A power system loss reduction regulation result visualization method and system

Abstract

The application discloses a power system loss reduction regulation result visualization method and system, comprising: performing abnormal diagnosis on line loss of a power system according to operation data of the power system; when the diagnosis result is abnormal, performing loss reduction analysis on the operation data, and performing loss reduction regulation on the power system based on a loss reduction strategy obtained through the analysis; modeling and processing a display model of the loss reduction regulation result; measuring current display brightness of the display model, current illumination intensity of an environment where the display model is located and a scene picture in a display range in real time, then adjusting display brightness of the display model according to a difference between the current illumination intensity and the current display brightness, and extracting feature information of a head of a person in the scene picture to adjust a display angle of the display model. According to the illumination intensity of the environment where the display model is located and the feature information of the head of the person in the scene picture in the display range obtained in real time, the application adjusts the display brightness and the angle of the display model, so that the visual effect of different users watching the model can be adaptively met.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of electric power data analysis processing, and in particular to a power system loss reduction regulation result visualization method and system. BACKGROUND

[0002] With the continuous development of the national economy, the load of the distribution network gradually increases. In the past, people tend to focus on the safety and reliability of the distribution network system, ignoring the economy, that is, lacking a unified power grid planning, so that the operation of the distribution network system is not economical. Among them, the power loss of the distribution network system is the power loss of each power transmission and transformation element and each transmission link in the entire power transmission process. Since the operation economy of the distribution network is an important guarantee for realizing energy saving and loss reduction of the power industry and an important means for power enterprises to improve competitiveness, in order to fully master the economy of the distribution network operation and guide the operation, construction and transformation of the power grid, it is necessary to reduce the loss of the distribution network.

[0003] The existing visualization distribution network loss reduction system mainly performs abnormal diagnosis and formulates loss reduction strategies after line loss monitoring and calculation of the power system, and finally performs loss reduction regulation according to the feasible loss reduction strategies, so as to realize accurate loss reduction regulation. However, the regulation result is in the form of data report, which needs to spend time to read the data and understand the regulation result, thereby reducing the reading efficiency of the regulation result. SUMMARY

[0004] The embodiment of the present application provides a power system loss reduction regulation result visualization method and system, which adjusts the display brightness and display angle of the display model by using the display brightness of the display model, the light intensity of the environment and the feature information of the head of the person in the scene picture of the display range collected in real time, so as to adaptively meet the visual perception of different users watching the display model.

[0005] In order to solve the above technical problems, the embodiment of the present application provides a power system loss reduction regulation result visualization method, which comprises:

[0006] According to the operation data of the power system, the line loss of the power system is abnormally diagnosed to obtain a diagnosis result;

[0007] When the diagnosis result is abnormal, the operation data is analyzed for loss reduction, and the power system is regulated for loss reduction based on the loss reduction strategy obtained by the analysis to obtain a regulation result;

[0008] The regulation result is modeled to obtain a display model;

[0009] real-time measure the current display brightness of the display model and the current illumination intensity of the environment where the display model is located, and adjust the display brightness of the display model according to the difference between the current illumination intensity and the current display brightness;

[0010] real-time collect a scene picture of the display model in a display range, then extract features from the scene picture, and adjust the display angle of the display model according to the position information of the first head feature extracted.

[0011] The operation data is obtained by real-time monitoring of line loss of the power system, and the first head feature is feature information of a head of a person in the scene picture.

[0012] According to the operation data obtained by real-time monitoring of line loss of the power system, the embodiment of the application performs abnormal diagnosis on the power system, performs loss reduction analysis on the operation data when the diagnosis result is abnormal, performs loss reduction regulation on the power system based on the loss reduction strategy obtained by the analysis to obtain a regulation result, models the regulation result to obtain a display model, and realizes visualization of the regulation result, so that managers or other personnel can intuitively read the regulation result of the power system. In addition, the current display brightness of the display model and the current illumination intensity of the environment where the display model is located are measured in real time, the display brightness of the display model is adjusted according to the difference between the current illumination intensity and the current display brightness, a scene picture of the display model in a display range is collected in real time, features of the scene picture are extracted, and the display angle of the display model is adjusted according to the position information of the first head feature extracted, so as to provide a display model with suitable display brightness and display angle for a user watching the display model, so as to meet the needs of different users watching the display model.

