Substation safety management system and method based on embedded device

Abstract

The invention discloses a substation safety management system and method based on embedded devices, relates to the field of substation management, solves the problem that the current monitoring method of the area where the substation is located is prone to waste of resources and cannot timely warn, comprises an area division module, a configuration optimization module, a video processing module, an intelligent distance measurement module, and a model construction module, wherein the area division module is used to divide the supervision area corresponding to the substation, the configuration optimization module is used to optimize the supervision configuration in the peripheral area and the core area, the model construction module is used to construct a target intrusion detection algorithm large model used for a data frame diagram, the video processing module is used to process the real-time monitoring video data of the supervision area corresponding to the substation, the intelligent distance measurement module is used to measure the distance between the intrusion target within the supervision range and the substation, the invention intelligently configures the monitoring equipment in the area where the substation is located, and realizes timely and accurate warning of the area where the substation is located.

Description

Technical Field

[0001] The present invention relates to substation safety management technology, and in particular to a substation safety management system and method based on embedded devices. Background Art

[0002] With the development of economy, the demand for electric energy is increasing. The stable, reliable and safe operation of power system affects the social development and people's life. As the core node of power system, the stable operation of substation is very important. In today's society, the wide application of intelligent monitoring system in substation monitoring has brought great convenience to people. The existing technology of intelligent monitoring of substation mainly adopts centralized data processing, which transmits image data to centralized GPU equipment to process and analyze the image.

[0003] Currently, when monitoring the area where the substation is located, monitoring equipment with unified standards is usually used to monitor the substation area. The substation area is not finely divided based on the intrusion situation, nor is monitoring equipment with appropriate configuration selected for monitoring based on the intrusion situation, resulting in a waste of resources. At the same time, the current monitoring method for the substation area is mostly regular patrol of the monitoring screen, which cannot effectively achieve timely and accurate early warning of the substation area.

[0004] To this end, the present invention proposes a substation safety management system and method based on embedded devices. Summary of the invention

[0005] The purpose of the present invention is to provide a substation safety management system and method based on embedded devices in view of the deficiencies in the prior art.

[0006] The technical problems to be solved by the present invention are:

[0007] How to realize intelligent configuration of monitoring equipment in the area where the substation is located based on intrusion data and how to realize timely and accurate early warning in the area where the substation is located.

[0008] The purpose of the present invention can be achieved through the following technical solutions:

[0009] The substation safety management system based on embedded devices includes a regional division module, a database, a configuration optimization module, a data acquisition module, a video processing module, an intelligent distance measurement module, a model building module and an alarm display terminal;

[0010] The area division module is used to divide the supervision area corresponding to the substation into a core area and a peripheral area;

[0011] The database is used to record historical intrusion data of different peripheral areas and different core areas, and the configuration optimization module is used to optimize the supervision configuration in the peripheral area and the core area to obtain the supervision plan corresponding to the peripheral area or the core area;

[0012] After the corresponding supervision scheme is implemented in the core area and the peripheral area, the data acquisition module is used to collect the real-time monitoring video data of the corresponding supervision area of ​​the substation in real time and send it to the video processing module; the model construction module is used to construct the target intrusion detection algorithm large model used in the data frame diagram, and send the target intrusion detection algorithm large model to the video processing module;

[0013] The video processing module is used to process the real-time monitoring video data of the supervision area corresponding to the substation, generate a ranging signal or do not perform any operation. If a ranging signal is generated, it is sent to the intelligent ranging module; the intelligent ranging module is used to measure the distance between the intrusion target and the substation within the supervision range, do not perform any operation or generate a safety warning signal. If a safety warning signal is generated and the corresponding abnormal picture is sent to the alarm display terminal, the alarm display terminal is used to alarm the safety warning signal and abnormal picture of the substation.

[0014] Furthermore, the division process of the core area and the peripheral area is as follows:

[0015] Set warning distances within the supervision area;

[0016] Expand outward from the substation as the center until the expansion reaches the warning line where the warning distance is located;

[0017] The cordon surrounds the substation. The supervision area surrounded by the cordon is the core area, and the supervision area excluding the core area is the peripheral area. The core area and the peripheral area are both rectangular.

[0018] Connect the right angle of the peripheral area with the right angle of the core area on the same side through a straight line, and extend the straight line to the substation. The straight line is used as a set of dividing lines of the regulatory area. Similarly, three additional sets of dividing lines of the regulatory area are obtained.

