A digital monitoring method, device and medium for power transmission and transformation engineering

Abstract

The present invention discloses a digital monitoring method, device and medium for power transmission and transformation projects, and relates to the technical field of transformer monitoring. The method collects historical maintenance data of each transformer to obtain transformer classification data. For each operating transformer, historical power operation data is collected, and a power transmission state judgment model for predicting whether power transmission is abnormal is trained for each operating transformer. After the training of the power transmission state judgment model is completed, the real-time power operation data and real-time operation state data of each operating transformer are collected. For the operating transformers of each monitoring level, clustering is performed according to the real-time operation state data to obtain transformer clustering data, thereby improving the accuracy of the prediction of power fluctuation abnormality and reducing the prediction misjudgment rate.

Description

Technical Field

[0001] The present invention relates to the technical field of transformer operation monitoring, and in particular to a digital monitoring method, equipment and medium for power transmission and transformation engineering. Background Art

[0002] Transmission and transformation engineering is an important part of the power grid. The power grid usually includes four main links: power generation, transmission, transformation and distribution. Transmission and transformation engineering involves the transmission of electricity from power plants to substations for voltage conversion and distribution to different places. Therefore, transmission and transformation engineering is part of the power grid system. Its function is to transmit electricity from power plants to substations, and then distribute it to various power consumption areas such as cities, industrial areas, and rural areas.

[0003] In power transmission and transformation projects, transformers play a key role in voltage conversion and power distribution. Their specific functions include:

[0004] Including voltage conversion: Transformers can convert the voltage in the power grid from one voltage level to another. This is because different voltage levels are required for power transmission and distribution. Usually, the voltage is stepped down from high voltage (such as transmission lines) to low voltage (for home, industrial and commercial use).

[0005] Current distribution: Transformers can not only convert voltage, but also distribute current. Through transformers, the power grid system can distribute electrical energy to different users, ensuring that each user can get the power they need.

[0006] Improve the efficiency of power transmission: By reducing the loss of current and voltage, transformers can improve the efficiency of power transmission. High-voltage transmission lines reduce resistance losses, thereby reducing energy waste;

[0007] However, as a device in a high temperature and high pressure environment, the transformer ages quickly. When the transformer ages, the transmission efficiency will decrease, the voltage and current fluctuations in the power grid will be abnormal, and the power grid will be unstable.

[0008] Before the popularization of data analysis technology, the monitoring of transformer aging problems was usually carried out through regular inspections, but there was a lack of real-time inference based on the power operation status of the power grid whether the transformer aging has affected the normal operation of the power grid, so as to timely discover the aging of the transformer;

[0009] At present, some data analysis methods are also used to analyze the aging of transformers, but they are often based on the historical performance of a single transformer to infer whether the power operation is normal. However, because the aging degree and maintenance conditions of transformers are different at different times, the abnormal or normal power operation performance caused by aging of the same transformer at different times is also different. Therefore, using the historical data of a single transformer for abnormal analysis has a high misjudgment rate;

[0010] To this end, the present invention proposes a digital monitoring method, device and medium for power transmission and transformation projects. Summary of the invention

[0011] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a digital monitoring method, device and medium for power transmission and transformation engineering, which improves the accuracy of power fluctuation abnormality prediction and reduces the prediction error rate.

[0012] To achieve the above object, a digital monitoring method for power transmission and transformation engineering is proposed according to Embodiment 1 of the present invention, comprising the following steps:

[0013] Collect the operating transformers in the transmission system in advance and collect the historical maintenance data of each operating transformer;

[0014] Based on historical maintenance data, transformers are divided into several monitoring levels to obtain transformer classification data;

[0015] For each transformer, collect historical power operation data;

[0016] Based on historical power operation data, a power transmission status judgment model is trained for each transformer to predict whether power transmission is abnormal;

[0017] After the training of the power transmission status judgment model is completed, the real-time power operation data and real-time operation status data of each transformer are collected;

[0018] For each transformer at each monitoring level, cluster the real-time operating status data to obtain transformer cluster data;

[0019] Based on the real-time power operation data and the set of power transmission state judgment models corresponding to the clustered transformer set, a predicted value of the real-time power transmission state of the transformer is obtained;

[0020] The method of collecting the historical maintenance data of each transformer is as follows:

