudgment and missed judgment. Automated image data processing reduces manual intervention and improves maintenance efficiency. At the same time, by providing clear image data, it helps maintenance personnel make decisions and take measures more quickly. The technical solution analyzes and processes real-time data, providing data-driven decision support for maintenance personnel. Through contrast enhancement and grayscale adjustment, the image data becomes more intuitive and easier to understand, helping maintenance personnel to more accurately assess the galloping situation and formulate corresponding countermeasures.

[0093] Specifically, based on the similarity value between the two grayscale processed image data contained in the target image data group, the grayscale values ​​of the two image data contained in the target image data group are adjusted respectively, including:

[0094] Retrieve the two grayscale processed image data contained in the target image data group;

[0095] For each grayscale processed image data, obtain its corresponding average grayscale value.

[0096] The difference between the average gray values ​​of the two gray-scale processed image data contained in the target image data group is processed, and the absolute value of the difference between the average gray values ​​of the two gray-scale processed image data contained in the target image data group is obtained.

[0097] The absolute value of the difference is compared with a preset difference threshold;

[0098] When the absolute value of the difference does not exceed the preset difference threshold, the first contrast enhancement model is used to perform contrast enhancement processing on the two image data contained in the target image data group.

[0099] The structure of the first contrast enhancement model is as follows:

[0100]

[0101] Among them, D t01D0 represents the contrast value after contrast enhancement processing using the first contrast enhancement model; J represents the original contrast value of the image data before contrast enhancement processing; J represents the absolute value of the difference. y This represents the preset difference threshold; S represents the similarity value between the two grayscale processed image data contained in the target image data group; specifically, This part is a comprehensive adjustment factor. This represents the ratio of the absolute value of the difference to a preset threshold. Multiplying this by the similarity S ensures that the adjustment factor considers not only grayscale differences but also the similarity of image content. This improves the adaptability of contrast enhancement to the actual image data. This is an enhancement coefficient, which adds an adjustment factor to a base of 1. When the adjustment factor is positive, the enhancement coefficient is greater than 1, thus enhancing the original contrast D0. This model can adaptively adjust the degree of contrast enhancement based on the difference in average grayscale values ​​and similarity between images, avoiding image distortion caused by over-enhancement. Simultaneously, the contrast enhancement adjustment makes image details clearer, improving image recognizability. Because it considers image similarity, it can better preserve the original features of the image while enhancing contrast, without destroying important information in the image due to over-enhancement.

[0102] When the absolute value of the difference exceeds the preset difference threshold, the second contrast enhancement model is used to perform contrast enhancement processing on the two image data contained in the target image data group.

[0103] The structure of the second contrast enhancement model is as follows:

[0104]

[0105] Among them, D t02 D0 represents the contrast value after contrast enhancement processing using the second contrast enhancement model; J represents the original contrast value of the image data before contrast enhancement processing; J represents the absolute value of the difference. y S represents the preset difference threshold; S represents the similarity value between the two grayscale processed image data contained in the target image data group; S y This indicates the preset similarity threshold. This section reflects the degree of difference between the current similarity and the preset similarity threshold. This indicates the ratio of the preset difference threshold to the absolute value of the current difference; This means taking the minimum of the two values ​​mentioned above as the adjustment factor. This is done to comprehensively consider the effects of image similarity and the difference in average grayscale values, ensuring appropriate contrast enhancement when image differences are large, while avoiding over-enhancement. This is an enhancement coefficient, an adjustment factor added to the base value of 1, used to enhance the original contrast D0. This model comprehensively considers image similarity and the difference in average grayscale values, enabling a more comprehensive assessment of differences between images. When image similarity is low and the difference in average grayscale values ​​is large, appropriate contrast enhancement is applied, making the differences in the images more apparent and aiding in observation and analysis. By using a smaller value as the adjustment factor, excessive contrast enhancement is avoided when image differences are too large, thus maintaining a reasonable visual effect and preventing unnatural distortion. It enhances the contrast between different parts of the image, making details and features more prominent, improving the analytical capabilities for images such as power grid transmission line galloping, and helping maintenance personnel to more accurately determine the galloping situation.