[0019] The peripheral area and the core area are divided by a dividing line, and the peripheral areas are numbered in a clockwise direction to obtain a first peripheral area, a second peripheral area, a third peripheral area and a fourth peripheral area. Similarly, a first core area, a second core area, a third core area and a fourth core area are obtained.

[0020] Furthermore, the historical intrusion data includes the number of foreign object intrusions into different peripheral areas and different core areas, as well as the intrusion residence time and intrusion distance of each foreign object intrusion. The intrusion distance is the distance between the substation and the supervision area after entering the supervision area.

[0021] Furthermore, the optimization process of the configuration optimization module is specifically as follows:

[0022] Get the number of foreign objects intrusion into the peripheral area and the core area RCu, where u is the number of the peripheral area and the core area;

[0023] Then, the invasion residence time and invasion distance of each foreign object invasion in the peripheral area and the core area are obtained, and the invasion residence time of each foreign object invasion is traversed and compared, and the maximum invasion residence time is taken as the foreign object invasion time RTu of the peripheral area and the core area. At the same time, the invasion distance of each foreign object invasion is traversed and compared, and the minimum invasion distance is taken as the foreign object invasion distance RJu of the peripheral area and the core area;

[0024] The regional intrusion values ​​QRu of the peripheral area and the core area are calculated by the formula. The specific formula is as follows:

[0025] QRu=(RCu×a1+RTu×a2+1 / RJu)×λ; where a1 and a2 are weight coefficients with fixed values, a2>a1, λ is the safety warning coefficient, and the safety warning coefficient of the core area is greater than that of the peripheral area;

[0026] The regional intrusion value is compared with the regional intrusion threshold, where the regional intrusion threshold includes a first regional intrusion threshold and a second regional intrusion threshold, and the second regional intrusion threshold is greater than the first regional intrusion threshold;

[0027] If the regional intrusion value is less than the first regional intrusion threshold, the peripheral area or the core area implements the third supervision scheme;

[0028] If the regional intrusion value is greater than or equal to the first regional intrusion threshold and less than the second regional intrusion threshold, the peripheral area or the core area implements the second supervision plan;

[0029] If the regional intrusion value is greater than or equal to the second regional intrusion threshold, the peripheral area or the core area executes the first supervision scheme.

[0030] Furthermore, the first monitoring solution is: using multiple high-configuration surveillance cameras;

[0031] The second monitoring solution is: using multiple surveillance cameras with common configuration;

[0032] The third monitoring solution is to use a single-channel surveillance camera with ordinary configuration.

[0033] Furthermore, the regulatory intensity of the first regulatory scheme is higher than that of the second regulatory scheme, and the regulatory intensity of the second regulatory scheme is higher than that of the third regulatory scheme.

[0034] Furthermore, the processing process of the video processing module is specifically as follows:

[0035] Obtain the real-time monitoring video data of the corresponding supervision area of ​​the substation, and extract the real-time monitoring video data frame by frame to obtain a frame-by-frame image;

[0036] Performing grayscale processing on each frame of the image to obtain a first data frame image;

[0037] Then, the first data frame is subjected to denoising to obtain a second data frame, and the target intrusion detection algorithm model is used to perform intrusion target detection on the second data frame;

[0038] If an intruder is detected in any core area or peripheral area, a ranging signal is generated;

[0039] If no intrusion target is detected in all core areas and peripheral areas, no action is taken.

[0040] Furthermore, the working process of the intelligent ranging module is as follows:

[0041] Obtain monitoring images in real-time monitoring video data based on the current moment;

[0042] A rectangular coordinate system corresponding to the monitoring screen is established with the origin of the monitoring camera, the vertical direction of the monitoring camera as the Y axis, and the connecting line between the lower end of the monitoring camera and the lower end of the intrusion target as the X axis;

[0043] With O as the origin, CO as the Y-axis, and OM as the X-axis, a rectangular coordinate system corresponding to the monitoring screen is established. CO is the height of the monitoring camera. The monitoring camera is vertically set at the substation. CO is the height H of the monitoring camera. Point C is the upper end of the monitoring camera. Point 0 is the foot of the vertical line of the monitoring camera on the ground where the substation is located. RQ is the intrusion height h of the intrusion target. Point R is the upper end of the intrusion target. Point Q is the foot of the vertical line of the intrusion target on the plane where the substation is located. Connect the upper end C of the monitoring camera with the upper end R of the intrusion target, and make an extension line of the upper end of the monitoring camera and the upper end of the intrusion target until they intersect with the ground where the substation is located. The intersection point is point M.

