A method, device and system for controlling and managing a converter station during a maintenance phase

Abstract

The embodiment of the application provides a control method, device and system for a converter station maintenance stage, and belongs to the field of converter station operation and maintenance. The control method comprises the following steps: acquiring video monitoring data of a converter station site; extracting feature information of a reflective vest; performing data analysis on the feature information of the reflective vest and classifying to obtain a corresponding work type category; combining the classified result with regional intrusion of the video monitoring data of the converter station site to obtain personnel intrusion conditions in different operation regions. The control method can monitor the site picture in real time and timely issue an alarm.

Description

Technical Field

[0001] This invention relates to the field of converter station operation and maintenance, and specifically to a control method, device and system for the overhaul phase of a converter station. Background Technology

[0002] Converter stations (substations) are a crucial component of power systems, used to convert, distribute, and control electrical energy from transmission lines. They serve as the intermediate link connecting high-voltage transmission lines to low-voltage distribution networks or consumers. Substations typically include a range of equipment and structures, such as transformers, switchgear, circuit breakers, relays, overhead supports, and control rooms. The size and function of a substation vary depending on the needs of the power system it serves. During operation, converter stations require maintenance personnel to ensure stable operation. However, due to the unique characteristics of converter stations, the safety of maintenance personnel must be guaranteed. Currently, some substations require back-end personnel to monitor and analyze surveillance footage during maintenance, making real-time monitoring of the field impossible and potentially leading to oversights and safety incidents. Summary of the Invention

[0003] The purpose of this invention is to provide a control method, device, and system for the maintenance phase of a converter station. This control method can monitor the site in real time and issue alarms promptly.

[0004] To achieve the above objectives, in one aspect, embodiments of the present invention provide a management and control method for the maintenance phase of a converter station, the management and control method comprising: Acquire video surveillance data at the converter station site; Extract feature information from reflective vests; The feature information of the reflective vest is analyzed and categorized to obtain the corresponding job category; The classification results are combined with the regional intrusion data from the video surveillance data at the converter station to obtain information on personnel intrusion in different work areas.

[0005] Optionally, the feature information extracted from the reflective vest includes: Collect image data of reflective clothing from different angles and under different lighting conditions; The image data of the reflective vest is subjected to image enhancement, noise filtering, and denoising processing. Extract sample features from the reflective parts of the reflective clothing in the image data.

[0006] Optionally, extracting sample features of the light-emitting parts of the reflective vest from the image data includes: Obtain the processed image data; Predict the location of the reflective part of the image data containing the reflective clothing and mark it to obtain a prediction box; The overlap ratio between the predicted bounding box and the marker bounding box is calculated using formula (1): (1) in, The overlap rate, For the prediction box, For the marker box; The classification scores of the predicted boxes to be eliminated are calculated based on the overlap rate, and the highest value of the previously calculated highest classification score is used to retain the predicted box with the highest classification score. Extract the image of the illuminated part from the image data based on the prediction box; The image of the light-emitting part extracted from the prediction box is used to locate the position of the light-emitting part by bilinear interpolation. Extract the sample features of the light-emitting region.

[0007] Optionally, the feature information of the reflective vest can be analyzed and categorized to obtain the corresponding job categories, including: Obtain sample features from the light-emitting area; The sample features of the light-emitting part are cropped into image samples of a certain size; Extract the color of the pixels in the image sample; The colors of the pixels in the image sample are matched with preset colors in the database. Obtain the corresponding job category based on the matching results.

[0008] Optionally, matching the colors of pixels in the image sample with preset colors in the database includes: Obtain the color of the pixels in the image sample; Indicate the X-axis and Y-axis directions of the image sample; The pixels of the image sample are obtained along the Y-axis direction of the image sample; Based on the order of the colors of the pixels that appear consecutively along the Y-axis in the image sample, a preset color with the same color order in the database is matched to obtain the corresponding job category.