[0013] As a preferred solution, the real-time collection of the scene picture of the display model in the display range, the feature extraction from the scene picture, and the adjustment of the display angle of the display model according to the position information of the first head feature extracted are specifically as follows:

[0014] The scene picture of the display model in the display range is collected in real time, and features of the scene picture are extracted to obtain the corresponding first head feature.

[0015] The head position information corresponding to the first head feature is compared with the position information of the center point of the scene picture to obtain a comparison difference, and the display angle of the display model is adjusted until the head position of the first head feature is at the center point of the scene picture.

[0016] According to the preferred solution of the embodiment of the present application, the scene picture of the display model in the display range is collected in real time, feature extraction is performed on the scene picture to obtain the corresponding first head feature, then the head position information corresponding to the first head feature is compared with the position information of the center point of the scene picture, and the display angle of the display model is controlled and adjusted according to the comparison difference obtained through the comparison, until the head position of the first head feature is at the center point of the scene picture, so that the visual center of the user watching the display model and the center of the display model are as close as possible, thereby improving the user's visual experience.

[0017] As a preferred solution, the feature extraction performed on the scene picture to obtain the corresponding first head feature is specifically as follows:

[0018] The pre-constructed neural network model is trained and verified using the first sample data set to obtain a feature extraction model.

[0019] The format of the scene picture is adjusted based on the input format requirement of the feature extraction model, and the scene picture after the format adjustment is input into the feature extraction model, so that the feature extraction model performs feature extraction on the scene picture after the format adjustment and finally outputs the first head feature corresponding to the scene picture.

[0020] The first sample data set includes a plurality of to-be-processed pictures and second head features corresponding to each to-be-processed picture, and the to-be-processed picture is an image containing a human face.

[0021] According to the preferred solution of the embodiment of the present application, the pre-constructed neural network model is trained and verified using the first sample data set including a plurality of to-be-processed pictures and second head features corresponding to each to-be-processed picture to obtain a feature extraction model, so as to improve the head feature extraction performance of the feature extraction model on images, and then the format of the scene picture is adjusted based on the input format requirement of the feature extraction model, so that the feature extraction model can better extract the first head feature corresponding to the scene picture.

[0022] As a preferred solution, the current display brightness of the display model and the current illumination intensity of the environment where the display model is located are measured in real time, and the display brightness of the display model is adjusted according to the difference between the current illumination intensity and the current display brightness, specifically as follows:

[0023] The current display brightness of the display model and the current illumination intensity of the environment where the display model is located are measured in real time, and the current illumination intensity is subtracted from the current display brightness to obtain a corresponding brightness difference.

[0024] When the luminance difference value is not zero, the display luminance of the display model is controlled to be adjusted until the display luminance of the display model reaches the target luminance.

[0025] According to a preferred scheme of the embodiment of the present application, the current display luminance of the display model and the current illumination intensity of the environment where the display model is located are measured in real time, the current illumination intensity is subtracted from the current display luminance to obtain a corresponding luminance difference value, and then when the luminance difference value is not zero, the display luminance of the display model is controlled to be adjusted until the display luminance of the display model reaches the target luminance, so as to protect the vision of the user who is watching the display model and save electric energy.

[0026] As a preferred scheme, when the diagnostic result is abnormal, loss reduction analysis is performed on the operation data, and based on a loss reduction strategy obtained through the analysis, loss reduction regulation is performed on the power system to obtain a regulation result, specifically:

[0027] When the diagnostic result is abnormal, a plurality of line loss influencing factors are extracted from the operation data, and a high-dimensional random matrix is constructed according to all the line loss influencing factors;

[0028] The high-dimensional random matrix is subjected to feature extraction to obtain a high-dimensional random matrix feature for representing the operation state of the power distribution network, and then the high-dimensional random matrix feature is input into a loss reduction analysis model, so that the loss reduction analysis model outputs a corresponding loss reduction strategy;

[0029] The loss reduction strategy is used to perform loss reduction regulation on the power system to obtain the regulation result.