[0106] The technical effects of the above solution are as follows: The solution adaptively selects a contrast enhancement model based on the similarity between images and the absolute value of the difference between their average grayscale values, thus obtaining clear images under different dancing amplitudes and environmental conditions. Contrast enhancement makes the dancing details in the image more apparent, improving image recognition. This helps maintenance personnel more accurately judge the dancing situation and promptly identify potential safety hazards. The first contrast enhancement model uses the above enhancement method when the grayscale difference is small, helping to reduce noise interference and avoid image distortion. The second contrast enhancement model uses the above enhancement method when the grayscale difference is large, highlighting the details in the image and making features such as dancing amplitude and direction clearer. By preset a difference threshold, the solution can automatically adapt to the image processing needs of different scenarios, improving the system's robustness and adaptability. Automated image data processing reduces manual intervention and improves maintenance efficiency. Simultaneously, by providing clear image data, it helps maintenance personnel make decisions and take measures more quickly.

[0107] In this embodiment, assessing whether power grid transmission line galloping is abnormal includes:

[0108] Based on the vision-based online monitoring and early warning requirements for power grid transmission line galloping, an online monitoring risk prediction model for power grid transmission line galloping is established, and the online monitoring risk prediction model for power grid transmission line galloping is deployed in a real online monitoring and early warning environment for power grid transmission lines galloping.

[0109] The characteristic data of power transmission line galloping are input into the online monitoring risk prediction model for power transmission line galloping. The model is then used to analyze and predict the characteristic data of power transmission line galloping to assess whether the galloping is abnormal, and finally determine the online monitoring risk prediction result for power transmission line galloping.

[0110] In this embodiment, an online monitoring risk prediction model for power grid transmission line galloping is established, and the following operations are performed:

[0111] Collect historical data on power grid transmission line galloping, and divide the collected historical data to determine the training set and test set;

[0112] Based on machine learning technology, a training set is used to train the machine learning model, enabling the machine learning model to autonomously learn the risk prediction behavior of online monitoring of power grid transmission line galloping from the training set and evaluate whether the power grid transmission line galloping is abnormal, thereby determining the online monitoring risk prediction model of power grid transmission line galloping based on machine learning.

[0113] The machine learning-based online monitoring risk prediction model for power grid transmission line galloping was tested using a test set to evaluate its performance and determine whether it can achieve the expected effect of predicting the risk of power grid transmission line galloping.

[0114] When the machine learning-based online monitoring risk prediction model for power grid transmission line galloping fails to achieve the expected prediction effect, the parameters of the machine learning-based online monitoring risk prediction model for power grid transmission line galloping are adjusted, and the model is continuously iterated and optimized until it can achieve the expected prediction effect, thereby determining the optimal online monitoring risk prediction model for power grid transmission line galloping.

[0115] The system provides visualized output of online monitoring and risk prediction results for power grid transmission line galloping, and automatically triggers early warning alarms when power grid transmission line galloping is abnormal.

[0116] In this embodiment, an early warning alarm is automatically triggered when the power grid transmission line gallops abnormally, including:

[0117] The system displays the online monitoring and risk prediction results of power grid transmission line galloping in a visual format to managers in real time. When abnormal behavior of power grid transmission line galloping is detected, an early warning alarm is automatically triggered, prompting managers to carry out timely maintenance and management of the power grid transmission lines.

[0118] In summary, it can perform real-time online monitoring of power grid transmission line galloping and provide timely early warning and control of abnormal power grid transmission line galloping behavior, which can fully ensure the safe and stable operation of the power grid.