[0009] Optionally, matching the colors of pixels in the image sample with preset colors in the database includes: Obtain the color of the pixels in the image sample; The pixels arranged in each column of the image sample are obtained at equal intervals along the X-axis direction of the image sample; Get the number of pixels with consecutive identical colors in each column of pixels; Calculate the maximum and minimum number of pixels with the same color that appear in the same order in different columns; Traverse all columns of the image sample along the X-axis, and subtract the minimum number of pixels with the same color from the number of pixels with the same color that appear sequentially in each column. Calculate the result of the subtraction of each column and divide it by the difference between the maximum and minimum values ​​to obtain the conversion value of pixels with the same color in each column; The number of pixels with the same color that appear sequentially and whose conversion values ​​are greater than a preset threshold is determined as the fixed value; Obtain the order of colors corresponding to the sequentially occurring fixed values, and match them with preset colors that have the same color order in the database to obtain the corresponding job category.

[0010] Optionally, combining the classification results with the regional intrusion data from the converter station's on-site video surveillance to obtain information on personnel intrusion in different work areas includes: The work area is set based on real-time camera footage, and the types of work that can be performed in the work area are also set. Visual recognition technology is used to measure the intensity and angle of light reflected by the reflective vest through a camera in order to determine the presence and location of the reflective vest; Obtain the job category corresponding to the reflective vest; Determine whether the worker wearing the reflective vest is present in the work area; When a worker wearing the reflective vest appears in the work area, it is determined whether the job category corresponding to the reflective vest can be present in the work area; An intrusion warning is issued when the job category corresponding to the reflective vest cannot be found in the work area.

[0011] On the other hand, the present invention also provides a control device for the maintenance phase of a converter station, the control device comprising: Cameras are deployed at the converter station site to acquire on-site information; A background monitor is connected to the camera to receive on-site information from the camera and execute the control method described above.

[0012] Furthermore, the present invention also provides a control system for the maintenance phase of a converter station, the system comprising: The video monitoring module is used to acquire on-site video of the converter station; The data transmission module is used to transmit the live video acquired by the video monitoring module; The background monitoring module is used to execute the control methods described above.

[0013] Through the above technical solution, this invention provides a control method for the maintenance phase of a converter station. This method extracts the characteristic information of reflective vests from video surveillance data at the converter station. After obtaining the reflective vest information, the characteristic information can be analyzed and categorized to determine the corresponding job type. After obtaining the job type corresponding to the reflective vest, the categorization result can be combined with the area intrusion data from the video surveillance data at the converter station to determine the personnel intrusion situation in different work areas. This control method can effectively monitor the on-site operation of the converter station in a timely manner and issue timely alarms to remind staff.

[0014] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0015] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of a control method for the maintenance phase of a converter station according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating the extraction of features of reflective vests in a control method for the maintenance phase of a converter station according to an embodiment of the present invention. Figure 3 This is a flowchart illustrating the extraction of sample features from reflective vests in a control method for the maintenance phase of a converter station according to an embodiment of the present invention. Figure 4 This is a first flowchart of a control method for the maintenance phase of a converter station according to an embodiment of the present invention, which matches the job categories. Figure 5 This is a second flowchart of a control method for the maintenance phase of a converter station according to an embodiment of the present invention, which matches the job categories. Figure 6 This is a third flowchart of a control method for the maintenance phase of a converter station according to an embodiment of the present invention, which matches the types of work. Figure 7 This is a flowchart of a control method for the maintenance phase of a converter station according to an embodiment of the present invention, which is used to determine the intrusion into the control area. Detailed Implementation

[0016] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0017] Figure 1 This is a flowchart of a control method for the maintenance phase of a converter station according to an embodiment of the present invention. In this invention, the control method may include the following steps: In step S1, video monitoring data of the converter station site is acquired.

[0018] In step S2, the feature information of the reflective vest is extracted.

[0019] In step S3, the feature information of the reflective vest is analyzed and categorized to obtain the corresponding job category.

[0020] In step S4, the classification results are combined with the area intrusion data from the video surveillance data of the converter station site to obtain the personnel intrusion situation in different work areas.

