Method, device, medium and equipment for evaluating running state of transformer distribution area

Abstract

The invention discloses a transformer power distribution area operation state evaluation method and device, a medium and equipment, and relates to the technical field of transformer power distribution area monitoring, and the method comprises the steps: obtaining the real-time operation parameters of a transformer power distribution area; performing feature extraction on the real-time operation parameters to obtain operation features; adopting a preset LSTM neural network fault prediction model to carry out fault prediction on the operation characteristics to obtain a fault probability value; and when the fault probability value is greater than or equal to a preset threshold value, performing operation state evaluation on the transformer power distribution area by adopting a fuzzy comprehensive evaluation method based on the real-time operation parameters to obtain an operation state evaluation result of the transformer power distribution area. The method provided by the invention can improve the evaluation accuracy of the operation state of the transformer distribution area, avoids the fault loss of the power distribution network caused by misjudgment, and saves the operation and maintenance cost.

Description

Technical Field

[0001] This invention relates to the field of transformer distribution substation monitoring technology, and in particular to a method, apparatus, medium, and equipment for assessing the operating status of transformer distribution substations. Background Technology

[0002] With the advancement of the construction of new power systems, the operational reliability of distribution transformer areas, as the final nodes of power transmission, directly affects the quality of power supply. According to statistics, power outages caused by distribution transformer failures account for approximately 63% of distribution network failures, and traditional monitoring methods are no longer sufficient to meet the needs of intelligent operation and maintenance. Currently, there are three major pain points: (1) insufficient fusion of multi-source heterogeneous data leads to distorted state assessment; (2) weak edge computing capabilities restrict the efficiency of real-time diagnosis; and (3) predictive maintenance lacks effective data support. Developing new monitoring methods is of great significance for improving equipment lifespan, reducing operation and maintenance costs, and ensuring power supply reliability. Summary of the Invention

[0003] In view of this, the present invention provides a method, apparatus, medium and equipment for evaluating the operating status of transformer distribution substations. The main purpose is to solve the problem that insufficient fusion of multi-source heterogeneous data leads to distorted status evaluation, resulting in inaccurate evaluation of the operating status of transformer distribution substations.

[0004] To address the aforementioned problems, this application provides a method for assessing the operating status of a transformer distribution substation, comprising:

[0005] Obtain real-time operating parameters of the transformer distribution substation;

[0006] Feature extraction is performed on the real-time operating parameters to obtain operating features;

[0007] A preset LSTM neural network fault prediction model is used to predict the faults of the operational characteristics to obtain fault probability values.

[0008] When the fault probability value is greater than or equal to a preset threshold, the fuzzy comprehensive evaluation method is used to evaluate the operating status of the transformer distribution area based on the real-time operating parameters, and the operating status evaluation result of the transformer distribution area is obtained.

[0009] Optionally, the real-time operating parameters include electrical quantity parameters, temperature parameters, vibration parameters, and humidity parameters;

[0010] The electrical parameters include voltage parameters, current parameters, active power parameters, reactive power parameters, frequency parameters, and harmonic distortion rate parameters.

[0011] The temperature parameters include winding temperature parameters and ambient temperature parameters.

[0012] Optionally, the step of extracting features from the real-time operating parameters to obtain operating features specifically includes:

[0013] The real-time operating parameters are time-aligned according to different parameter categories to obtain a first parameter array;

[0014] The first parameter array is normalized to obtain the second parameter array;

[0015] The second parameter array is subjected to outlier removal processing to obtain the third parameter array;

[0016] The operational features are obtained by extracting features based on the third parameter array;

[0017] The operating characteristics include voltage fluctuation rate, current harmonic content, temperature rise rate, and frequency domain characteristics.

[0018] Optionally, before using a preset LSTM neural network fault prediction model to predict the operational features, the method further includes: constructing a preset LSTM neural network fault prediction model;

[0019] The construction of the preset LSTM neural network fault prediction model specifically includes:

[0020] Historical operating parameters are obtained, including historical electrical quantity parameters, historical temperature parameters, historical vibration parameters, and historical humidity parameters;

[0021] The historical operating parameters are labeled to obtain label data; the label data includes positive sample label data representing fault data and negative sample label data representing normal data.

[0022] The LSTM neural network, including two LSTM network layers, is trained using the labeled data as training samples to obtain the preset LSTM neural network fault prediction model that meets the preset model accuracy.

[0023] Optionally, the step of using a preset LSTM neural network fault prediction model to predict the operational characteristics and obtain a fault probability value specifically includes:

[0024] Temporal features are extracted using the first LSTM network layer of the preset LSTM neural network fault prediction model to obtain the current hidden state containing temporal features;

[0025] The second LSTM network layer of the preset LSTM neural network fault prediction model is used to perform feature fusion processing on the current hidden state to obtain a multi-dimensional global temporal feature vector.

[0026] The global time-series feature vector is subjected to a linear transformation to obtain the fault probability value.