[0021] Collect maintenance-related data of all transformers in the transformer set;

[0022] The maintenance-related data includes the usage duration, the number of failures, the number of repairs, and the number of component replacements;

[0023] The maintenance-related data of all transformers constitutes historical maintenance data;

[0024] Based on the historical maintenance data, the transformers are divided into several monitoring levels, and the way to obtain the transformer classification data is as follows:

[0025] Mark the numbers of the transformers in the transformer set as i;

[0026] Mark the usage duration of the i-th transformer as Si, the number of failures as Gi, the number of repairs as Wi, and the number of component replacements as Bi;

[0027] For the i-th transformer, calculate the corresponding aging degree index Li; among them, the calculation formula of the aging degree index Li is: Li = a1×Si + a2×Gi + a3×Wi + a4×Bi, where a1, a2, a3, and a4 are respectively preset proportionality coefficients;

[0028] Preset N monitoring levels and set an aging degree threshold for each level;

[0029] Classify the i-th transformer into the corresponding monitoring level according to the aging degree index and the aging degree threshold;

[0030] In a preferred embodiment, the way to classify into the corresponding monitoring level is: the aging degree index of the i-th transformer satisfies the condition: Yn ≤ Li < Y(n + 1); where n is the number of the monitoring level into which the i-th transformer is classified, n = 1, 2, 3,... N; Yn is the aging degree threshold of the n-th monitoring index;

[0031] The corresponding relationship between the numbers of all transformers and the monitoring levels constitutes the transformer classification data;

[0032] For each transformer, the way to collect historical power operation data is as follows:

[0033] Preset the sliding window time length;

[0034] Every other sliding window time length, collect the power feature sequence set and the actual operation status label set of each transformer; the power feature sequence set includes power feature sequences corresponding to several power features, the length of the power feature sequence is one sliding window time length, and each element in the power feature sequence is the power feature value per unit time; among them, the actual operation status label represents the label value of whether each transformer is abnormal during the duration corresponding to each power feature sequence set, and the label value is 0 or 1;

[0035] The set of all power feature sequences corresponding to all transformers and the set of actual operation status labels constitute the historical power operation data;

[0036] The method of training a power transmission state judgment model for each transformer to predict whether power transmission is abnormal is as follows:

[0037] For each transformer:

[0038] Each power feature sequence set corresponding to the transformer is used as the input of the transmission state judgment model. The transmission state judgment model takes the predicted abnormal probability of the transformer operation state within each sliding window time length as the output, and the value range of the predicted abnormal probability is [0,1]. The actual operation state label corresponding to the power feature sequence set of the transformer in the historical power operation data is used as the prediction target, and the sum of the prediction errors of the predicted abnormal probability to the actual operation state label is minimized as the training target; the transmission state judgment model is trained until the sum of the prediction errors reaches convergence and the training is stopped;

[0039] The method of collecting the real-time power operation data and real-time operation status data of each transformer is as follows:

[0040] At every sliding window time length, the power feature sequences corresponding to the power features of each transformer in the time period are collected to form a power feature sequence set, and the power feature sequence set is used as real-time power operation data;

[0041] Every time the sliding window has a certain length, the operating state characteristic values ​​corresponding to the operating state characteristics of each transformer in the time period are collected to form a set of operating state characteristic value sets;

[0042] For each time period of the sliding window length, the set of operating status feature values ​​of all transformers constitutes real-time operating status data;

[0043] The method of obtaining transformer clustering data is:

[0044] Classify transformers according to monitoring levels;

[0045] For all transformers within each monitoring level:

[0046] The number of clusters k is preset, and the set of operating status feature values ​​of each transformer in the time period of each sliding window length in the real-time operating status data is taken as a data point;

[0047] All data points within each sliding window time period are divided into k clusters using a clustering algorithm;

[0048] The serial numbers of all transformers corresponding to each cluster divided by the clustering algorithm constitute transformer clustering data;

[0049] The method of obtaining the predicted value of the real-time power transmission state of the transformer is:

[0050] For the i-th transformer, the set of all transformers corresponding to the cluster corresponding to the transformer is marked as Ci, and the transformers in the set Ci of transformers except the i-th transformer are marked as ci;

[0051] In the time period corresponding to the length of each sliding window, the real-time power operation data corresponding to the i-th transformer is input into the power transmission state judgment model corresponding to all transformers in the transformer set Ci to obtain the predicted abnormal probability of the output;

[0052] The predicted abnormal probability of the power transmission state judgment model of the i-th transformer is marked as Pi, and the predicted abnormal probability of the power transmission state judgment model of the transformer ci is marked as Pci;

[0053] Calculate the predicted value Fi of the real-time power transmission state of the i-th transformer, the calculation formula of the predicted value Fi of the real-time power transmission state is: Fi = a5×pi+a6×∑ ci Pci; wherein a5 and a6 are preset proportional coefficients respectively.