[0030] According to a preferred scheme of the embodiment of the present application, when the diagnostic result is abnormal, a plurality of line loss influencing factors are extracted from the operation data, and a high-dimensional random matrix is constructed according to all the line loss influencing factors, then the high-dimensional random matrix is subjected to feature extraction to obtain a high-dimensional random matrix feature for representing the operation state of the power distribution network, and then the high-dimensional random matrix feature is input into a loss reduction analysis model, so that the loss reduction analysis model outputs an effective loss reduction strategy for the operation state of the power distribution network, and the loss reduction strategy is used to perform loss reduction regulation on the power system, thereby improving the loss reduction regulation effect on the power system.

[0031] As a preferred scheme, the line loss of the power system is subjected to abnormal diagnosis according to the operation data of the power system to obtain a diagnostic result, specifically:

[0032] The operation data is subjected to feature extraction to obtain key features in the operation data;

[0033] The key features are adjusted in format based on input format requirements of the anomaly diagnosis model, to obtain corresponding input features, and the input features are input into the anomaly diagnosis model, so that the anomaly diagnosis model performs anomaly diagnosis on the input features and finally outputs corresponding diagnosis results.

[0034] The anomaly diagnosis model is obtained by training a pre-constructed learning vector quantization network.

[0035] According to the preferred scheme of the embodiment of the present application, the key features in the operation data are extracted, then the key features are adjusted in format based on input format requirements of the anomaly diagnosis model, to obtain corresponding input features, and the input features are input into the anomaly diagnosis model, so as to improve the anomaly diagnosis accuracy of the anomaly diagnosis model. In addition, the anomaly diagnosis model is obtained by training a pre-constructed learning vector quantization network, and the learning vector quantization network can be trained by randomly selecting suspended training sample data, and the use of the learning vector quantization network can reduce the memory requirement for storing the entire training data set.

[0036] As a preferred scheme, the anomaly diagnosis model is obtained by:

[0037] A learning vector quantization network is pre-constructed, and a second sample data set is divided into a training set and a validation set.

[0038] The learning vector quantization network is iteratively trained and validated, and in each iteration, the current training set is used to train the current learning vector quantization network to obtain a pre-training model, and then the current validation set is used to validate the current pre-training model to obtain the diagnosis accuracy of the current pre-training model, and then the sample data in the current training set is updated, and the current pre-training model is used as the learning vector quantization network, until the iteration number reaches a preset number, and the iteration is stopped, and the pre-training model with the highest diagnosis accuracy in the iteration process is used as the anomaly diagnosis model.

[0039] The second sample data set includes a plurality of sample data, and the sample data is obtained by preprocessing historical operation data of line loss anomalies.

[0040] According to the preferred scheme of the embodiment of the present application, the historical line loss anomaly problem and the processing work order are preprocessed to obtain sample data, and a second sample data set including a plurality of sample data is divided into a training set and a validation set, and the training set and the validation set are used to iteratively train and validate a pre-constructed learning vector quantization network, so as to optimize the anomaly diagnosis accuracy and efficiency of the anomaly diagnosis model.

[0041] To solve the same technical problems, the embodiment of the present application also provides a power system loss reduction regulation result visualization system, comprising:

[0042] An abnormality diagnosis module is configured to perform abnormality diagnosis on line loss of the power system according to operation data of the power system, and obtain a diagnosis result; wherein the operation data is obtained by real-time monitoring of the line loss of the power system;

[0043] A loss reduction regulation module is configured to perform loss reduction analysis on the operation data when the diagnosis result is abnormal, and perform loss reduction regulation on the power system based on a loss reduction strategy obtained by the analysis, and obtain a regulation result;

[0044] A modeling module is configured to perform modeling processing on the regulation result, and obtain a display model;

[0045] A brightness adjustment module is configured to measure current display brightness of the display model and current illumination intensity of an environment where the display model is located in real time, and adjust display brightness of the display model according to a difference between the current illumination intensity and the current display brightness;

[0046] An angle adjustment module is configured to collect a scene picture of the display model in a display range in real time, then perform feature extraction on the scene picture, and adjust a display angle of the display model according to position information of a first head feature extracted; wherein the first head feature is feature information of a head of a person in the scene picture.