[0027] Optionally, when the fault probability value is greater than or equal to a preset threshold, the transformer distribution substation is evaluated for its operational status using a fuzzy comprehensive evaluation method based on the real-time operating parameters, resulting in an evaluation result for the transformer distribution substation's operational status. This specifically includes:

[0028] When the fault probability value is greater than or equal to a preset threshold, the preset weights corresponding to different types of real-time status parameters are updated based on the fault probability value to obtain an updated weight matrix.

[0029] Based on the weight matrix, the fuzzy comprehensive evaluation method is used to score the different warning levels of the transformer distribution area, and the comprehensive scores corresponding to the different warning levels are obtained.

[0030] The warning level with the highest comprehensive score is determined as the target warning level, and the operation status assessment result of the transformer distribution area is obtained.

[0031] Optionally, the step of using a fuzzy comprehensive evaluation method based on the weight matrix to score different early warning levels of the transformer distribution area, and obtaining comprehensive scores corresponding to different early warning levels, specifically includes:

[0032] The membership function corresponding to different types of feature parameters is used for calculation to obtain the membership values ​​representing different evaluation indicators corresponding to the different types of feature parameters;

[0033] Construct a fuzzy relation matrix based on each of the membership values;

[0034] The comprehensive scores corresponding to different warning levels are obtained by performing calculations based on the weight matrix and the fuzzy relation matrix.

[0035] To address the aforementioned problems, this application provides a transformer distribution substation operation status assessment device, comprising:

[0036] The acquisition module is used to acquire real-time operating parameters of the transformer distribution substation.

[0037] The feature extraction module is used to extract features from the real-time operating parameters to obtain operating features;

[0038] The fault prediction module is used to predict the faults of the operating characteristics using a preset LSTM neural network fault prediction model to obtain fault probability values.

[0039] The status assessment module is used to assess the operating status of the transformer distribution area based on the real-time operating parameters using a fuzzy comprehensive evaluation method when the fault probability value is greater than or equal to a preset threshold, and to obtain the operating status assessment result of the transformer distribution area.

[0040] To address the aforementioned problems, this application provides a storage medium storing a computer program that, when executed by a processor, implements the steps of the transformer distribution substation operation status assessment method described above.

[0041] To address the aforementioned problems, this application provides an electronic device, comprising at least a memory and a processor. The memory stores a computer program, and the processor, when executing the computer program in the memory, implements the steps of the aforementioned transformer distribution substation operation status assessment method.

[0042] The beneficial effects of this application are as follows: This application obtains real-time operating parameters of the transformer distribution area; the real-time operating parameters include electrical quantity parameters, temperature parameters, vibration parameters, and humidity parameters; the electrical quantity parameters include voltage parameters, current parameters, active power parameters, reactive power parameters, frequency parameters, and harmonic distortion rate parameters; the temperature parameters include winding temperature parameters and ambient temperature parameters. Analysis of multi-dimensional real-time data effectively solves the problem of distorted state assessment caused by insufficient heterogeneous data fusion, and analysis of multi-dimensional parameters can improve the accuracy of subsequent transformer distribution area operation state assessment results. Feature extraction is performed on the real-time operating parameters to obtain operating features; a preset LSTM neural network fault prediction model is used to predict faults based on the operating features to obtain fault probability values; the preset LSTM neural network fault prediction model is used to predict fault probability, improving the accuracy and real-time performance of fault prediction. When the fault probability value is greater than or equal to a preset threshold, a fuzzy comprehensive evaluation method is used based on the real-time operating parameters to assess the operating state of the transformer distribution area, obtaining the transformer distribution area operation state assessment result. By combining neural network models and fuzzy comprehensive evaluation methods, accurate state assessment results can be provided, avoiding losses from power distribution network faults caused by misjudgments and saving operation and maintenance costs.

[0043] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0044] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0045] Figure 1 A flowchart illustrating a method for evaluating the operating status of a transformer distribution substation according to an embodiment of this application is shown.

[0046] Figure 2 A flowchart illustrating a method for evaluating the operating status of a transformer distribution substation according to another embodiment of this application is shown.

[0047] Figure 3 A structural block diagram of a transformer distribution substation operation status assessment device provided in another embodiment of this application is shown. Detailed Implementation

[0048] Various embodiments and features of this application are described herein with reference to the accompanying drawings.

[0049] It should be understood that various modifications can be made to the embodiments described herein. Therefore, the above description should not be considered as limiting, but merely as an example of embodiments. Other modifications within the scope and spirit of this application will be apparent to those skilled in the art.

[0050] The accompanying drawings, which are included in and form part of this specification, illustrate embodiments of the present application and, together with the general description of the present application given above and the detailed description of the embodiments given below, serve to explain the principles of the present application.

[0051] These and other features of this application will become apparent from the following description of preferred forms of embodiments given as non-limiting examples, with reference to the accompanying drawings.

[0052] It should also be understood that although this application has been described with reference to some specific examples, those skilled in the art can certainly implement many other equivalent forms of this application.

[0053] The above and other aspects, features and advantages of this application will become more apparent when taken in conjunction with the accompanying drawings and in view of the following detailed description.