[0044] The coordinates of point O are known to be (x1, y1) and the angle α, where the angle α is the depression angle of the surveillance camera.

[0045] Furthermore, the working process of the intelligent ranging module also includes:

[0046] The length of OQ is calculated by the formula, which is as follows:

[0047] OQ = MO-MQ;

[0048] That is, OQ = H / tanα-h / tanα;

[0049] It is known that the offset angle of the surveillance camera is β, and point P is the center point of the core area, so the angle of the angle QOP is 180°-β;

[0050] Therefore, the radius OP of the core region is L and the length of OQ is H / tanα-h / tanα;

[0051] The length of QP is:

[0052] α1 is the depression angle of the surveillance camera, and α1=α;

[0053] Since the radius of the core area is known, the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the current moment is:

[0054] Real-time distance = QP-L;

[0055] If the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the current moment is greater than or equal to the safe distance, the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the next moment is calculated;

[0056] If the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the current moment is less than the safety distance, the time when the intrusion target enters the screen is recorded, and the residence time of the intrusion target is obtained by subtracting the entry time from the current time. When the residence time exceeds the duration threshold, a security warning signal is generated, and the current monitoring screen is recorded as an abnormal screen. When the residence time does not exceed the duration threshold, no operation is performed.

[0057] Second, the substation safety management method based on embedded devices is as follows:

[0058] Step S100, dividing the corresponding supervision area of ​​the substation, optimizing the supervision configuration in the peripheral area and the core area according to the historical intrusion data, and obtaining the supervision plan of the peripheral area or the core area;

[0059] Step S200, after executing the corresponding supervision scheme in the core area and the peripheral area, real-time monitoring video data of the corresponding supervision area of ​​the substation is collected in real time;

[0060] Step S300, constructing a target intrusion detection algorithm large model used in the data frame diagram, processing the real-time monitoring video data of the corresponding supervision area of ​​the substation in combination with the target intrusion detection algorithm large model, generating a ranging signal or performing no operation;

[0061] Step S400: If a distance measurement signal is generated, the distance between the intrusion target and the substation within the supervision range is measured; if a safety warning signal and a corresponding abnormal screen are generated, an alarm is displayed.

[0062] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0063] 1. The present invention first divides the corresponding supervision area of ​​the substation, and optimizes the supervision configuration in the peripheral area and the core area according to the historical intrusion data to obtain the supervision plan of the peripheral area or the core area. The present invention realizes the intelligent configuration of the monitoring equipment in the area where the substation is located in combination with the intrusion data;

[0064] 2. After executing the corresponding supervision scheme in the core area and the peripheral area, the present invention collects real-time monitoring video data of the supervision area corresponding to the substation in real time, and constructs a target intrusion detection algorithm large model used in the data frame diagram. The real-time monitoring video data of the supervision area corresponding to the substation is processed in combination with the target intrusion detection algorithm large model, and a ranging signal is generated or no operation is performed. If a ranging signal is generated, the distance between the intrusion target and the substation within the supervision range is calculated. If a safety warning signal and a corresponding abnormal picture are generated, an alarm is displayed. The present invention realizes timely and accurate warning of the area where the substation is located. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0066] Figure 1 is the overall system block diagram of the present invention;

[0067] Figure 2 It is a schematic diagram of the division of the peripheral area and the core area in the present invention;

[0068] Figure 3 It is a schematic diagram of the arrangement of multiple surveillance cameras in the present invention;

[0069] Figure 4 The figure is a schematic diagram of the arrangement of a single-channel surveillance camera in the present invention.

[0070] Figure 5 It is a schematic diagram of the distance measurement between the intrusion target and the substation in the present invention.

[0071] Figure 6 It is a side schematic diagram of the intrusion target and the surveillance camera in the present invention.