[0054] An electronic device according to Embodiment 2 of the present invention includes: a processor and a memory, wherein the memory stores a computer program that can be called by the processor;

[0055] The processor executes the above-mentioned digital monitoring method for power transmission and transformation projects based on augmented reality technology by calling the computer program stored in the memory.

[0056] A computer-readable storage medium according to embodiment 3 of the present invention stores a rewritable computer program;

[0057] When the computer program is executed on a computer device, the computer device is enabled to execute the above-mentioned digital monitoring method for power transmission and transformation projects based on augmented reality technology.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] The present invention collects a set of operating transformers in a power transmission system in advance, and collects historical maintenance data of each transformer. Based on the historical maintenance data, the transformers are divided into several monitoring levels to obtain transformer classification data. For each transformer, historical power operation data is collected. Based on the historical power operation data, a transmission state judgment model for predicting whether power transmission is abnormal is trained for each transformer. After the training of the transmission state judgment model is completed, the real-time power operation data and real-time operation state data of each transformer are collected. For each monitoring level of transformers, clustering is performed according to the real-time operation state data to obtain transformer clustering data. Based on the real-time power operation data and the set of transmission state judgment models corresponding to the clustered transformer set, a predicted value of the real-time transmission state of the transformer is obtained. The abnormal prediction of transformers with similar operation states and similar aging degrees to the i-th transformer is comprehensively referred to, thereby improving the accuracy of abnormal power fluctuation prediction and reducing the prediction misjudgment rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a flow chart of a digital monitoring method for power transmission and transformation engineering in an embodiment of the present invention;

[0061] Figure 2 This is a schematic diagram of the structure of an electronic device in Embodiment 2 of the present invention;

[0062] Figure 3 This is a schematic diagram of the computer-readable storage medium structure in Example 3 of the present invention. DETAILED DESCRIPTION

[0063] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] Example 1

[0065] like Figure 1 As shown, a digital monitoring method for power transmission and transformation engineering includes the following steps:

[0066] Step 1: Collect the operating transformers in the transmission system in advance and collect the historical maintenance data of each operating transformer;

[0067] Step 2: Based on historical maintenance data, the operating transformers are divided into several monitoring levels to obtain transformer classification data;

[0068] Step 3: For each transformer, collect historical power operation data;

[0069] Step 4: Based on historical power operation data, a power transmission status judgment model is trained for each transformer to predict whether power transmission is abnormal;

[0070] Step 5: After the power transmission status judgment model training is completed, the real-time power operation data and real-time operation status data of each transformer are collected;

[0071] Step 6: For each transformer at each monitoring level, cluster the real-time operating status data to obtain transformer clustering data;

[0072] Step 7: Based on the real-time power operation data and the transmission state judgment model set corresponding to the clustered transformer set, obtain the predicted value of the real-time transmission state of the transformer; it can be understood that the predicted value of the real-time transmission state can indicate whether each transformer has an abnormality, thereby achieving the monitoring effect of the transformer state;

[0073] The method of pre-collecting the set of operating transformers in the power transmission system is as follows:

[0074] Based on the actual architecture of the power transmission and transformation project in the power grid system to be monitored, the locations of all transformers in use in the power grid system, the power area responsible for transmitting power, and their numbers are collected, and the corresponding relationship between the locations of all transformers and the transformer numbers is combined into a transformer set;

[0075] It is understandable that in the power grid system, each transformer is usually responsible for power transmission in a fixed area, and each transformer is installed at a different time and has different maintenance times, resulting in different degrees of aging. Therefore, the degree of influence of transformer aging on power transmission efficiency is also different;

[0076] Furthermore, the method of collecting the historical maintenance data of each transformer is as follows:

[0077] Collect maintenance-related data of all transformers in the transformer set;

[0078] Specifically, the maintenance-related data includes usage time, number of failures, number of repairs, and number of parts replacements;

[0079] The maintenance-related data of all transformers constitute the historical maintenance data;

[0080] Furthermore, based on the historical maintenance data, the transformers are divided into several monitoring levels, and the transformer classification data is obtained in the following manner:

[0081] The number of the transformer in the transformer set is marked as i;

[0082] Mark the usage duration of the i-th transformer as Si, the number of failures as Gi, the number of repairs as Wi, and the number of component replacements as Bi. It can be understood that the larger any one of the parameters of usage duration, number of failures, number of repairs, or number of component replacements, the more serious the aging degree of this transformer, and the more likely it is to have abnormal power transmission situations.

[0083] For the i-th transformer, calculate the corresponding aging degree index Li. Among them, the calculation formula of the aging degree index Li is: Li = a1×Si + a2×Gi + a3×Wi + a4×Bi, where a1, a2, a3, and a4 are respectively preset proportionality coefficients.

[0084] Preset N monitoring levels and set an aging degree threshold for each level.

[0085] Classify the i-th transformer into the corresponding monitoring level according to the aging degree index and the aging degree threshold.

[0086] In a preferred embodiment, the method of classifying into the corresponding monitoring level is: the aging degree index of the i-th transformer satisfies the condition: Yn ≤ Li < Y(n + 1); where n is the number of the monitoring level to which the i-th transformer is classified, n = 1, 2, 3,... N; Yn is the aging degree threshold of the n-th monitoring index. It can be understood that transformers in the same monitoring level have similar aging degrees.

[0087] Then the corresponding relationship between the numbers of all transformers and the monitoring levels constitutes the transformer classification data.

[0088] Furthermore, the method of collecting historical power operation data for each transformer is as follows:

[0089] Preset the sliding window time length. It should be noted that the sliding window time length represents several unit time lengths, and the unit time length is set according to actual needs, such as per second or per minute.

[0090] Every other sliding window time length, collect the power feature sequence set and the actual operation status label set of each transformer. The power feature sequence set includes power feature sequences corresponding to several power features. The length of the power feature sequence is one sliding window time length, and each element in the power feature sequence is the power feature value of each unit time. Among them, the actual operation status label represents the label value of whether each transformer is abnormal during the duration corresponding to each power feature sequence set. The label value is 0 or 1. When the transformer is working abnormally, the label value is 1, and when the transformer is working normally, the label value is 0. The label value is manually marked by the monitoring personnel.

[0091] The power characteristics include but are not limited to voltage input and output difference, current input and output difference, and power input and output difference; the voltage input and output difference, current input and output difference, and power input and output difference respectively represent the difference between input voltage and output voltage, the difference between input current and output current, and the difference between input power and output power; it can be understood that the voltage, current, and power can all be obtained in real time using corresponding sensors; and each of the power characteristic sequences can represent the efficiency and volatility of the transformer for power transmission within the time length of each sliding window;

[0092] The set of all power feature sequences corresponding to all transformers and the set of actual operation status labels constitute the historical power operation data;

[0093] The method of training a power transmission state judgment model for predicting whether power transmission is abnormal for each transformer based on historical power operation data is as follows:

[0094] For each transformer:

[0095] Each power feature sequence set corresponding to the transformer is used as the input of the transmission state judgment model. The transmission state judgment model takes the predicted abnormal probability of the transformer operation state within each sliding window time length as the output, and the value range of the predicted abnormal probability is [0,1]. The actual operation state label corresponding to the power feature sequence set of the transformer in the historical power operation data is used as the prediction target, and the sum of the prediction errors of the actual operation state label with respect to the predicted abnormal probability is minimized as the training target; the transmission state judgment model is trained until the sum of the prediction errors reaches convergence, and the training is stopped, and a transmission state judgment model that outputs the predicted operation state according to the power feature sequence set is trained; the transmission state judgment model is any one of the time series classification neural network models; the time series classification neural network model includes but is not limited to the neural network models such as RNN and LSTM; the sum of the prediction errors is any one of the mean square error or the cross entropy;

[0096] Furthermore, the method of collecting the real-time power operation data and real-time operation status data of each transformer is:

[0097] At every sliding window time length, the power feature sequences corresponding to the power features of each transformer in the time period are collected to form a power feature sequence set, and the power feature sequence set is used as real-time power operation data;

[0098] Every time the sliding window has a certain length, the operating state characteristic values ​​corresponding to the operating state characteristics of each transformer in the time period are collected to form a set of operating state characteristic value sets;

[0099] The operating state characteristics include but are not limited to the working state data measuring the performance of the transformer, such as the average temperature, the average gas pressure, and the gas release frequency; it can be understood that the fluctuation of the above-collected power characteristic sequence represents the harmful influence of the abnormality caused by the aging of the transformer, and the collected operating state characteristic value represents the observable phenomenon directly caused by the aging of the transformer, that is, the observable phenomenon can more directly reflect the aging phenomenon of the transformer, thereby providing a more accurate basis for judging whether the power characteristic fluctuation is affected by the aging of the transformer;

[0100] For each time period of the sliding window length, the set of operating status feature values ​​of all transformers constitutes real-time operating status data;

[0101] Furthermore, for each monitoring level of transformers, clustering is performed according to real-time operating status data, and the method for obtaining transformer clustering data is as follows:

[0102] Classify transformers according to monitoring levels;

[0103] For all transformers within each monitoring level:

[0104] The number of clusters k is preset, and the set of operating status feature values ​​of each transformer in the time period of each sliding window length in the real-time operating status data is taken as a data point; as an example of a data point, within a certain sliding window time period, the average temperature of a transformer oil is 60°C, the average gas pressure is 0.2 bar, and the gas release frequency is 1, then the coordinates of the data point can be expressed as (60, 0.2, 1);

[0105] All data points within the time period of each sliding window length are divided into k clusters using a clustering algorithm; it can be understood that each cluster corresponds to a number of transformers; through the setting of clusters, on the basis of dividing the transformers according to the monitoring level, they are further classified according to the performance of the operating status, that is, transformers with similar operating performance are grouped as a group, which can serve as references to each other and provide more accurate reference data; the clustering algorithm is K-means or FCM algorithm;

[0106] The serial numbers of all transformers corresponding to each cluster divided by the clustering algorithm constitute transformer clustering data;

[0107] Furthermore, the method of obtaining the predicted value of the real-time power transmission state of the transformer based on the real-time power operation data and the power transmission state judgment model set corresponding to the clustered transformer set is:

[0108] For the i-th transformer, the set of all transformers corresponding to the cluster corresponding to the transformer is marked as Ci, and the transformers in the set Ci of transformers except the i-th transformer are marked as ci;

[0109] In the time period corresponding to the length of each sliding window, the real-time power operation data corresponding to the i-th transformer is input into the power transmission state judgment model corresponding to all transformers in the transformer set Ci to obtain the predicted abnormal probability of the output;

[0110] The predicted abnormal probability of the power transmission state judgment model of the i-th transformer is marked as Pi, and the predicted abnormal probability of the power transmission state judgment model of the transformer ci is marked as Pci;

[0111] Calculate the predicted value Fi of the real-time power transmission state of the i-th transformer, the calculation formula of the predicted value Fi of the real-time power transmission state is: Fi = a5×pi+a6×∑ ci Pci; wherein a5 and a6 are preset proportional coefficients respectively; it can be understood that the predicted value Fi of the real-time power transmission status represents the abnormal prediction of the transformer with a comprehensive reference to the operating status and aging degree similar to the i-th transformer, thereby improving the accuracy of the prediction of power fluctuation anomalies and reducing the prediction error rate.

[0112] Example 2

[0113] Figure 2 1 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Figure 2 As shown, according to another aspect of the present application, an electronic device 100 is also provided. The electronic device 100 may include one or more processors and one or more memories. The memories store computer readable codes, and when the computer readable codes are executed by one or more processors, the digital monitoring method for power transmission and transformation engineering as described above may be executed.