[0054] Specific embodiments of this application are described thereafter with reference to the accompanying drawings; however, it should be understood that the claimed embodiments are merely examples of this application, which can be implemented in various ways. Well-known and / or repeated functions and structures are not described in detail to avoid unnecessary or redundant details that could obscure the application. Therefore, the specific structural and functional details claimed herein are not intended to be limiting, but merely serve as the basis and representative basis for the claims to teach those skilled in the art to use this application in a variety of substantially any suitable detailed structures.

[0055] This specification may use the phrases “in one embodiment,” “in another embodiment,” “in yet another embodiment,” or “in other embodiments,” all of which may refer to one or more of the same or different embodiments according to this application.

[0056] This application provides a method for evaluating the operating status of a transformer distribution substation, such as... Figure 1 As shown, it includes:

[0057] Step S101: Obtain the real-time operating parameters of the transformer distribution substation;

[0058] In this specific implementation process, the real-time operating parameters include electrical quantity parameters, temperature parameters, vibration parameters, and humidity parameters. The electrical quantity parameters include voltage parameters, current parameters, active power parameters, reactive power parameters, frequency parameters, and harmonic distortion rate parameters. The temperature parameters include winding temperature parameters and ambient temperature parameters. Voltage transformers can be used to collect the input and output voltage of the distribution area; current transformers can be used to collect the current signal of the distribution area; temperature sensors installed at the transformer windings, core, and cable connections can be used to collect the temperature parameters of the distribution area; and temperature rise and humidity sensors (accuracy ±5%RH) can be used to monitor the ambient humidity of the distribution area to prevent insulation from becoming damp. Vibration sensors (sensitivity ≥100mV / g) are used to detect abnormal mechanical vibrations in the transformer (such as loose windings). The data acquisition and transmission equipment uses intelligent terminals (RTU / FTU): integrating analog-to-digital converters (ADC, sampling rate ≥1kHz) and edge computing modules to achieve data preprocessing. Communication modules: 4G / 5G, LoRa, or fiber optic, supporting real-time data upload to the cloud platform (transmission latency ≤200ms).

[0059] Step S102: Extract features from the real-time operating parameters to obtain operating features;

[0060] In this step, the real-time operating parameters are time-aligned according to different parameter categories to obtain a first parameter array. Different parameter categories include electrical quantities, temperature, vibration, humidity, and other parameter types. A preset time protocol is used to align the real-time operating parameters according to different parameter categories, for example, using the IEEE 1588 Precise Time Protocol (PTP) to ensure multi-source data time synchronization (error < 1ms). The first parameter array is normalized to obtain a second parameter array; each parameter is normalized to the [0,1] interval to eliminate dimensional influence. Outlier removal is performed on the second parameter array to obtain a third parameter array; outliers are removed based on the 3σ principle or box plot method. Feature extraction is performed based on the third parameter array to obtain the operating features; these operating features include characteristic parameters such as voltage fluctuation rate, current harmonic content, temperature rise rate, and frequency domain characteristics.

[0061] Step S103: Use a preset LSTM neural network fault prediction model to predict the faults of the operating features and obtain the fault probability value;

[0062] In this step, the first LSTM network layer of the preset LSTM neural network fault prediction model is used to extract temporal features, obtaining the current hidden state containing temporal features. The second LSTM network layer of the preset LSTM neural network fault prediction model is then used to perform feature fusion processing on the current hidden state, obtaining a multi-dimensional global temporal feature vector. A linear transformation is then performed on the global temporal feature vector to obtain the fault probability value. Specifically, a preset sliding window is used to collect the operational feature data. The edge computing terminal inputs 60-second window data every minute, and the preset LSTM neural network fault prediction model is used to predict faults in the operational features to obtain the fault probability value.

[0063] Step S104: When the fault probability value is greater than or equal to a preset threshold, the transformer distribution area is evaluated for its operating status based on the real-time operating parameters using a fuzzy comprehensive evaluation method, and the operating status evaluation result of the transformer distribution area is obtained.

[0064] In this step, the preset threshold can be 0.3, which can be set according to actual needs. When the fault probability value is greater than or equal to the preset threshold, the preset weights corresponding to different types of real-time status parameters are updated based on the fault probability value to obtain an updated weight matrix. Based on the weight matrix, a fuzzy comprehensive evaluation method is used to score different warning levels of the transformer distribution area, obtaining comprehensive scores corresponding to different warning levels. The warning level with the highest comprehensive score is determined as the target warning level, and the operating status evaluation result of the transformer distribution area is obtained. When the fault probability value is less than the preset threshold, it is judged as low risk, only data is recorded, and no warning is triggered.

[0065] This application acquires real-time operating parameters of a transformer distribution substation. These real-time operating parameters include electrical quantity parameters, temperature parameters, vibration parameters, and humidity parameters. The electrical quantity parameters include voltage parameters, current parameters, active power parameters, reactive power parameters, frequency parameters, and harmonic distortion rate parameters. The temperature parameters include winding temperature parameters and ambient temperature parameters. Analyzing multi-dimensional real-time data effectively solves the problem of distorted state assessment caused by insufficient heterogeneous data fusion. Analyzing multi-dimensional parameters can improve the accuracy of subsequent transformer distribution substation operating state assessment results. Feature extraction is performed on the real-time operating parameters to obtain operating features. A preset LSTM neural network fault prediction model is used to predict faults based on these operating features, obtaining fault probability values. The preset LSTM neural network fault prediction model is then used to predict fault probabilities, improving the accuracy and real-time performance of fault prediction. When the fault probability value is greater than or equal to a preset threshold, a fuzzy comprehensive evaluation method is used based on the real-time operating parameters to assess the operating state of the transformer distribution substation, obtaining the transformer distribution substation operating state assessment result. By combining neural network models and fuzzy comprehensive evaluation methods, accurate state assessment results can be provided, avoiding losses from power distribution network faults caused by misjudgments and saving operation and maintenance costs.