[0119] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0120] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A vision-based online monitoring and early warning method for power transmission line galloping, characterized in that, include: Real-time data on power grid transmission line galloping was collected and preprocessed to extract key features of power grid transmission line galloping and determine the characteristic data of power grid transmission line galloping. A risk prediction model for online monitoring of power transmission line galloping was established to analyze and predict the characteristic data of power transmission line galloping, assess whether the power transmission line galloping is abnormal, and determine the risk prediction results of online monitoring of power transmission line galloping. The system provides visualized output of online monitoring and risk prediction results for power grid transmission line galloping, and automatically triggers early warning alarms when power grid transmission line galloping is abnormal.

2. The vision-based online monitoring and early warning method for power transmission line galloping as described in claim 1, characterized in that, Collect real-time data on power grid transmission line galloping, including: Non-contact real-time monitoring of power grid transmission line galloping is carried out using high-definition cameras installed on power grid transmission towers, capturing the movement trajectory of power grid transmission lines and collecting video data of power grid transmission line galloping; Real-time monitoring of wind speed, wind direction, temperature, and humidity parameters during power grid transmission line galloping is achieved using environmental sensors installed on power grid transmission towers, thereby collecting environmental data related to power grid transmission line galloping. Based on video data of power transmission line galloping and environmental data of power transmission line galloping, real-time data of power transmission line galloping is determined.

3. The vision-based online monitoring and early warning method for power transmission line galloping as described in claim 2, characterized in that, Preprocessing of real-time data on power grid transmission line galloping includes: Clean the real-time data of power grid transmission line galloping to remove noise and identify missing and outlier values. The missing and outlier values ​​in the real-time data of power grid transmission line galloping are evaluated to determine whether the missing and outlier values ​​in the real-time data of power grid transmission line galloping are valuable for online monitoring and early warning of power grid transmission line galloping. If missing values ​​and outliers in the real-time data of power transmission line galloping are valuable for online monitoring and early warning of power transmission line galloping, then the mean is used to fill in the missing values ​​and the median is used to replace the outliers. If missing or outlier values ​​in the real-time data of power transmission line galloping are of no value for online monitoring and early warning of power transmission line galloping, then the missing or outlier values ​​should be deleted. The real-time data of power grid transmission line galloping is normalized to eliminate the dimensional differences in the real-time data of power grid transmission line galloping, thus forming standardized real-time data of power grid transmission line galloping.

4. The vision-based online monitoring and early warning method for power transmission line galloping as described in claim 3, characterized in that, Preprocessing of real-time data on power grid transmission line galloping also includes: Contrast enhancement is performed on real-time data of power grid transmission line galloping to highlight the detailed features in the data. Target detection is performed on real-time data of power grid transmission line galloping. The galloping behavior of power grid transmission lines is detected and identified in real time. The galloping behavior of power grid transmission lines is analyzed to extract key features of power grid transmission line galloping and determine the characteristic data of power grid transmission line galloping, including amplitude, frequency and phase.

5. The vision-based online monitoring and early warning method for power transmission line galloping as described in claim 4, characterized in that, Contrast enhancement is performed on real-time data of power grid transmission line galloping, including: Extract image data corresponding to real-time data on power grid transmission line galloping; Each pair of adjacent image data acquisition times is treated as a single image data group. Perform grayscale processing on the two image data contained in the image data group to obtain the grayscale processed image data; The similarity between two grayscale processed image data contained in the image data group is compared to obtain the similarity value between the two grayscale processed image data. The similarity value is compared with a preset similarity threshold; When the similarity value is not lower than a preset similarity threshold, the real-time data of power grid transmission line galloping is enhanced in contrast according to a preset contrast enhancement range, wherein the preset contrast enhancement range is 1%-6%; When the similarity value is lower than the preset similarity threshold, the image data group with the similarity value lower than the preset similarity threshold is taken as the target image data group. For a target image dataset, retrieve the similarity value between two grayscale processed image data points contained within it; Based on the similarity value between the two grayscale processed image data contained in the target image data group, the grayscale values ​​of the two image data contained in the target image data group are adjusted respectively.