[0047] As a preferred scheme, the angle adjustment module specifically comprises:

[0048] A feature extraction unit is configured to collect the scene picture of the display model in the display range in real time, and perform feature extraction on the scene picture, and obtain the first head feature corresponding thereto;

[0049] An angle adjustment unit is configured to compare head position information corresponding to the first head feature with position information of a center point of the scene picture, obtain a comparison difference, and control adjustment of the display angle of the display model according to the comparison difference, until the head position of the first head feature is at the center point of the scene picture.

[0050] As a preferred scheme, the loss reduction regulation module specifically comprises:

[0051] A matrix construction unit is configured to extract a plurality of line loss influence factors from the operation data when the diagnosis result is abnormal, and construct a high-dimensional random matrix according to all the line loss influence factors;

[0052] The loss reduction analysis unit is configured to perform feature extraction on the high-dimensional random matrix to obtain a high-dimensional random matrix feature representing an operating state of the power distribution network, and then input the high-dimensional random matrix feature into a loss reduction analysis model to cause the loss reduction analysis model to output a corresponding loss reduction strategy.

[0053] The loss reduction regulation unit is configured to use the loss reduction strategy to perform loss reduction regulation on the power system to obtain the regulation result. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 FIG. 1 is a flowchart of a power system loss reduction regulation result visualization method according to an embodiment of the present application.

[0055] Figure 2 FIG. 2 is a structural diagram of a power system loss reduction regulation result visualization system according to an embodiment of the present application. DETAILED DESCRIPTION

[0056] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0057] Embodiment One:

[0058] Please refer to Figure 1 A power system loss reduction regulation result visualization method according to an embodiment of the present application includes steps S1 to S3, and each step is specifically as follows.

[0059] S1. According to the operating data of the power system, the line loss of the power system is diagnosed to obtain a diagnosis result.

[0060] The operating data is obtained by real-time monitoring of the line loss of the power distribution network.

[0061] As a preferred solution, step S1 includes steps S11 to S12, and each step is specifically as follows.

[0062] S11. The operating data is subjected to feature extraction to obtain key features in the operating data, so as to improve the effectiveness of the data input into the abnormal diagnosis model.

[0063] In step S12, the key features are adjusted in format based on the input format requirement of the anomaly diagnosis model, to obtain corresponding input features, and the input features are input into the anomaly diagnosis model, so that the anomaly diagnosis model performs anomaly diagnosis on the input features and finally outputs corresponding diagnosis results.

[0064] It should be noted that by adjusting the key features in format, the processing fluency of the anomaly diagnosis model on the input data can be improved, thereby improving the diagnosis efficiency of the anomaly diagnosis model.

[0065] The anomaly diagnosis model is obtained by training a pre-constructed learning vector quantization network, and the obtaining process of the anomaly diagnosis model includes steps S01 to S02, which are as follows.

[0066] In step S01, a learning vector quantization network is pre-constructed, and a second sample data set is divided into a training set and a validation set.

[0067] The second sample data set includes a plurality of sample data, which is obtained by preprocessing historical operation data of line loss anomaly.

[0068] It should be noted that the historical operation data of line loss anomaly can be selected as recent line loss anomaly work sheets, such as line loss anomaly work sheets in the last 3 months. The preprocessing includes but is not limited to data cleaning and phenomenon feature labeling.

[0069] In step S02, the learning vector quantization network is iteratively trained and verified. In each iteration, the current training set is used to train the current learning vector quantization network to obtain a pre-training model, and then the current validation set is used to verify the current pre-training model to obtain the diagnosis accuracy of the current pre-training model. Then, the sample data in the current training set is updated, and the current pre-training model is used as the learning vector quantization network. The iteration is stopped when the number of iterations reaches a preset number, and the pre-training model with the highest diagnosis accuracy in the iteration process is used as the anomaly diagnosis model.

[0070] In step S2, when the diagnosis result is abnormal, loss reduction analysis is performed on the operation data, and based on the loss reduction strategy obtained by the analysis, loss reduction control is performed on the power system to obtain a control result.