[0072] Figure 7 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0073] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0074] Example 1: Please refer to Figure 1-Figure 6 As shown, the technical solution provided by the present invention is: a substation safety management system based on an embedded device, the range of the substation that needs to be managed safely is marked as a supervision area, and the substation is centrally set in the supervision area. The supervision area can be rectangular or circular. In this embodiment, the outer area in the supervision area is preferably rectangular, and the core area in the supervision area is preferably circular;

[0075] In this embodiment, it includes a region division module, a database, a configuration optimization module, a data acquisition module, a video processing module, an intelligent distance measurement module, a model construction module and an alarm display terminal;

[0076] The area division module is used to divide the supervision area corresponding to the substation into a core area and a peripheral area;

[0077] Among them, Figure 2 As shown in the figure, the division process of the core area and the peripheral area is as follows:

[0078] Set warning distances within the supervision area;

[0079] Expand outward from the substation as the center until the expansion reaches the warning line where the warning distance is located;

[0080] The cordon surrounds the substation. The supervision area surrounded by the cordon is the core area, and the supervision area excluding the core area is the peripheral area. The core area and the peripheral area are both rectangular.

[0081] Connect the right angle of the peripheral area with the right angle of the core area on the same side through a straight line, and extend the straight line to the substation. The straight line is used as a set of dividing lines of the regulatory area. Similarly, three additional sets of dividing lines of the regulatory area are obtained.

[0082] The peripheral area and the core area are divided by a dividing line, and the peripheral areas are numbered in a clockwise direction to obtain a first peripheral area, a second peripheral area, a third peripheral area, and a fourth peripheral area. Similarly, a first core area, a second core area, a third core area, and a fourth core area are obtained.

[0083] In this embodiment, the database is connected to the configuration optimization module, and the database is used to record the historical intrusion data of different peripheral areas and different core areas, wherein the historical intrusion data is the number of foreign object intrusions in different peripheral areas and different core areas, and the intrusion residence time and intrusion distance of each foreign object intrusion, and the intrusion distance is the interval distance from the substation after entering the supervision area;

[0084] The configuration optimization module is used to optimize the supervision configuration in the peripheral area and the core area. The optimization process is as follows:

[0085] Get the number of foreign objects intrusion into the peripheral area and the core area RCu, where u is the number of the peripheral area and the core area;

[0086] Then, the invasion residence time and invasion distance of each foreign object invasion in the peripheral area and the core area are obtained, and the invasion residence time of each foreign object invasion is traversed and compared, and the maximum invasion residence time is taken as the foreign object invasion time RTu of the peripheral area and the core area. At the same time, the invasion distance of each foreign object invasion is traversed and compared, and the minimum invasion distance is taken as the foreign object invasion distance RJu of the peripheral area and the core area;

[0087] The regional intrusion values ​​QRu of the peripheral area and the core area are calculated by the formula. The specific formula is as follows:

[0088] QRu=(RCu×a1+RTu×a2+1 / RJu)×λ; where a1 and a2 are weight coefficients with fixed values, a2>a1, λ is the safety warning coefficient, and the safety warning coefficient of the core area is greater than that of the peripheral area;

[0089] It can be understood that the regional intrusion value is used to reflect the degree of intrusion in the peripheral area and the core area. Therefore, the number of foreign object intrusions and the length of intrusion stay are proportional to the regional intrusion value, and the intrusion distance is inversely proportional to the regional intrusion value;

[0090] The regional intrusion value is compared with the regional intrusion threshold, where the regional intrusion threshold includes a first regional intrusion threshold and a second regional intrusion threshold, and the second regional intrusion threshold is greater than the first regional intrusion threshold;

[0091] If the regional intrusion value is less than the first regional intrusion threshold, the peripheral area or the core area implements the third supervision scheme;

[0092] If the regional intrusion value is greater than or equal to the first regional intrusion threshold and less than the second regional intrusion threshold, the peripheral area or the core area implements the second supervision plan;

[0093] If the regional intrusion value is greater than or equal to the second regional intrusion threshold, the peripheral area or the core area implements the first supervision plan;

[0094] The first monitoring solution is: using multiple high-configuration surveillance cameras;

[0095] The second monitoring solution is: using multiple surveillance cameras with common configuration;

[0096] The third monitoring solution is: using a single-channel surveillance camera with ordinary configuration;

[0097] It should be specifically noted that the supervision intensity of the first supervision scheme is higher than that of the second supervision scheme, and the supervision intensity of the second supervision scheme is higher than that of the third supervision scheme; the configuration difference of the surveillance cameras is reflected in the graphics processor, and the difference between single-channel and multi-channel is reflected in the number of surveillance cameras. In this embodiment, Figure 3 and Figure 4 As shown, the number of multi-channel surveillance cameras can be three groups. The multi-channel surveillance camera setting points and single-channel surveillance camera setting points in the core area are at the edge of the substation, and the multi-channel surveillance camera setting points and single-channel surveillance camera setting points in the peripheral area are on the cordon.