[0066] Another embodiment of this application provides another method for evaluating the operating status of transformer distribution substations, such as... Figure 2 As shown, it includes:

[0067] Step S201: Obtain the real-time operating parameters of the transformer distribution substation;

[0068] In this specific implementation process, the electrical parameters include voltage, current, active power, reactive power, frequency, and harmonic distortion rate; the temperature parameters include winding temperature and ambient temperature. Voltage transformers can be used to collect the input and output voltages of the distribution area; current transformers can be used to collect the current signals of the distribution area; temperature sensors installed on the transformer windings, core, and cable connections can be used to collect the temperature parameters of the distribution area; and temperature rise and humidity sensors (accuracy ±5%RH) can be used to monitor the ambient humidity of the distribution area to prevent insulation from getting damp. Vibration sensors (sensitivity ≥100mV / g) are used to detect abnormal mechanical vibrations in the transformer (such as loose windings). Data acquisition and transmission equipment uses intelligent terminals (RTU / FTU): integrating analog-to-digital converters (ADC, sampling rate ≥1kHz) and edge computing modules to achieve data preprocessing. Communication modules: 4G / 5G, LoRa, or fiber optic, supporting real-time data upload to the cloud platform (transmission latency ≤200ms).

[0069] Step S202: Extract features from the real-time operating parameters to obtain operating features;

[0070] In this step, the real-time operating parameters are time-aligned according to different parameter categories to obtain a first parameter array. Different parameter categories include electrical quantities, temperature, vibration, humidity, and other parameter types. A preset time protocol is used to align the real-time operating parameters according to different parameter categories, for example, using the IEEE 1588 Precise Time Protocol (PTP) to ensure multi-source data time synchronization (error < 1ms). The first parameter array is normalized to obtain a second parameter array; each parameter is normalized to the [0,1] interval to eliminate dimensional influence. Outlier removal is performed on the second parameter array to obtain a third parameter array; outliers are removed based on the 3σ principle or box plot method. Feature extraction is performed based on the third parameter array to obtain the operating features; these operating features include characteristic parameters such as voltage fluctuation rate, current harmonic content, temperature rise rate, and frequency domain characteristics.

[0071] Step S203: Construct a pre-defined LSTM neural network fault prediction model;

[0072] In the specific implementation process of this step, the construction of the preset LSTM neural network fault prediction model specifically includes: acquiring historical operating parameters, including historical electrical quantity parameters, historical temperature parameters, historical vibration parameters, and historical humidity parameters; labeling the historical operating parameters to obtain label data; the label data includes positive sample label data representing fault data and negative sample label data representing normal data; training the LSTM neural network including a two-layer LSTM network layer based on the label data as training samples to obtain the preset LSTM neural network fault prediction model that meets the preset model accuracy. Each LSTM network layer has 64 neurons. Specifically, the labels of the positive sample data include the time of fault occurrence and the fault type, including categories such as overload and insulation aging; extracting time-series data (60 time steps × 12-dimensional features) one hour before the fault occurred, labeling it as 1, and performing data augmentation processing on the data of a few fault categories to obtain positive sample label data representing fault data; the data augmentation processing method can be oversampling or adding Gaussian noise, and oversampling can use the SMOTE algorithm. Negative sample normal data is obtained by randomly sampling data from normal operating periods and labeling it as 0. The ratio of negative samples to positive samples is 3:1 to achieve a balanced processing effect. Both positive and negative sample labeled data undergo preprocessing to obtain preprocessed data. The preprocessing process includes normalization and sequence segmentation. Normalization is performed using Min-Max normalization based on the feature dimension (e.g., voltage is normalized to [0,1]). The time-series segmentation process involves: generating samples using a sliding window (window = 60 seconds, step size = 1 second), resulting in an input matrix of shape (N_samples, 60); dividing the dataset into training, test, and validation sets in a ratio of 7:1.5:1.5; training the preset LSTM neural network using the training set, testing the trained model using the test set, and validating the model loss using the validation set. Model training stops when the validation set loss does not decrease for 5 consecutive rounds, resulting in the preset LSTM neural network fault prediction model.