[0071] As a preferred solution, step S2 includes steps S21 to S23, which are as follows.

[0072] In step S21, when the diagnosis result is abnormal, a plurality of line loss influencing factors are extracted from the operation data according to a preset period, and a high-dimensional random matrix is constructed based on all line loss influencing factors.

[0073] Step S22, feature extraction is performed on the high-dimensional random matrix to obtain high-dimensional random matrix features for representing the operation state of the power distribution network, and then the high-dimensional random matrix features are input into the loss reduction analysis model to enable the loss reduction analysis model to output a corresponding loss reduction strategy.

[0074] Step S23, using the loss reduction strategy, loss reduction regulation is performed on the power system to obtain a regulation result.

[0075] Step S3, modeling processing is performed on the regulation result to obtain a display model.

[0076] It should be noted that the modeling processing includes but is not limited to the following technical means: generating a line loss column chart, a load density distribution chart, a high temperature distribution chart, a rainfall distribution chart and a lightning distribution chart in the form of a GIS graph. By using the above technical means, the failure caused by the power intensive area, high temperature, lightning and heavy rain is intuitively displayed, and the problem of reducing the reading efficiency of the regulation result by displaying the regulation result in the form of data report in the existing power distribution network system is solved.

[0077] Step S4, the current display brightness of the display model and the current illumination intensity of the environment where the display model is located are measured in real time, and the display brightness of the display model is adjusted according to the difference between the current illumination intensity and the current display brightness.

[0078] As a preferred scheme, step S4 includes steps S41 to S42, and each step is specifically as follows:

[0079] Step S41, the current display brightness of the display model and the current illumination intensity of the environment where the display model is located are measured in real time, and the current illumination intensity is subtracted from the current display brightness to obtain a corresponding brightness difference.

[0080] Step S42, when the brightness difference is not zero, the display brightness of the display model is controlled to be adjusted until the display brightness of the display model reaches a target brightness, so that the current display brightness of the display model is more suitable for users to watch.

[0081] In this embodiment, when the brightness difference is positive, it indicates that the display model is in a strong light environment, and at this time the display brightness of the display model is controlled to be brightened (increased) until the display brightness of the display model reaches the target brightness; when the brightness difference is negative, it indicates that the display model is in a weak light environment, and at this time the display brightness of the display model is controlled to be dimmed (decreased) until the display brightness of the display model reaches the target brightness.

[0082] It should be noted that the target brightness can be equal to the current light intensity of the environment in which the display model is located, so as to protect the vision of the user who is watching the display model, or can be set by the user according to the user's own needs, such as setting the target brightness as the current light intensity + X (X>0), so as to improve the clarity of the display model.

[0083] In step S5, the scene picture of the display model in the display range is collected in real time, and then the scene picture is feature extracted, and the display angle of the display model is adjusted according to the position information of the first head feature extracted.

[0084] It should be noted that the light compensation when collecting the scene picture can improve the clarity of the collected scene picture.

[0085] As a preferred solution, step S5 includes steps S51 to S53, and each step is specifically as follows:

[0086] In step S51, the scene picture of the display model in the display range is collected in real time.

[0087] In step S52, the scene picture is feature extracted to obtain the corresponding first head feature.

[0088] The first head feature is the feature information of the head of the personnel in the scene picture.

[0089] As a preferred solution, step S52 includes steps S521 to S522, and each step is specifically as follows:

[0090] In step S521, the pre-constructed neural network model is trained and verified using the first sample data set to obtain a feature extraction model, so as to improve the feature extraction accuracy of the feature extraction model.

[0091] The first sample data set includes a plurality of to-be-processed pictures and second head features corresponding to each to-be-processed picture, and the to-be-processed picture is an image containing a face.

[0092] In step S522, the format of the scene picture is adjusted based on the input format requirement of the feature extraction model, and the scene picture after format adjustment is input into the feature extraction model, so that the feature extraction model performs feature extraction on the scene picture after format adjustment and finally outputs the first head feature corresponding to the scene picture.