[0021] In this invention, to further improve the control methods and approaches during the maintenance phase of converter stations (substations) and strengthen the risk management of directly managed projects, a large-scale integrated digital safety management system for converter stations (substations) based on 5G networks is implemented. This system allows for timely system maintenance and remote safety monitoring of key work areas at the converter station, enabling the timely detection of violations and unsafe conditions and triggering audible and visual alarms. During management, video surveillance data from the converter station can be acquired and analyzed to extract the characteristic information of reflective vests. This characteristic information is then analyzed and categorized to determine the corresponding job categories. These job categories are then combined with the area intrusion data from the converter station's video surveillance to assess intrusion activity in different work areas. When personnel not matching their job category are detected in a work area, a warning is issued to alert staff. This management method effectively monitors the on-site operations of the converter station and promptly issues alarms to alert staff. In one embodiment of the present invention, such as Figure 2 As shown, the process of extracting feature information from reflective vests may include: In step S5, image data of the reflective vest under different angles and lighting conditions are collected.

[0022] In step S6, the image data of the reflective vest is subjected to image enhancement, noise filtering, and denoising processing.

[0023] In step S7, sample features of the reflective parts of the reflective clothing in the image data are extracted.

[0024] In this invention, after acquiring video surveillance data from the converter station, the data can be analyzed to obtain information about the reflective vest. Image data of the reflective vest under different angles and lighting conditions can be collected. After obtaining the image data, image enhancement, noise filtering, and denoising processes can be performed to make the image data more suitable for the requirements. After processing the image data, sample features of the light-emitting parts of the reflective vest can be extracted. These sample features represent the characteristic information of the reflective vest.

[0025] In one embodiment of the invention, such as Figure 3 As shown, the process for extracting sample features from reflective clothing may include: In step S8, the processed image data is acquired.

[0026] In step S9, the position of the reflective part with reflective clothing in the image data is predicted and marked to obtain the prediction box.

[0027] In step S10, the overlap rate between the predicted bounding box and the marker bounding box is calculated using formula (1): (1) in, The overlap rate, For the prediction box, For the marker box.

[0028] In step S11, the classification score of the predicted box to be removed is calculated based on the overlap rate, and the highest value of the previously calculated highest classification score is used to retain the predicted box with the highest classification score.

[0029] In step S12, the image of the illuminated part in the image data is extracted based on the prediction box.

[0030] In step S13, the position of the light-emitting part is located in the image of the light-emitting part extracted from the prediction box using the bilinear interpolation method.

[0031] In step S14, the sample features of the light-emitting part are extracted.

[0032] In this invention, after obtaining the processed image data, the processed image data is extracted to obtain sample features of the reflective parts. During this processing, the position of the reflective parts with reflective clothing in the image data can be predicted and marked to obtain prediction boxes. Then, the overlap rate between the prediction boxes and the marked boxes can be obtained through formula (1). After obtaining the overlap rate, prediction boxes that meet the requirements can be filtered according to the size of the overlap rate. When filtering by overlap rate, the classification score of the prediction box to be eliminated and the highest value of the previously calculated highest classification score can be calculated based on the overlap rate of the prediction boxes. In this way, the classification score corresponding to the prediction box is not necessarily the original score of this prediction box, but the highest classification score of the box proposed based on the prediction box, making the accuracy of the prediction box more accurate. After obtaining the prediction box, the image of the reflective parts in the image data can be extracted according to the prediction box. Then, the position of the reflective parts can be located by the bilinear interpolation method on the image of the reflective parts extracted from the prediction box. After locating the image of the reflective parts, the sample features of the reflective parts can be extracted.

[0033] In one embodiment of the present invention, such as Figure 4 As shown, the first step in matching job categories may include: In step S15, the sample features of the light-emitting part are obtained.

[0034] In step S16, the sample features of the light-emitting part are cropped into image samples of a certain size.

[0035] In step S17, the color of the pixels in the image sample is extracted.