[0098] As a further solution of the present invention, after the corresponding supervision scheme is implemented in the core area and the peripheral area, during actual operation, the data acquisition module is used to collect real-time monitoring video data of the corresponding supervision area of ​​the substation in real time, and send the real-time monitoring video data to the video processing module;

[0099] In practice, the data acquisition module is a surveillance camera installed in the peripheral area and the core area, and the surveillance camera is used to collect surveillance video data in real time;

[0100] In this embodiment, the video processing module is used to process the real-time monitoring video data of the supervision area corresponding to the substation, and the processing process is as follows:

[0101] Acquire the real-time monitoring video data of the supervision area corresponding to the substation, and extract the real-time monitoring video data frame by frame to obtain a frame-by-frame graph Wj; wherein j is the number of the frame-by-frame graph, j=1, 2, ..., m, and m is a positive integer;

[0102] Using a weighted average algorithm to perform grayscale processing on each frame of the image to obtain a first data frame image;

[0103] The pixel values ​​of all pixels in the frame-by-frame image are calculated by the formula, and the pixel values ​​corresponding to all pixels in the frame-by-frame image are replaced by the gray value Gary. The calculation formula is as follows:

[0104] Gary=0.299*RWj+0.587*GWj+0.114*BWj;

[0105] Among them, RWj represents the R channel value of the pixel point in the frame-by-frame image, GWj represents the G channel value of the pixel point in the frame-by-frame image, and BWj represents the B channel value of the pixel point in the frame-by-frame image;

[0106] It should be noted that the weighted average algorithm extracts brightness information from color channels by using different weights. Usually, according to the different settings of the human eye's sensitivity to different colors, the green channel has the highest weight, the red channel has the second highest weight, and the blue channel has the lowest weight.

[0107] The first data frame is de-noised using a Gaussian filter to obtain a second data frame, and the second data frame is detected as an intrusion target using a large model of a target intrusion detection algorithm based on yolo;

[0108] If an intruder is detected in any core area or peripheral area, a ranging signal is generated;

[0109] If no intrusion target is detected in all core areas and peripheral areas, no operation will be performed;

[0110] The video processing module sends the ranging signal to the intelligent ranging module;

[0111] In this embodiment, the model building module is used to build a large model of the target intrusion detection algorithm used in the data frame graph, and the construction process is as follows:

[0112] Step T1, collect a large amount of image data containing intrusion scenes from public resources;

[0113] Step T2: Use the labeling tool LabelImg to label the image, mark the location and category of the intrusion target, and obtain the image labeling dataset in YOLO format;

[0114] Step T3, rotating, scaling and inverting the image annotation dataset to obtain an image dataset, thereby increasing the diversity of the image data;

[0115] Step T4, dividing the image data set into a ratio of 9:1 to obtain a training set and a validation set;

[0116] Step T5, using the YOLO network to train based on the training set and the validation set to obtain a large model of the YOLO-based target intrusion detection algorithm;

[0117] It should be specifically noted that Yolo is a convolutional neural network with regression function, which can detect the position and category of multiple target objects in real time at one time;

[0118] The step T5 includes the following sub-steps:

[0119] Step T501, the yolo network divides the images in the training set into S*S grid units, where S is a positive integer;

[0120] It should be specifically noted that the grid unit is used to detect objects within the grid unit. During the training process, each grid unit predicts K bounding boxes and the confidence of the bounding boxes, where K is a positive integer. The confidence includes two aspects: one is the possibility that the bounding box contains the object Pr(object), and the other is the accuracy of the bounding box Where truth represents the actual box, pred represents the predicted box, and The intersection-over-union ratio is calculated to measure the overlap between the predicted box and the actual box, and then the accuracy of the prediction is determined. The confidence reflects the possibility of the target object existing in the current bounding box and the accuracy of the bounding box predicting the position of the target object. Therefore, the confidence can be defined as When there is a target object in the bounding box, Pr(object)=1, and the confidence is further calculated based on the predicted bounding box and the actual box; when there is no target object in the bounding box, Pr(object)=0;

[0121] Step T502, setting the hyperparameters for model training, including: setting the intersection-over-union ratio threshold between the predicted box and the actual box to determine positive and negative samples; setting the number and size of the first inspection boxes to detect objects of different sizes and shapes; setting the number of samples in each training batch and the number of training rounds; setting the learning rate for training to control the update rate of the model weights;

[0122] Step T503, training is performed according to the set hyperparameters of the model training to obtain a large model of the target intrusion detection algorithm based on YOLO;

[0123] Step T504, input the verification set into the YOLO-based target intrusion detection algorithm large model for training verification, and then adjust the hyperparameters of the model training to improve the recognition accuracy of the YOLO-based target intrusion detection algorithm large model;

[0124] The module building module sends the target intrusion detection algorithm model to the video processing module.