[0093] Step S53, comparing the head position information corresponding to the first head feature with the position information of the center point of the scene picture to obtain a comparison difference value, and controlling the display angle of the display model to be adjusted until the head position of the first head feature is at the center point of the scene picture, so that the visual center of the user watching the display model and the center of the display model are as close as possible, thereby improving the user's visual experience and enabling the user to clearly watch the complete display model.

[0094] Please refer to Figure 2 A structure schematic diagram of a power system loss reduction regulation result visualization system provided by the embodiment of the present application is provided, the system comprises an abnormality diagnosis module M1, a loss reduction regulation module M2, a modeling module M3, a brightness adjustment module M4 and an angle adjustment module M5, and each module is specifically as follows:

[0095] The abnormality diagnosis module M1 is used for performing abnormality diagnosis on the line loss of the power system according to the operation data of the power system to obtain a diagnosis result, wherein the operation data is obtained by real-time monitoring on the line loss of the power system;

[0096] The loss reduction regulation module M2 is used for performing loss reduction analysis on the operation data when the diagnosis result is abnormal, and performing loss reduction regulation on the power system based on the loss reduction strategy obtained by the analysis to obtain a regulation result;

[0097] The modeling module M3 is used for performing modeling processing on the regulation result to obtain a display model;

[0098] The brightness adjustment module M4 is used for measuring the current display brightness of the display model and the current illumination intensity of the environment where the display model is located in real time, and adjusting the display brightness of the display model according to the difference between the current illumination intensity and the current display brightness;

[0099] The angle adjustment module M5 is used for collecting a scene picture of the display model in a display range in real time, then performing feature extraction on the scene picture, and adjusting the display angle of the display model according to the position information of a first head feature obtained by the extraction, wherein the first head feature is the feature information of the head of a person in the scene picture.

[0100] As a preferred scheme, the angle adjustment module M5 specifically comprises a feature extraction unit 51 and an angle adjustment unit 52, and each unit is specifically as follows:

[0101] The feature extraction unit 51 is used for collecting a scene picture of the display model in a display range in real time, and performing feature extraction on the scene picture to obtain a corresponding first head feature;

[0102] The angle adjusting unit 52 is used for comparing the head position information corresponding to the first head feature with the position information of the center point of the scene picture to obtain a comparison difference value, and controlling the display angle of the display model according to the comparison difference value until the head position of the first head feature is at the center point of the scene picture.

[0103] As a preferred solution, the loss reduction control module M2 specifically comprises a matrix construction unit 21, a loss reduction analysis unit 22 and a loss reduction control unit 23, and each unit is specifically as follows:

[0104] The matrix construction unit 21 is used for extracting a plurality of line loss influencing factors from the operation data when the diagnosis result is abnormal, and constructing a high-dimensional random matrix according to all the line loss influencing factors;

[0105] The loss reduction analysis unit 22 is used for performing feature extraction on the high-dimensional random matrix to obtain high-dimensional random matrix features for representing the operation state of the power distribution network, and then inputting the high-dimensional random matrix features into a loss reduction analysis model to make the loss reduction analysis model output a corresponding loss reduction strategy;

[0106] The loss reduction control unit 23 is used for using the loss reduction strategy to perform loss reduction control on the power system to obtain a control result.

[0107] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system described above can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0108] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0109] The present application provides a power system loss reduction control result visualization method and system, according to the operation data obtained by real-time monitoring of the line loss of the power system, performing abnormal diagnosis on the power system, when the diagnosis result is abnormal, performing loss reduction analysis on the operation data, and based on the loss reduction strategy obtained by the analysis, performing loss reduction control on the power system to obtain a control result, then modeling the control result to obtain a display model, realizing the visualization of the control result, so that the manager or other personnel can intuitively read the control result of the power system. In addition, the current display brightness of the display model and the current illumination intensity of the environment where the display model is located are measured in real time, and according to the difference between the current illumination intensity and the current display brightness, the display brightness of the display model is adjusted, and the scene picture of the display model in the display range is collected in real time, then the scene picture is feature extracted, and according to the position information of the first head feature obtained by the extraction, the display angle of the display model is adjusted, so as to provide a display model with suitable display brightness and display angle for the user who is watching the display model, so as to meet the needs of different users who watch the display model.