[0073] Step S204: Use a preset LSTM neural network fault prediction model to predict the faults of the operating characteristics and obtain the fault probability value;

[0074] In this step, the first LSTM network layer of the preset LSTM neural network fault prediction model is used to extract temporal features, obtaining the current hidden state containing temporal features. The second LSTM network layer of the preset LSTM neural network fault prediction model is used to perform feature fusion processing on the current hidden state, obtaining a multi-dimensional global temporal feature vector. The global temporal feature vector is then subjected to linear transformation to obtain the fault probability value. Specifically, the forget gate of the first LSTM network layer is used to process the voltage fluctuation rate of the operating feature, obtaining the voltage fluctuation state feature. The input gate of the first LSTM network layer is used to process the temperature rise rate and current harmonic interaction of the state feature, obtaining the temperature rise rate state feature and the current harmonic interaction state feature. The cell state of the first LSTM network layer is updated to obtain the frequency domain state feature. The correlation between the encoded voltage fluctuation and harmonics and the trend of the temperature rise rate are analyzed to obtain the current hidden state containing temporal features. The current hidden state is subjected to high-order temporal dependency capture to obtain the hidden state corresponding to each time step; the hidden state corresponding to the last time step is extracted to obtain a multi-dimensional global temporal feature vector; the global temporal feature vector is subjected to linear transformation to obtain the fault probability value.

[0075] Step S205: When the fault probability value is greater than or equal to a preset threshold, the preset weights corresponding to different types of real-time state parameters are updated based on the fault probability value to obtain the updated weight matrix;

[0076] In this step, when the fault probability value is greater than or equal to a preset threshold, the preset weights corresponding to the feature parameters are updated according to the fault probability value to obtain an updated weight matrix. For example, the feature parameters include load rate, temperature rise, vibration, and humidity; the weights obtained based on historical fault statistics are 0.4, 0.3, 0.2, and 0.1, respectively; the preset weight for load rate is 0.4; the preset weight for temperature rise is 0.3; the preset weight for vibration is 0.2; and the preset weight for humidity is 0.1. When the fault probability value is greater than 0.6, the vibration weight is increased to 0.3, and so on. Specifically, an adjustment range for the feature parameters can be preset. When the probability value falls into the adjustment range, the probability of the corresponding adjustment range is updated to obtain an updated weight matrix.

[0077] Step S206: Based on the weight matrix, the fuzzy comprehensive evaluation method is used to score the different warning levels of the transformer distribution area, and the comprehensive scores corresponding to the different warning levels are obtained respectively;

[0078] In this step, membership functions corresponding to different types of feature parameters are used for calculation to obtain membership values ​​representing different evaluation indicators for each type of feature parameter. For example, for the load rate feature parameter: the mathematical expression of the membership function for low load rate (0-60%) is: membership degree = 1 - load rate / 60; the mathematical expression of the membership function for high load rate (>60%) is: load rate / 100; for the temperature feature parameter: the mathematical expression of the membership function for normal temperature (ΔT < 40K) is: membership degree = 1 - ΔT / 80; the mathematical expression of the membership function for dangerous temperature (ΔT ≥ 60K) is: membership degree = 1, etc. A fuzzy relation matrix is ​​constructed based on each membership value; multiplication is performed on the weight matrix and the fuzzy relation matrix to obtain the comprehensive score corresponding to different warning levels. Different warning levels can be divided into Level 1, Level 2 and Level 3 according to the order of danger from high to low; the first comprehensive score corresponding to Level 1 warning level, the second comprehensive score corresponding to Level 2 warning level and the third comprehensive score corresponding to Level 3 warning level can be obtained.

[0079] Step S207: Determine the warning level with the highest comprehensive score as the target warning level to obtain the transformer distribution area operation status assessment result;

[0080] In this step, the warning level with the highest comprehensive score is determined as the target warning level, thus obtaining the transformer distribution area operation status assessment result. For example, if the first comprehensive score corresponding to the Level 1 warning level is 0.2175, the second comprehensive score corresponding to the Level 2 warning level is 0.5425, and the third comprehensive score corresponding to the Level 3 warning level is 0.24, then the Level 2 warning level corresponding to the highest comprehensive score of 0.5425 is determined as the target warning level. Subsequently, the operation parameters of the transformer distribution area can be controlled according to different warning levels. For example, when the target warning level is Level 1, an immediate shutdown and on-site maintenance within 30 minutes are implemented; when the target warning level is Level 2, a 24-hour inspection and load adjustment are implemented; when the target warning level is Level 3, remote activation of the dehumidification device and trend recording are used, etc., to control the current warning status, prevent accidents, and proactively prevent faults.

[0081] Step S208: When the fault probability value is less than the preset threshold, the current transformer distribution area is determined to be in a low-risk state.

[0082] In the specific implementation process of this step, when the fault probability value is less than the preset threshold, the current operating status of the transformer distribution area is determined to be a low-risk state, and the current operating status data is recorded to provide data support for subsequent fault analysis.