[0125] The intelligent distance measurement module is used to measure the distance between the intrusion target and the substation within the supervision range. The specific measurement process is as follows:

[0126] Obtain monitoring images in real-time monitoring video data based on the current moment;

[0127] A rectangular coordinate system corresponding to the monitoring screen is established with the origin of the monitoring camera, the vertical direction of the monitoring camera as the Y axis, and the connecting line between the lower end of the monitoring camera and the lower end of the intrusion target as the X axis;

[0128] like Figure 5 and Figure 6 As shown, a rectangular coordinate system corresponding to the monitoring screen is established with O as the origin, CO as the Y axis, and OM as the X axis. CO is the height of the monitoring camera. The monitoring camera is vertically set at the substation. CO is the height H of the monitoring camera. Point C is the upper end of the monitoring camera. Point 0 is the foot of the vertical of the monitoring camera on the ground where the substation is located. RQ is the intrusion height h of the intrusion target. Point R is the upper end of the intrusion target. Point Q is the foot of the vertical on the plane where the intrusion target is located at the substation. Connect the upper end C of the monitoring camera with the upper end R of the intrusion target, and make an extension line of the upper end of the monitoring camera and the upper end of the intrusion target until they intersect with the ground where the substation is located. The intersection is point M.

[0129] The coordinates of point O are known to be (x1, y1) and the angle α, where the angle α is the depression angle of the surveillance camera; the depression angle can change as the intrusion target moves;

[0130] According to the geometric relationship, triangle OCM is similar to triangle QRM. The angle α1 of triangle QRM is equal to the angle α. α1 is also the depression angle of the surveillance camera. The length of OQ is calculated by the formula. The formula is as follows:

[0131] OQ = MO-MQ;

[0132] That is, OQ = H / tanα-h / tanα;

[0133] It is known that the offset angle of the surveillance camera is β, and point P is the center point of the core area, so the angle of the angle QOP is 180°-β; in the initial state, the shooting angle of the surveillance camera is straight ahead. When an intrusion target is detected, the shooting angle of the surveillance camera moves, and then the offset angle is obtained;

[0134] Therefore, the radius OP of the core region is L and the length of OQ is H / tanα-h / tanα;

[0135] Therefore, the length of QP is:

[0136]

[0137] Since the radius of the core area is known, the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the current moment is:

[0138] Real-time distance = QP-L;

[0139] If the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the current moment is greater than or equal to the safe distance, the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the next moment is calculated;

[0140] If the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the current moment is less than the safety distance, the time when the intrusion target enters the screen is recorded, and the residence time of the intrusion target is obtained by subtracting the entry time from the current time. When the residence time exceeds the duration threshold, a safety warning signal is generated, and the current monitoring screen is recorded as an abnormal screen. When the residence time does not exceed the duration threshold, no operation is performed;

[0141] The intelligent distance measurement module sends the safety warning signal and the corresponding abnormal picture to the alarm display terminal, and the alarm display terminal is used to display the safety warning signal and the abnormal picture of the substation as an alarm.

[0142] Example 2: Figure 7 As shown, based on another concept of the same invention, a substation safety management method based on embedded devices is now proposed, and the method is specifically as follows:

[0143] Step S100, dividing the corresponding supervision area of ​​the substation, optimizing the supervision configuration in the peripheral area and the core area according to the historical intrusion data, and obtaining the supervision plan of the peripheral area or the core area;

[0144] Step S200, after executing the corresponding supervision scheme in the core area and the peripheral area, real-time monitoring video data of the corresponding supervision area of ​​the substation is collected in real time;

[0145] Step S300, constructing a target intrusion detection algorithm large model used in the data frame diagram, processing the real-time monitoring video data of the corresponding supervision area of ​​the substation in combination with the target intrusion detection algorithm large model, generating a ranging signal or performing no operation;

[0146] Step S400: If a distance measurement signal is generated, the distance between the intrusion target and the substation within the supervision range is measured; if a safety warning signal and a corresponding abnormal screen are generated, an alarm is displayed.