[0114] The method or system according to the embodiment of the present application can also be used by Figure 2 The electronic device architecture shown in FIG. Figure 2As shown, the electronic device 100 may include a bus 101, one or more CPUs 102, a read-only memory (ROM) 103, a random access memory (RAM) 104, a communication port 105 connected to a network, an input / output component 106, a hard disk 107, etc. The storage device in the electronic device 100, such as the ROM 103 or the hard disk 107, may store the digital monitoring method for power transmission and transformation engineering provided in the present application. The digital monitoring method for power transmission and transformation projects may, for example, include the following steps: Step 1: Collect in advance a set of operating transformers in the transmission system, and collect historical maintenance data of each transformer; Step 2: Based on the historical maintenance data, divide the transformers into several monitoring levels to obtain transformer classification data; Step 3: For each transformer, collect historical power operation data; Step 4: Based on the historical power operation data, train a transmission status judgment model for each transformer to predict whether the power transmission is abnormal; Step 5: After the training of the transmission status judgment model is completed, collect the real-time power operation data and real-time operation status data of each transformer; Step 6: For each monitoring level of transformers, cluster them according to the real-time operation status data to obtain transformer clustering data; Step 7: Based on the real-time power operation data and the set of transmission status judgment models corresponding to the clustered transformer set, obtain a predicted value of the real-time transmission status of the transformer.

[0115] Furthermore, the electronic device 100 may further include a user interface 108. Figure 2 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 2 One or more components of an electronic device are shown.

[0116] Example 3

[0117] Figure 3 Schematic diagram of a computer-readable storage medium structure provided by an embodiment of the present application. Figure 3 As shown, a computer-readable storage medium 200 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 200. When the computer-readable instructions are executed by the processor, the digital monitoring method for power transmission and transformation engineering according to the embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium 200 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0118] In addition, according to the implementation of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, the non-transitory machine-readable storage medium stores machine-readable instructions, the machine-readable instructions can be run by a processor to execute instructions corresponding to the method steps provided by the present application, and when the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are executed.

[0119] The methods, apparatuses, and devices of the present application may be implemented in many ways. For example, the methods, apparatuses, and devices of the present application may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is for illustration only, and the steps of the method of the present application are not limited to the order specifically described above unless otherwise specified. In addition, in some embodiments, the present application may also be implemented as programs recorded in a recording medium, which include machine-readable instructions for implementing the method according to the present application. Therefore, the present application also covers recording media storing programs for executing the method according to the present application.

[0120] In addition, the parts of the above-mentioned technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0121] The specific implementation modes as described above further describe the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation mode of the present invention and is not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0122] The above preset parameters or preset thresholds are all set by technicians in this field according to actual conditions or obtained through large-scale data simulation.

[0123] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A digital monitoring method for power transmission and transformation engineering, characterized in that: The following steps are involved: Collect the operating transformers in the transmission system in advance and collect the historical maintenance data of each transformer; Based on historical maintenance data, transformers are divided into several monitoring levels to obtain transformer classification data; For each transformer, collect historical power operation data; Based on historical power operation data, a power transmission status judgment model is trained for each transformer to predict whether power transmission is abnormal; After the training of the power transmission status judgment model is completed, the real-time power operation data and real-time operation status data of each transformer are collected; For each transformer at each monitoring level, cluster the real-time operating status data to obtain transformer cluster data; Based on the real-time power operation data and the set of power transmission state judgment models corresponding to the clustered transformer set, a predicted value of the real-time power transmission state of the transformer is obtained; For each transformer, the method of collecting historical power operation data is as follows: Preset sliding window time length; Every other sliding window time length, collect the power feature sequence set and the actual operation status label set of each transformer; the power feature sequence set includes a power feature sequence corresponding to a number of power features, the length of the power feature sequence is a sliding window time length, and each element in the power feature sequence is a power feature value per unit time; wherein the actual operation status label represents the label value of whether each transformer is abnormal within the time length corresponding to each power feature sequence set, and the label value is 0 or 1; The set of all power feature sequences corresponding to all transformers and the set of actual operation status labels constitute the historical power operation data; The method of training a power transmission state judgment model for each transformer to predict whether power transmission is abnormal is as follows: For each transformer: Each power feature sequence set corresponding to the transformer is used as the input of the power transmission state judgment model. The power transmission state judgment model takes the predicted abnormal probability of the transformer operation state within each sliding window time length as the output, and the value range of the predicted abnormal probability is [0,1]. The actual operation state label corresponding to the power feature sequence set of the transformer in the historical power operation data is used as the prediction target, and the sum of the prediction errors of the predicted abnormal probability to the actual operation state label is minimized as the training target; The transmission state judgment model is trained until the sum of the prediction errors reaches convergence and then the training is stopped.