[0083] This application acquires real-time operating parameters of a transformer distribution substation. These real-time operating parameters include electrical quantity parameters, temperature parameters, vibration parameters, and humidity parameters. The electrical quantity parameters include voltage parameters, current parameters, active power parameters, reactive power parameters, frequency parameters, and harmonic distortion rate parameters. The temperature parameters include winding temperature parameters and ambient temperature parameters. Analyzing multi-dimensional real-time data effectively solves the problem of distorted state assessment caused by insufficient heterogeneous data fusion. Analyzing multi-dimensional parameters can improve the accuracy of subsequent transformer distribution substation operating state assessment results. Feature extraction is performed on the real-time operating parameters to obtain operating features. A preset LSTM neural network fault prediction model is used to predict faults based on these operating features, obtaining fault probability values. The preset LSTM neural network fault prediction model is then used to predict fault probabilities, improving the accuracy and real-time performance of fault prediction. When the fault probability value is greater than or equal to a preset threshold, a fuzzy comprehensive evaluation method is used based on the real-time operating parameters to assess the operating state of the transformer distribution substation, obtaining the transformer distribution substation operating state assessment result. By combining neural network models and fuzzy comprehensive evaluation methods, accurate state assessment results can be provided, avoiding losses from power distribution network faults caused by misjudgments and saving operation and maintenance costs.

[0084] Another embodiment of this application provides a transformer distribution substation operation status assessment device, such as... Figure 3 As shown, it includes:

[0085] Module 1 is used to acquire real-time operating parameters of the transformer distribution area;

[0086] Feature extraction module 2 is used to extract features from the real-time operating parameters to obtain operating features;

[0087] Fault prediction module 3 is used to predict the faults of the operating features using a preset LSTM neural network fault prediction model to obtain fault probability values;

[0088] The status assessment module 4 is used to assess the operating status of the transformer distribution area based on the real-time operating parameters using a fuzzy comprehensive evaluation method when the fault probability value is greater than or equal to a preset threshold, and to obtain the operating status assessment result of the transformer distribution area.

[0089] In specific implementation, the acquisition module 1 is specifically used for: the real-time operating parameters include electrical quantity parameters, temperature parameters, vibration parameters, and humidity parameters; the electrical quantity parameters include voltage parameters, current parameters, active power parameters, reactive power parameters, frequency parameters, and harmonic distortion rate parameters; the temperature parameters include winding temperature parameters and ambient temperature parameters.

[0090] In the specific implementation process, the feature extraction module 2 is specifically used to: perform time alignment processing on the real-time operating parameters according to different parameter categories to obtain a first parameter array; perform normalization processing on the first parameter array to obtain a second parameter array; perform outlier removal processing on the second parameter array to obtain a third parameter array; and perform feature extraction based on the third parameter array to obtain the operating features; wherein, the operating features include voltage fluctuation rate, current harmonic content, temperature rise rate, and frequency domain features.

[0091] In specific implementation, the device further includes a model building module, which is specifically used for: building a preset LSTM neural network fault prediction model, specifically including: acquiring historical operating parameters, including historical electrical quantity parameters, historical temperature parameters, historical vibration parameters, and historical humidity parameters; labeling the historical operating parameters to obtain label data; the label data includes positive sample label data representing fault data and negative sample label data representing normal data; training an LSTM neural network including a two-layer LSTM network based on the label data as training samples to obtain the preset LSTM neural network fault prediction model that meets the preset model accuracy.

[0092] In the specific implementation process, the fault prediction module 3 is specifically used to: extract temporal features using the first LSTM network layer of the preset LSTM neural network fault prediction model to obtain the current hidden state containing temporal features; perform feature fusion processing on the current hidden state using the second LSTM network layer of the preset LSTM neural network fault prediction model to obtain a multi-dimensional global temporal feature vector; and perform linear transformation processing on the global temporal feature vector to obtain the fault probability value.

[0093] In the specific implementation process, the status assessment module 4 is specifically used for: when the fault probability value is greater than or equal to a preset threshold, updating the preset weights corresponding to different types of real-time status parameters based on the fault probability value to obtain an updated weight matrix; using a fuzzy comprehensive evaluation method based on the weight matrix to score different warning levels of the transformer distribution area to obtain comprehensive scores corresponding to different warning levels; determining the warning level with the highest comprehensive score as the target warning level to obtain the transformer distribution area operation status assessment result.

[0094] In the specific implementation process, the state assessment module 4 is also used to: perform calculation processing using membership functions corresponding to different types of feature parameters to obtain membership values ​​representing different evaluation indicators corresponding to different types of feature parameters; construct a fuzzy relation matrix based on each membership value; and perform calculation processing based on the weight matrix and the fuzzy relation matrix to obtain comprehensive scores corresponding to different warning levels.

[0095] This application acquires real-time operating parameters of a transformer distribution substation. These real-time operating parameters include electrical quantity parameters, temperature parameters, vibration parameters, and humidity parameters. The electrical quantity parameters include voltage parameters, current parameters, active power parameters, reactive power parameters, frequency parameters, and harmonic distortion rate parameters. The temperature parameters include winding temperature parameters and ambient temperature parameters. Analyzing multi-dimensional real-time data effectively solves the problem of distorted state assessment caused by insufficient heterogeneous data fusion. Analyzing multi-dimensional parameters can improve the accuracy of subsequent transformer distribution substation operating state assessment results. Feature extraction is performed on the real-time operating parameters to obtain operating features. A preset LSTM neural network fault prediction model is used to predict faults based on these operating features, obtaining fault probability values. The preset LSTM neural network fault prediction model is then used to predict fault probabilities, improving the accuracy and real-time performance of fault prediction. When the fault probability value is greater than or equal to a preset threshold, a fuzzy comprehensive evaluation method is used based on the real-time operating parameters to assess the operating state of the transformer distribution substation, obtaining the transformer distribution substation operating state assessment result. By combining neural network models and fuzzy comprehensive evaluation methods, accurate state assessment results can be provided, avoiding losses from power distribution network faults caused by misjudgments and saving operation and maintenance costs.