[0147] In this application, if corresponding calculation formulas appear, the above calculation formulas are all dimensionless and take their numerical calculations. The weight coefficients, proportional coefficients and other coefficients in the formulas are set to a result value obtained by quantifying each parameter. The size of the weight coefficient and the proportional coefficient can be determined as long as it does not affect the proportional relationship between the parameter and the result value.

[0148] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. Substation safety management system based on embedded devices, characterized by: It includes area division module, database, configuration optimization module, data acquisition module, video processing module, intelligent distance measurement module, model building module and alarm display terminal; The area division module is used to divide the supervision area corresponding to the substation into a core area and a peripheral area; The database is used to record the historical intrusion data of different peripheral areas and different core areas, the historical intrusion data is the number of foreign object intrusions in different peripheral areas and different core areas, and the intrusion residence time and intrusion distance of each foreign object intrusion, the intrusion distance is the interval distance from the substation after entering the supervision area, and the configuration optimization module is used to optimize the supervision configuration in the peripheral area and the core area to obtain the supervision plan corresponding to the peripheral area or the core area; The optimization process of the configuration optimization module is as follows: Get the number of foreign objects intrusion into the peripheral area and the core area RCu, where u is the number of the peripheral area and the core area; Then, the invasion residence time and invasion distance of each foreign object invasion in the peripheral area and the core area are obtained, and the invasion residence time of each foreign object invasion is traversed and compared, and the maximum invasion residence time is taken as the foreign object invasion time RTu of the peripheral area and the core area. At the same time, the invasion distance of each foreign object invasion is traversed and compared, and the minimum invasion distance is taken as the foreign object invasion distance RJu of the peripheral area and the core area; The regional intrusion values ​​QRu of the peripheral area and the core area are calculated by the formula. The specific formula is as follows: QRu=(RCu×a1+RTu×a2+1 / RJu)×λ; where a1 and a2 are weight coefficients with fixed values, a2>a1, λ is the safety warning coefficient, and the safety warning coefficient of the core area is greater than that of the peripheral area; The regional intrusion value is compared with the regional intrusion threshold, where the regional intrusion threshold includes a first regional intrusion threshold and a second regional intrusion threshold, and the second regional intrusion threshold is greater than the first regional intrusion threshold; If the regional intrusion value is less than the first regional intrusion threshold, the peripheral area or the core area implements the third supervision scheme; If the regional intrusion value is greater than or equal to the first regional intrusion threshold and less than the second regional intrusion threshold, the peripheral area or the core area implements the second supervision plan; If the regional intrusion value is greater than or equal to the second regional intrusion threshold, the peripheral area or the core area implements the first supervision plan; After the corresponding supervision scheme is implemented in the core area and the peripheral area, the data acquisition module is used to collect the real-time monitoring video data of the corresponding supervision area of ​​the substation in real time and send it to the video processing module; the model construction module is used to construct the target intrusion detection algorithm large model used in the data frame diagram, and send the target intrusion detection algorithm large model to the video processing module; The video processing module is used to process the real-time monitoring video data of the supervision area corresponding to the substation, generate a distance measurement signal or do not perform any operation, and if a distance measurement signal is generated, it is sent to the intelligent distance measurement module; the intelligent distance measurement module is used to measure the distance between the intrusion target and the substation within the supervision range, do not perform any operation or generate a safety warning signal, and if a safety warning signal is generated and the corresponding abnormal picture is sent to the alarm display terminal, the alarm display terminal is used to alarm the safety warning signal and abnormal picture of the substation; The processing process of the video processing module is as follows: Obtain the real-time monitoring video data of the corresponding supervision area of ​​the substation, and extract the real-time monitoring video data frame by frame to obtain a frame-by-frame image; Performing grayscale processing on each frame of the image to obtain a first data frame image; Then, the first data frame is subjected to denoising to obtain a second data frame, and the target intrusion detection algorithm model is used to perform intrusion target detection on the second data frame; If an intruder is detected in any core area or peripheral area, a ranging signal is generated; If no intrusion target is detected in all core areas and peripheral areas, no operation will be performed; The working process of the intelligent ranging module is as follows: Obtain monitoring images in real-time monitoring video data based on the current moment; A rectangular coordinate system corresponding to the monitoring screen is established with the origin of the monitoring camera, the vertical direction of the monitoring camera as the Y axis, and the connecting line between the lower end of the monitoring camera and the lower end of the intrusion target as the X axis; With O as the origin, CO as the Y-axis, and OM as the X-axis, a rectangular coordinate system corresponding to the monitoring screen is established. CO is the height of the monitoring camera. The monitoring camera is vertically set at the substation. CO is the height H of the monitoring camera. Point C is the upper end of the monitoring camera. Point 0 is the foot of the vertical line of the monitoring camera on the ground where the substation is located. RQ is the intrusion height h of the intrusion target. Point R is the upper end of the intrusion target. Point Q is the foot of the vertical line of the intrusion target on the plane where the substation is located. Connect the upper end C of the monitoring camera with the upper end R of the intrusion target, and make an extension line of the upper end of the monitoring camera and the upper end of the intrusion target until they intersect with the ground where the substation is located. The intersection point is point M. The coordinates of point O are known to be (x1, y1) and the angle α, where the angle α is the depression angle of the surveillance camera; The length of OQ is calculated by the formula, which is as follows: OQ = MO-MQ; That is, OQ = H / tanα-h / tanα; It is known that the offset angle of the surveillance camera is β, and point P is the center point of the core area, so the angle of the angle QOP is 180°-β; Therefore, the radius OP of the core region is L and the length of OQ is H / tanα-h / tanα; The length of QP is: α1 is the depression angle of the surveillance camera, and α1=α; Since the radius of the core area is known, the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the current moment is: Real-time distance = QP-L; If the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the current moment is greater than or equal to the safe distance, the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the next moment is calculated; If the real-time distance between the intrusion target and the core area in the corresponding monitoring screen at the current moment is less than the safety distance, the time when the intrusion target enters the screen is recorded, and the residence time of the intrusion target is obtained by subtracting the entry time from the current time. When the residence time exceeds the duration threshold, a security warning signal is generated, and the current monitoring screen is recorded as an abnormal screen. When the residence time does not exceed the duration threshold, no operation is performed.