[0110] Further, when the diagnosis result is abnormal, a plurality of line loss influencing factors are extracted from the operation data, a high-dimensional random matrix is constructed according to all the line loss influencing factors, then feature extraction is performed on the high-dimensional random matrix to obtain high-dimensional random matrix features for representing the operation state of the power distribution network, and then the high-dimensional random matrix features are input into the loss reduction analysis model, so that the loss reduction analysis model outputs an effective loss reduction strategy for the operation state of the power distribution network, and the loss reduction strategy is used to perform loss reduction regulation and control on the power system, thereby improving the loss reduction regulation and control effect on the power system.

[0111] The above specific embodiments further specifically describe the purposes, technical solutions and beneficial effects of the present application. It should be understood that the above description is only for specific embodiments of the present application and is not intended to limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for visualizing the results of power system loss reduction control, characterized in that, include: Based on the power system's operating data, anomaly diagnosis is performed on the power system's line losses to obtain diagnostic results. Specifically, feature extraction is performed on the operating data to obtain key features in the operating data. Based on the input format requirements of the anomaly diagnosis model, the key features are formatted to obtain corresponding input features, which are then input into the anomaly diagnosis model so that the model can perform anomaly diagnosis on the input features and finally output the corresponding diagnosis result; wherein, the anomaly diagnosis model is obtained by training a pre-constructed learning vector quantization network; When the diagnostic result is abnormal, loss reduction analysis is performed on the operating data, and loss reduction regulation is carried out on the power system based on the loss reduction strategy obtained from the analysis to obtain the regulation result; The regulation results are modeled to obtain a display model; The current display brightness of the display model and the current light intensity of the environment in which the display model is located are measured in real time, and the display brightness of the display model is adjusted according to the difference between the current light intensity and the current display brightness; The scene image of the display model within the display range is acquired in real time, then the scene image is processed for feature extraction, and the display angle of the display model is adjusted according to the position information of the extracted first head feature. The operational data is obtained by real-time monitoring of the line loss of the power system, and the first head feature is the feature information of the person's head in the scene image.

2. The method for visualizing power system loss reduction control results according to claim 1, characterized in that, The process involves real-time acquisition of scene images of the display model within the display range, extraction of features from the scene images, and adjustment of the display angle of the display model based on the extracted first head feature position information. Specifically: The scene images of the display model within the display range are acquired in real time, and feature extraction is performed on the scene images to obtain the corresponding first head features; The head position information corresponding to the first head feature is compared with the position information of the center point of the scene image to obtain a comparison difference. Based on the comparison difference, the display angle of the display model is controlled and adjusted until the head position of the first head feature is at the center point of the scene image.

3. The method for visualizing power system loss reduction control results according to claim 2, characterized in that, The step of extracting features from the scene image to obtain the corresponding first head feature specifically involves: Using the first sample dataset, the pre-built neural network model is trained and validated to obtain the feature extraction model; Based on the input format requirements of the feature extraction model, the format of the scene image is adjusted, and the scene image with the adjusted format is input into the feature extraction model, so that the feature extraction model extracts features from the scene image with the adjusted format and finally outputs the first head feature corresponding to the scene image. The first sample dataset includes several images to be processed and the second head features corresponding to each image to be processed, wherein the images to be processed are images containing human faces.

4. The method for visualizing power system loss reduction control results according to claim 1, characterized in that, The real-time measurement of the current display brightness of the display model and the current illuminance of the environment in which the display model is located, and the adjustment of the display brightness of the display model based on the difference between the current illuminance and the current display brightness, specifically involves: The current display brightness of the display model and the current illumination intensity of the environment in which the display model is located are measured in real time, and the current illumination intensity is subtracted from the current display brightness to obtain the corresponding brightness difference value; When the brightness difference is not zero, the display brightness of the display model is adjusted until the display brightness of the display model reaches the target brightness.