[0036] In step S18, the colors of the pixels in the image sample are matched with preset colors in the database.

[0037] In step S19, the corresponding job category is obtained based on the matching result.

[0038] In this invention, the reflective vest's light-emitting portion can be strip-shaped, worn on the back of the worker. This strip-shaped light-emitting portion can have different color combinations, each representing a different job category. Therefore, after obtaining sample features, the job category of the worker wearing the reflective vest can be determined based on the color combination of the light-emitting portion, thus determining whether the worker has intruded into the work area. To determine the job category corresponding to the light-emitting portion, sample features of the light-emitting portion can be obtained and then cropped into an image sample of a certain size. After obtaining the image sample, the colors of the pixels in the image sample can be matched with preset colors in a database. The preset colors in the database are pre-configured and can be different color combinations, not just a single color; different preset colors can represent different job categories. After obtaining the color of the image sample, this color is compared with the sample color. If the color order is consistent, the job category corresponding to the image sample can be the same as the preset color.

[0039] In one embodiment of the present invention, such as Figure 5 As shown, the second process for matching job categories may include: In step S20, the color of the pixel in the image sample is obtained.

[0040] In step S21, the X-axis direction and Y-axis direction of the image sample are indicated.

[0041] In step S22, the pixels of the image sample are obtained along the Y-axis direction of the image sample.

[0042] In step S23, based on the order of the colors of the pixels that appear consecutively along the Y-axis in the image sample, a preset color with the same color order in the database is matched to obtain the corresponding job category.

[0043] In this invention, when matching the colors of pixels in an image sample, the colors of the pixels in the image sample can be obtained first. The image sample can have a certain size, so its X-axis and Y-axis directions can be marked. After marking the directions of the image sample, since the reflective part of the worker's reflective vest in this invention is a multi-layered color combination of reflective strips, after cropping the reflective part, the color combination of the reflective part can be obtained by collecting data from the vertical direction. Pixels of the image sample can be obtained along the Y-axis direction. After obtaining the color order of the pixels that appear consecutively along the Y-axis direction of the image sample, a preset color with the same color appearance order can be matched in the database. The job category corresponding to this preset color can also be the job category corresponding to the image sample, thereby obtaining the job category of the worker wearing the reflective vest.

[0044] In one embodiment of the present invention, such as Figure 6 As shown, the third process for matching job categories may include: In step S24, the color of the pixel in the image sample is obtained.

[0045] In step S25, pixels arranged in each column of the image sample are obtained at equal intervals along the X-axis direction of the image sample.

[0046] In step S26, the number of pixels with consecutive identical colors in each column is obtained.

[0047] In step S27, the maximum and minimum values ​​of the number of pixels with the same color that appear in the same order in different columns are calculated.

[0048] In step S28, all columns of the image sample are traversed along the X-axis, and the minimum number of pixels with the same color that appear sequentially in each column is subtracted from the minimum number of pixels with the same color.

[0049] In step S29, the result of subtraction of each column is calculated and divided by the difference between the maximum and minimum values ​​to obtain the conversion value of pixels with the same color in each column.

[0050] In step S30, the number of pixels with the same color that appear sequentially and whose conversion values ​​are greater than a preset threshold is determined as a fixed value.

[0051] In step S31, the order of colors corresponding to the sequentially appearing fixed values ​​is obtained, and the preset colors with the same color order in the database are matched to obtain the corresponding job category.