[0096] Another embodiment of this application provides a storage medium storing a computer program, which, when executed by a processor, implements the following method steps:

[0097] Step 1: Obtain the real-time operating parameters of the transformer distribution area;

[0098] Step 2: Extract features from the real-time operating parameters to obtain operating features;

[0099] Step 3: Use a preset LSTM neural network fault prediction model to predict the faults of the operational features and obtain the fault probability value;

[0100] Step 4: When the fault probability value is greater than or equal to a preset threshold, the transformer distribution area is evaluated for its operating status based on the real-time operating parameters using a fuzzy comprehensive evaluation method, and the operating status evaluation result of the transformer distribution area is obtained.

[0101] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0102] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0103] The specific implementation process of the above method steps can be found in the embodiment of the above-mentioned method for evaluating the operating status of any transformer distribution area, which will not be repeated here.

[0104] This application acquires real-time operating parameters of a transformer distribution substation. These real-time operating parameters include electrical quantity parameters, temperature parameters, vibration parameters, and humidity parameters. The electrical quantity parameters include voltage parameters, current parameters, active power parameters, reactive power parameters, frequency parameters, and harmonic distortion rate parameters. The temperature parameters include winding temperature parameters and ambient temperature parameters. Analyzing multi-dimensional real-time data effectively solves the problem of distorted state assessment caused by insufficient heterogeneous data fusion. Analyzing multi-dimensional parameters can improve the accuracy of subsequent transformer distribution substation operating state assessment results. Feature extraction is performed on the real-time operating parameters to obtain operating features. A preset LSTM neural network fault prediction model is used to predict faults based on these operating features, obtaining fault probability values. The preset LSTM neural network fault prediction model is then used to predict fault probabilities, improving the accuracy and real-time performance of fault prediction. When the fault probability value is greater than or equal to a preset threshold, a fuzzy comprehensive evaluation method is used based on the real-time operating parameters to assess the operating state of the transformer distribution substation, obtaining the transformer distribution substation operating state assessment result. By combining neural network models and fuzzy comprehensive evaluation methods, accurate state assessment results can be provided, avoiding losses from power distribution network faults caused by misjudgments and saving operation and maintenance costs.

[0105] Another embodiment of this application provides an electronic device, which can be a server. The electronic device includes a processor, a memory, a network interface, and a database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the program is executed by the processor, it implements the functions or steps of a transformer substation operation status assessment method on the server side.

[0106] In one embodiment, an electronic device is provided, which can be a client. The electronic device includes a processor, memory, a network interface, a display screen, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with an external server via a network connection. When the program of the electronic device is executed by the processor, it implements the functions or steps of a transformer substation operation status assessment method on the client side.

[0107] Another embodiment of this application provides an electronic device, including at least a memory and a processor. The memory stores a computer program, and the processor, when executing the computer program in the memory, performs the following method steps:

[0108] Step 1: Obtain the real-time operating parameters of the transformer distribution area;

[0109] Step 2: Extract features from the real-time operating parameters to obtain operating features;

[0110] Step 3: Use a preset LSTM neural network fault prediction model to predict the faults of the operational features and obtain the fault probability value;

[0111] Step 4: When the fault probability value is greater than or equal to a preset threshold, the transformer distribution area is evaluated for its operating status based on the real-time operating parameters using a fuzzy comprehensive evaluation method, and the operating status evaluation result of the transformer distribution area is obtained.

[0112] The specific implementation process of the above method steps can be found in the embodiment of the above-mentioned method for evaluating the operating status of any transformer distribution area, which will not be repeated here.

[0113] This application acquires real-time operating parameters of a transformer distribution substation. These real-time operating parameters include electrical quantity parameters, temperature parameters, vibration parameters, and humidity parameters. The electrical quantity parameters include voltage parameters, current parameters, active power parameters, reactive power parameters, frequency parameters, and harmonic distortion rate parameters. The temperature parameters include winding temperature parameters and ambient temperature parameters. Analyzing multi-dimensional real-time data effectively solves the problem of distorted state assessment caused by insufficient heterogeneous data fusion. Analyzing multi-dimensional parameters can improve the accuracy of subsequent transformer distribution substation operating state assessment results. Feature extraction is performed on the real-time operating parameters to obtain operating features. A preset LSTM neural network fault prediction model is used to predict faults based on these operating features, obtaining fault probability values. The preset LSTM neural network fault prediction model is then used to predict fault probabilities, improving the accuracy and real-time performance of fault prediction. When the fault probability value is greater than or equal to a preset threshold, a fuzzy comprehensive evaluation method is used based on the real-time operating parameters to assess the operating state of the transformer distribution substation, obtaining the transformer distribution substation operating state assessment result. By combining neural network models and fuzzy comprehensive evaluation methods, accurate state assessment results can be provided, avoiding losses from power distribution network faults caused by misjudgments and saving operation and maintenance costs.