6. The vision-based online monitoring and early warning method for power transmission line galloping as described in claim 5, characterized in that, Based on the similarity value between the two grayscale processed image data contained in the target image data group, the grayscale values ​​of the two image data contained in the target image data group are adjusted respectively, including: Retrieve the two grayscale processed image data contained in the target image data group; For each grayscale processed image data, obtain its corresponding average grayscale value. The difference between the average gray values ​​of the two gray-scale processed image data contained in the target image data group is processed, and the absolute value of the difference between the average gray values ​​of the two gray-scale processed image data contained in the target image data group is obtained. The absolute value of the difference is compared with a preset difference threshold; When the absolute value of the difference does not exceed the preset difference threshold, the first contrast enhancement model is used to perform contrast enhancement processing on the two image data contained in the target image data group. When the absolute value of the difference exceeds the preset difference threshold, the second contrast enhancement model is used to perform contrast enhancement processing on the two image data contained in the target image data group.

7. The vision-based online monitoring and early warning method for power transmission line galloping as described in claim 4, characterized in that, Assess whether power grid transmission line galloping is abnormal, including: Based on the vision-based online monitoring and early warning requirements for power grid transmission line galloping, an online monitoring risk prediction model for power grid transmission line galloping is established, and the online monitoring risk prediction model for power grid transmission line galloping is deployed in a real online monitoring and early warning environment for power grid transmission lines galloping. The characteristic data of power transmission line galloping are input into the online monitoring risk prediction model for power transmission line galloping. The model is then used to analyze and predict the characteristic data of power transmission line galloping to assess whether the galloping is abnormal, and finally determine the online monitoring risk prediction result for power transmission line galloping.

8. The vision-based online monitoring and early warning method for power transmission line galloping as described in claim 7, characterized in that, Establish an online monitoring and risk prediction model for power grid transmission line galloping, and perform the following operations: Collect historical data on power grid transmission line galloping, and divide the collected historical data to determine the training set and test set; Based on machine learning technology, a training set is used to train the machine learning model, enabling the machine learning model to autonomously learn the risk prediction behavior of online monitoring of power grid transmission line galloping from the training set and evaluate whether the power grid transmission line galloping is abnormal, thereby determining the online monitoring risk prediction model of power grid transmission line galloping based on machine learning. The machine learning-based online monitoring risk prediction model for power grid transmission line galloping was tested using a test set to evaluate its performance and determine whether it can achieve the expected effect of predicting the risk of power grid transmission line galloping. When the machine learning-based online monitoring risk prediction model for power grid transmission line galloping fails to achieve the expected prediction effect, the parameters of the machine learning-based online monitoring risk prediction model for power grid transmission line galloping are adjusted, and the model is continuously iterated and optimized until it can achieve the expected prediction effect, thereby determining the optimal online monitoring risk prediction model for power grid transmission line galloping.

9. The vision-based online monitoring and early warning method for power transmission line galloping as described in claim 8, characterized in that, An early warning alarm is automatically triggered when there is abnormal galloping of power transmission lines, including: The system provides real-time visualization of the online monitoring and risk prediction results for power grid transmission line galloping to management personnel. When abnormal behavior is detected in the galloping of power grid transmission lines, an early warning alarm is automatically triggered, prompting management personnel to perform timely maintenance and management of the power grid transmission lines.

10. A vision-based online monitoring and early warning system for power grid transmission line galloping, used to implement the vision-based online monitoring and early warning method for power grid transmission line galloping as described in claim 9, characterized in that, include: The front-end acquisition module is used to collect the motion trajectory and environmental parameters of power grid transmission lines during galloping, and to determine the real-time data of power grid transmission line galloping. The data processing module is used to preprocess real-time data on power grid transmission line galloping and determine the characteristic data of power grid transmission line galloping. The risk assessment module is used to analyze and predict the characteristic data of power grid transmission line galloping, and to determine the risk prediction results of online monitoring of power grid transmission line galloping. The early warning output module is used to visualize and output the risk prediction results of online monitoring of power grid transmission line galloping, and automatically triggers an early warning alarm when the power grid transmission line galloping is abnormal.

Images