[0052] In this invention, when workers wear reflective vests, the reflective parts may be partially obscured by dirt. This dirt can affect the extraction of pixel colors from image samples, thus requiring efforts to minimize its impact. In this invention, when obtaining the color of pixels in an image sample, pixels arranged in each column of the image sample can be obtained at equal intervals along the X-axis. The number of pixels with consecutive identical colors in each column is then calculated. The maximum and minimum values ​​of the number of pixels with the same color appearing in the same order in different columns can then be calculated. After obtaining the maximum and minimum values, all columns of the image sample can be traversed along the X-axis. The minimum value of the number of pixels with the same color appearing sequentially in each column is subtracted from the maximum value. The result of the subtraction for each column is then divided by the difference between the maximum and minimum values ​​to obtain a conversion value for pixels with the same color in each column. This conversion value makes the distribution of pixels with the same color in each column more compact. After obtaining the conversion values ​​of the same color appearing sequentially in each column, the number of pixels with the same color whose conversion values ​​are greater than a preset threshold can be calculated as a definite value. When the conversion value of a pixel representing a color in a column is less than the preset value, that pixel is discarded. When the conversion value of a pixel representing a color in a column is greater than the preset value, that conversion value can be considered, and the definite value can be incremented by one. Then, the pixels in the next column are judged. When the definite value is greater than the preset threshold, the color corresponding to that definite value is the definite color. After obtaining the color corresponding to the definite value, the order of the colors corresponding to the sequentially appearing definite values ​​can be obtained. This color order is the color order in the image sample. Then, it is matched with the preset colors with the same color order in the database, and the corresponding job category can be obtained. This method can minimize the impact of stains and obscuring the image, so as to accurately determine the job category corresponding to the image sample.

[0053] In one embodiment of the present invention, such as Figure 7 As shown, the process for determining area intrusion may include: In step S32, the work area is set based on the real-time image from the camera, and the types of work that can be performed in the work area are set.

[0054] In step S33, visual recognition technology is used to measure the intensity and angle of light reflected by the reflective vest through a camera to determine the presence and location of the reflective vest.

[0055] In step S34, the job category corresponding to the reflective vest is obtained.

[0056] In step S35, it is determined whether the worker wearing reflective clothing is present in the work area.

[0057] In step S36, when a worker wearing a reflective vest appears in the work area, it is determined whether the job category corresponding to the reflective vest can appear in the work area.

[0058] In step S37, when the job category corresponding to the reflective vest cannot appear in the work area, an intrusion warning is output.

[0059] In this invention, when determining if an area is intruded, a work area can be set based on real-time camera footage, and the types of jobs that can be performed within that area can be defined. The location of a reflective vest in the camera's view can then be determined, thereby identifying the location of the worker wearing the vest. After obtaining the worker's job type, it can be determined whether the worker wearing the reflective vest is present in the work area. If the worker is present, it can be further determined whether the job type corresponding to the reflective vest is present in that area. If the job type corresponding to the reflective vest is not present in the work area, an intrusion warning can be issued promptly.

[0060] On the other hand, the present invention also provides a control device for the maintenance phase of a converter station, the control device including: Cameras are deployed at the converter station site to acquire on-site information; The background monitor, connected to the camera, can receive on-site information from the camera and execute the control methods described above.

[0061] Furthermore, the present invention also provides a control system for the maintenance phase of a converter station, which may include: The video monitoring module is used to acquire on-site video of the converter station; The data transmission module is used to transmit the live video acquired by the video surveillance module; The background monitoring module is used to execute the control methods described above.

[0062] Through the above technical solution, this invention provides a control method for the maintenance phase of a converter station. This method extracts the characteristic information of reflective vests from video surveillance data at the converter station. After obtaining the reflective vest information, the characteristic information can be analyzed and categorized to determine the corresponding job type. After obtaining the job type corresponding to the reflective vest, the categorization result can be combined with the area intrusion data from the video surveillance data at the converter station to determine the personnel intrusion situation in different work areas. This control method can effectively monitor the on-site operation of the converter station in a timely manner and issue timely alarms to remind staff.