[0114] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. The scope of protection of this application is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to this application within its substance and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. A method for evaluating the operating status of a transformer distribution substation, characterized in that, include: Obtain real-time operating parameters of the transformer distribution substation; Feature extraction is performed on the real-time operating parameters to obtain operating features; A preset LSTM neural network fault prediction model is used to predict the faults of the operational characteristics to obtain fault probability values. When the fault probability value is greater than or equal to a preset threshold, the fuzzy comprehensive evaluation method is used to evaluate the operating status of the transformer distribution area based on the real-time operating parameters, and the operating status evaluation result of the transformer distribution area is obtained.

2. The method as described in claim 1, characterized in that, The real-time operating parameters include electrical quantity parameters, temperature parameters, vibration parameters, and humidity parameters; The electrical parameters include voltage parameters, current parameters, active power parameters, reactive power parameters, frequency parameters, and harmonic distortion rate parameters. The temperature parameters include winding temperature parameters and ambient temperature parameters.

3. The method as described in claim 1, characterized in that, The step of extracting features from the real-time operating parameters to obtain operating features specifically includes: The real-time operating parameters are time-aligned according to different parameter categories to obtain a first parameter array; The first parameter array is normalized to obtain the second parameter array; The second parameter array is subjected to outlier removal processing to obtain the third parameter array; The operational features are obtained by extracting features based on the third parameter array; The operating characteristics include voltage fluctuation rate, current harmonic content, temperature rise rate, and frequency domain characteristics.

4. The method as described in claim 1, characterized in that, Before using a preset LSTM neural network fault prediction model to predict the operational characteristics, the method further includes: constructing a preset LSTM neural network fault prediction model; The construction of the preset LSTM neural network fault prediction model specifically includes: Historical operating parameters are obtained, including historical electrical quantity parameters, historical temperature parameters, historical vibration parameters, and historical humidity parameters; The historical operating parameters are labeled to obtain label data; the label data includes positive sample label data representing fault data and negative sample label data representing normal data. The LSTM neural network, including two LSTM network layers, is trained using the labeled data as training samples to obtain the preset LSTM neural network fault prediction model that meets the preset model accuracy.

5. The method as described in claim 1, characterized in that, The step of using a preset LSTM neural network fault prediction model to predict faults in the operational features and obtain fault probability values ​​specifically includes: Temporal features are extracted using the first LSTM network layer of the preset LSTM neural network fault prediction model to obtain the current hidden state containing temporal features; The second LSTM network layer of the preset LSTM neural network fault prediction model is used to perform feature fusion processing on the current hidden state to obtain a multi-dimensional global temporal feature vector. The global time-series feature vector is subjected to a linear transformation to obtain the fault probability value.

6. The method as described in claim 1, characterized in that, When the fault probability value is greater than or equal to a preset threshold, the transformer distribution substation is evaluated using a fuzzy comprehensive evaluation method based on the real-time operating parameters to obtain the transformer distribution substation operating status evaluation result, specifically including: When the fault probability value is greater than or equal to a preset threshold, the preset weights corresponding to different types of real-time status parameters are updated based on the fault probability value to obtain an updated weight matrix. Based on the weight matrix, the fuzzy comprehensive evaluation method is used to score the different warning levels of the transformer distribution area, and the comprehensive scores corresponding to the different warning levels are obtained. The warning level with the highest comprehensive score is determined as the target warning level, and the operation status assessment result of the transformer distribution area is obtained.

7. The method as described in claim 6, characterized in that, The fuzzy comprehensive evaluation method based on the weight matrix is ​​used to score different early warning levels of the transformer distribution area, obtaining comprehensive scores corresponding to different early warning levels, specifically including: The membership function corresponding to different types of feature parameters is used for calculation to obtain the membership values ​​representing different evaluation indicators corresponding to the different types of feature parameters; Construct a fuzzy relation matrix based on each of the membership values; The comprehensive scores corresponding to different warning levels are obtained by performing calculations based on the weight matrix and the fuzzy relation matrix.

8. A transformer distribution substation operation status assessment device, characterized in that, include: The acquisition module is used to acquire real-time operating parameters of the transformer distribution substation. The feature extraction module is used to extract features from the real-time operating parameters to obtain operating features; The fault prediction module is used to predict the faults of the operating characteristics using a preset LSTM neural network fault prediction model to obtain fault probability values. The status assessment module is used to assess the operating status of the transformer distribution area based on the real-time operating parameters using a fuzzy comprehensive evaluation method when the fault probability value is greater than or equal to a preset threshold, and to obtain the operating status assessment result of the transformer distribution area.

9. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the transformer distribution substation operation status assessment method according to any one of claims 1-7.

10. An electronic device, characterized in that, It includes at least a memory and a processor, wherein the memory stores a computer program, and the processor, when executing the computer program in the memory, implements the steps of the transformer distribution substation operation status assessment method according to any one of claims 1-7.

Images