2. The substation safety management system based on embedded devices according to claim 1 is characterized in that: The division process of the core area and the peripheral area is as follows: Set warning distances within the supervision area; Expand outward from the substation as the center until the expansion reaches the warning line where the warning distance is located; The cordon surrounds the substation. The supervision area surrounded by the cordon is the core area, and the supervision area excluding the core area is the peripheral area. The core area and the peripheral area are both rectangular. Connect the right angle of the peripheral area with the right angle of the core area on the same side through a straight line, and extend the straight line to the substation. The straight line is used as a set of dividing lines of the regulatory area. Similarly, three additional sets of dividing lines of the regulatory area are obtained. The peripheral area and the core area are divided by a dividing line, and the peripheral areas are numbered in a clockwise direction to obtain a first peripheral area, a second peripheral area, a third peripheral area and a fourth peripheral area. Similarly, a first core area, a second core area, a third core area and a fourth core area are obtained.

3. The substation safety management system based on embedded devices according to claim 1 is characterized in that: The first monitoring solution is: using multiple high-configuration surveillance cameras; The second monitoring solution is: using multiple surveillance cameras with common configuration; The third monitoring solution is to use a single-channel surveillance camera with ordinary configuration.

4. The substation safety management system based on embedded devices according to claim 3 is characterized in that: The regulatory intensity of the first regulatory scheme is higher than that of the second regulatory scheme, and the regulatory intensity of the second regulatory scheme is higher than that of the third regulatory scheme.

5. A substation safety management method based on embedded devices, characterized in that: Using the substation safety management system based on embedded devices as described in any one of claims 1 to 4, the method is as follows: Step S100, dividing the corresponding supervision area of ​​the substation, optimizing the supervision configuration in the peripheral area and the core area according to the historical intrusion data, and obtaining the supervision plan of the peripheral area or the core area; Step S200, after executing the corresponding supervision scheme in the core area and the peripheral area, real-time monitoring video data of the corresponding supervision area of ​​the substation is collected in real time; Step S300, constructing a target intrusion detection algorithm large model used in the data frame diagram, processing the real-time monitoring video data of the corresponding supervision area of ​​the substation in combination with the target intrusion detection algorithm large model, generating a ranging signal or performing no operation; Step S400: If a distance measurement signal is generated, the distance between the intrusion target and the substation within the supervision range is measured; if a safety warning signal and a corresponding abnormal screen are generated, an alarm is displayed.

Images