[0063] It should also be noted that 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. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0064] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A control method for the maintenance phase of a converter station, characterized in that, The control methods include: Acquire video surveillance data at the converter station site; Extract feature information from reflective vests; The feature information of the reflective vest is analyzed and categorized to obtain the corresponding job category; The classification results are combined with the regional intrusion data of the video surveillance data at the converter station to obtain the personnel intrusion situation in different work areas. The feature information of the reflective vest is analyzed and categorized to obtain the corresponding job categories, including: Obtain sample features from the light-emitting area; The sample features of the light-emitting part are cropped into image samples of a certain size; Extract the color of the pixels in the image sample; The colors of the pixels in the image sample are matched with preset colors in the database. Obtain the corresponding job category based on the matching results; The step of matching the color of the pixels in the image sample with a preset color in the database includes: Obtain the color of the pixels in the image sample; The pixels arranged in each column of the image sample are obtained at equal intervals along the X-axis direction of the image sample; Get the number of pixels with consecutive identical colors in each column of pixels; Calculate the maximum and minimum number of pixels with the same color that appear in the same order in different columns; Traverse all columns of the image sample along the X-axis, and subtract the minimum number of pixels with the same color from the number of pixels with the same color that appear sequentially in each column. Calculate the result of the subtraction of each column and divide it by the difference between the maximum and minimum values ​​to obtain the conversion value of pixels with the same color in each column; The number of pixels with the same color that appear sequentially and whose conversion values ​​are greater than a preset threshold is determined as the fixed value; Obtain the order of colors corresponding to the sequentially occurring fixed values, and match them with preset colors that have the same color order in the database to obtain the corresponding job category.

2. The control method according to claim 1, characterized in that, The feature information extracted from reflective vests includes: Collect image data of reflective clothing from different angles and under different lighting conditions; The image data of the reflective vest is subjected to image enhancement, noise filtering, and denoising processing. Extract sample features from the reflective parts of the reflective vest in the image data.

3. The control method according to claim 2, characterized in that, Extracting sample features of the light-emitting parts of the reflective vest from the image data includes: Obtain the processed image data; Predict the location of the reflective part of the image data containing the reflective clothing and mark it to obtain a prediction box; The overlap ratio between the predicted bounding box and the marker bounding box is calculated using formula (1): ，（1） in, The overlap rate, For the prediction box, For the marker box; The classification scores of the predicted boxes to be eliminated are calculated based on the overlap rate, and the highest value of the previously calculated highest classification score is used to retain the predicted box with the highest classification score. Extract the image of the illuminated part from the image data based on the prediction box; The position of the light-emitting part is located by bilinear interpolation of the image of the light-emitting part extracted from the prediction box. Extract the sample features of the light-emitting region.

4. The control method according to claim 1, characterized in that, The step of matching the color of the pixels in the image sample with a preset color in the database includes: Obtain the color of the pixels in the image sample; Indicate the X-axis and Y-axis directions of the image sample; The pixels of the image sample are obtained along the Y-axis direction of the image sample; Based on the order of the colors of the pixels that appear consecutively along the Y-axis in the image sample, a preset color with the same color order in the database is matched to obtain the corresponding job category.

5. The control method according to claim 1, characterized in that, The process of combining the classification results with the regional intrusion data from the converter station's video surveillance to obtain information on personnel intrusion in different work areas includes: The work area is set based on real-time camera footage, and the types of work that can be performed in the work area are also set. Visual recognition technology is used to measure the intensity and angle of light reflected by the reflective vest through a camera in order to determine the presence and location of the reflective vest; Obtain the job category corresponding to the reflective vest; Determine whether the worker wearing the reflective vest is present in the work area; When a worker wearing the reflective vest appears in the work area, it is determined whether the job category corresponding to the reflective vest can be present in the work area; An intrusion warning is issued when the job category corresponding to the reflective vest cannot be found in the work area.

6. A control device for the maintenance phase of a converter station, characterized in that, The control device includes: Cameras are deployed at the converter station site to acquire on-site information; A background monitor is connected to the camera to receive on-site information from the camera and execute the control method as described in any one of claims 1-5.

7. A control system for the maintenance phase of a converter station, characterized in that, The system includes: The video monitoring module is used to acquire on-site video of the converter station; The data transmission module is used to transmit the live video acquired by the video monitoring module; The background monitoring module is used to execute the control method as described in any one of claims 1-5.

Images