2. A digital monitoring method for power transmission and transformation engineering according to claim 1, characterized in that: The method of collecting the historical maintenance data of each transformer is as follows: Collect maintenance-related data of all transformers in the operating transformer set; The maintenance-related data include usage time, number of failures, number of repairs, and number of parts replacements; The maintenance-related data of all transformers constitute the historical maintenance data.

3. A digital monitoring method for power transmission and transformation engineering according to claim 2, characterized in that: Based on the historical maintenance data, the transformers are divided into several monitoring levels, and the transformer classification data is obtained as follows: The number of the transformer in the transformer set is marked as i; The service life of the i-th transformer is marked as Si, the number of failures is marked as Gi, the number of repairs is marked as Wi, and the number of parts replacements is marked as Bi; For the i-th transformer, calculate the corresponding aging degree index Li; where the calculation formula of the aging degree index Li is: Li = a1×Si + a2×Gi + a3×Wi + a4×Bi, where a1, a2, a3, and a4 are respectively preset proportionality coefficients; Preset N monitoring levels and set an aging degree threshold for each level; Classify the i-th transformer into the corresponding monitoring level according to the aging degree index and the aging degree threshold; The method of classifying into the corresponding monitoring level is: the aging degree index of the i-th transformer satisfies the condition: Yn ≤ Li < Y(n+1); where n is the number of the monitoring level to which the i-th transformer is classified, n = 1, 2, 3,... N; Yn is the aging degree threshold of the n-th monitoring index; The corresponding relationship between the numbers of all transformers and the monitoring levels constitutes the transformer classification data.

4. A digital monitoring method for power transmission and transformation engineering according to claim 3, characterized in that: The method of collecting the real-time power operation data and real-time operation status data of each transformer is: Every other sliding window time length, collect the power feature sequences corresponding to each power feature of each transformer within the time period of this sliding window time length, form a power feature sequence set, and use this power feature sequence set as the real-time power operation data; Every other sliding window time length, collect the operation status feature values corresponding to the operation status features of each transformer within the time period of this sliding window time length, form a set of operation status feature values; For each time period of the sliding window time length, the set of operation status feature values of all transformers constitutes the real-time operation status data.

5. A digital monitoring method for power transmission and transformation engineering according to claim 4, characterized in that: The method of obtaining the transformer clustering data is: Classify the transformers according to the monitoring levels; For all transformers within each monitoring level: Preset the number of clustering clusters k, and use the set of operation status feature values of each transformer within each time period of the real-time operation status data as a data point; Use the clustering algorithm to divide all data points within each time period of the sliding window time length into k clustering clusters; The numbers of all transformers corresponding to each clustering cluster after the clustering algorithm divides constitute the transformer clustering data.

6. A digital monitoring method for power transmission and transformation engineering according to claim 5, characterized in that: The method of obtaining the predicted value of the real-time power transmission status of the transformer is: For the i-th transformer, mark the set of all transformers corresponding to the clustering cluster to which this transformer belongs as Ci, and mark the transformers other than the i-th transformer in the set of transformers Ci as ci; Within each time period corresponding to the sliding window time length, input the real-time power operation data corresponding to the i-th transformer into the power transmission status judgment models corresponding to all transformers in the set of transformers Ci to obtain the predicted abnormal probability of the output; Mark the predicted abnormal probability of the power transmission status judgment model of the i-th transformer as Pi, and mark the predicted abnormal probability of the power transmission status judgment model of the transformer ci as Pci; Calculate the predicted value Fi of the real-time power transmission state of the i-th transformer, the calculation formula of the predicted value Fi of the real-time power transmission state is: Fi = a5×Pi+a6×∑ ci Pci; wherein a5 and a6 are preset proportional coefficients respectively.

7. An electronic device, characterized in that: Including: A processor and a memory, where: The memory stores a computer program that can be called by the processor; The processor executes the digital monitoring method for power transmission and transformation projects described in any one of claims 1 to 6 in the background by calling the computer program stored in the memory.

8. A computer-readable storage medium, characterized in that: A rewritable computer program is stored thereon; When the computer program is executed on a computer device, the computer device executes the digital monitoring method for power transmission and transformation projects as described in any one of claims 1 to 6 in the background.

Images