5. The method for visualizing power system loss reduction control results according to claim 1, characterized in that, When the diagnostic result is abnormal, loss reduction analysis is performed on the operating data, and based on the loss reduction strategy obtained from the analysis, loss reduction regulation is implemented on the power system to obtain the regulation result, specifically as follows: When the diagnostic result is abnormal, multiple line loss influencing factors are extracted from the operational data, and a high-dimensional random matrix is ​​constructed based on all the line loss influencing factors. Feature extraction is performed on the high-dimensional random matrix to obtain high-dimensional random matrix features that characterize the operating state of the distribution network. Then, the high-dimensional random matrix features are input into the loss reduction analysis model so that the loss reduction analysis model outputs the corresponding loss reduction strategy. Using the aforementioned loss reduction strategy, the power system is subjected to loss reduction regulation to obtain the regulation result.

6. The method for visualizing power system loss reduction control results according to claim 1, characterized in that, The acquisition of the anomaly diagnosis model is specifically as follows: A learning vector quantization network is pre-constructed, and the second sample dataset is divided into a training set and a validation set; The learning vector quantization network is iteratively trained and validated. In each iteration, the current learning vector quantization network is trained using the current training set to obtain a pre-trained model. Then, the current pre-trained model is validated using the current validation set to obtain the diagnostic accuracy of the current pre-trained model. Then, the sample data in the current training set is updated, and the current pre-trained model is used as the learning vector quantization network. The iteration continues until the preset number of iterations is reached, at which point the iteration stops. The pre-trained model with the highest diagnostic accuracy during the iteration process is used as the anomaly diagnosis model. The second sample dataset includes several sample data sets, which are obtained by preprocessing historical operational data of line loss anomalies.

7. A visualization system for power system loss reduction control results, characterized in that, The following is stated: The anomaly diagnosis module is used to perform anomaly diagnosis on the line loss of the power system based on the power system's operating data and obtain the diagnosis results. Specifically, it extracts features from the operating data to obtain the key features in the operating data. Based on the input format requirements of the anomaly diagnosis model, the key features are formatted to obtain corresponding input features, which are then input into the anomaly diagnosis model so that the model can perform anomaly diagnosis on the input features and finally output the corresponding diagnosis result. The anomaly diagnosis model is obtained by training a pre-constructed learning vector quantization network, and the running data is obtained by real-time monitoring of the line loss of the power system. The loss reduction and control module is used to perform loss reduction analysis on the operating data when the diagnostic result is abnormal, and to perform loss reduction control on the power system based on the loss reduction strategy obtained from the analysis, so as to obtain the control result; The modeling module is used to model the control results to obtain a display model; The brightness adjustment module is used to measure the current display brightness of the display model and the current light intensity of the environment in which the display model is located in real time, and adjust the display brightness of the display model according to the difference between the current light intensity and the current display brightness; An angle adjustment module is used to acquire scene images of the display model within the display range in real time, then extract features from the scene images, and adjust the display angle of the display model according to the position information of the extracted first head feature; wherein, the first head feature is the feature information of the person's head in the scene image.

8. The power system loss reduction control result visualization system as described in claim 7, characterized in that, The angle adjustment module specifically includes: The feature extraction unit is used to collect scene images of the display model within the display range in real time, and to extract features from the scene images to obtain the corresponding first head features; An angle adjustment unit is used to compare the head position information corresponding to the first head feature with the position information of the center point of the scene image to obtain a comparison difference. Based on the comparison difference, the display angle of the display model is controlled and adjusted until the head position of the first head feature is at the center point of the scene image.

9. The power system loss reduction control result visualization system as described in claim 7, characterized in that, The loss reduction control module specifically includes: A matrix construction unit is used to extract multiple line loss influencing factors from the running data when the diagnostic result is abnormal, and to construct a high-dimensional random matrix based on all the line loss influencing factors. The loss reduction analysis unit is used to extract features from the high-dimensional random matrix to obtain high-dimensional random matrix features that characterize the operating status of the distribution network. Then, the high-dimensional random matrix features are input into the loss reduction analysis model so that the loss reduction analysis model outputs the corresponding loss reduction strategy. The loss reduction control unit is used to perform loss reduction control on the power system using the loss reduction strategy to obtain the control result.

Images