A data processing and control system and method for intelligent substation

Abstract

The present invention belongs to the field of data processing technology, and specifically provides a data processing and control system and method for a smart substation, the method comprising: an edge processing terminal receives the operation data of each device in the smart substation collected in real time through a sensor network; the data is pre-processed and transmitted to a data processing server; a fault diagnosis model is preset in the data processing server, and the pre-processed data is input into the fault diagnosis model to output a fault diagnosis result; the fault diagnosis model is a model for predicting substation equipment failures obtained by training a decision tree model based on historical fault sample data of the substation; and the output fault diagnosis result is transmitted to a decision control terminal; the decision control terminal matches the acquired fault diagnosis result with a protection processing mode and generates a control instruction to send to the corresponding device in the substation, and outputs fault warning information based on the fault diagnosis result. The delay of data transmission is reduced, and the real-time performance of the system is improved.

Description

Background Art

[0002] With the rapid development of smart grid technology, smart substations have gradually become an indispensable key node in the power grid system. Smart substations not only carry the task of transmitting and converting electric energy, but are also an important embodiment of the intelligence and automation of the power grid. The performance of its data processing and control system is directly related to the stable operation, power supply quality and economic benefits of the power grid.

[0003] However, traditional substation data processing systems generally adopt a centralized architecture, that is, all data is collected to a central server through a communication network for unified processing. This architecture can still cope with processing small-scale, low-frequency data, but it is powerless when faced with large-scale, high-frequency data collection in smart grids. Data transmission delay has become an important factor restricting system performance, especially in scenarios with extremely high real-time requirements. This delay may lead to the loss or misjudgment of important information. In addition, the centralized architecture also faces the problem of low processing efficiency. When all data is concentrated on the central server, the processing capacity of the server often becomes a bottleneck, resulting in a decrease in data processing speed, which cannot meet the data processing speed requirements of the smart grid.

[0004] Due to data transmission delays and low processing efficiency, traditional substation data processing systems are often unable to detect equipment failures in a timely manner, which leads to untimely and inaccurate fault handling, posing a great threat to the stable operation of the power grid. Summary of the invention

[0005] Traditional substation data processing and control systems can no longer meet the requirements of modern smart grids for efficient and reliable operation. The present invention provides a smart substation data processing and control system.

[0006] In a first aspect, the technical solution of the present invention provides a smart substation data processing and control system, including an edge processing terminal, a data processing server and a decision control terminal; the number of edge processing terminals is several;

[0007] The edge processing terminal receives the operating data of each device in the smart substation collected in real time through the sensor network; the collected data is pre-processed and transmitted to the data processing server;

[0008] A fault diagnosis model is preset in the data processing server, and the pre-processed data is input into the fault diagnosis model to output a fault diagnosis result; the fault diagnosis model is a model for predicting substation equipment failures obtained by training a decision tree model based on historical fault sample data of the substation; and the output fault diagnosis result is transmitted to the decision control terminal;

[0009] The decision control terminal is preset with protection processing methods corresponding to different fault types; the decision control terminal matches the protection processing method corresponding to the fault diagnosis result according to the acquired fault diagnosis result and generates a control instruction to send to the corresponding equipment in the substation, and outputs fault warning information based on the fault diagnosis result.

[0010] As a further limitation of the technical solution of the present invention, the edge processing terminal includes a data receiving module, a data preprocessing module and a data transmission module;

[0011] The data receiving module receives the operation monitoring data from each device in the smart substation transmitted through the sensor network in real time;

[0012] The data preprocessing module preprocesses the received data, including data noise reduction and normalization;

[0013] The data transmission module transmits the preprocessed data to the data processing server according to the set protocol.

[0014] As a further limitation of the technical solution of the present invention, the process of data noise reduction by the data preprocessing module is as follows:

[0015] The substation equipment operation monitoring data is divided into M modal components, and the energy density of the nth modal component is calculated. and average period To calculate the modal component noise threshold ;

[0016]

[0017] D is the sampling density; V is the sampling volume; k is the amplitude coefficient; In order to remove the time required for noise contained in the substation equipment operation monitoring data, when the change in the modal component noise threshold is greater than the set threshold, the n-1th modal component is noise. The modal component noise threshold is replaced by the n-1th modal component to reconstruct the modal component of the substation equipment operation monitoring data to obtain noise-free substation equipment operation monitoring data.

[0018] As a further limitation of the technical solution of the present invention, the process of generating a fault diagnosis model by a data processing server includes: obtaining historical operating status data of the substation to perform data preprocessing to extract fault features and generate a fault sample data set; dividing the fault sample data set into a training data set and a test data set; using the C5.0 decision tree algorithm to calculate the information gain rate of the fault features of the training data set, and selecting the feature attribute with the largest information gain value as the root node partitioning attribute of the decision tree model; in the branches generated after the root node partition, recursively dividing each branch node using the same method as the root node partition; when all samples of a branch node after the parent node partition are of the same category, stop further partitioning and use the node as a leaf node to generate a trained decision tree; input the test set data into the trained decision tree, the nodes and edges in the decision tree divide each test set data into corresponding categories, and obtain the final fault diagnosis result, and evaluate the model performance by analyzing the amount of data diagnosed correctly and incorrectly in the test set, and output the model with the best performance as the fault diagnosis model.

[0019] As a further limitation of the technical solution of the present invention, the data processing server performs the following process to calculate the gain rate of each fault feature:

[0020] Calculate the empirical entropy H(R) of the fault sample data set:

[0021]

[0022] Calculate the empirical conditional entropy H(R|A) of fault feature A for the training data set:

[0023]

[0024] Calculate information gain:

[0025]

[0026] Calculate the entropy of fault feature A:

[0027]

[0028] Calculate the information gain rate of fault feature A for the sample data set:

[0029]

[0030] In the formula, R is the fault sample data set, |R| is the number of samples, and F types of fault type samples are ,| | is the number of samples of each type of fault, k=1,2,…,F; F | |The sum is equal to the number of samples |R|; R is divided into N subsets according to the value of the fault feature, and the i-th subset is ,| |For The number of samples, for Fault type A sample collection of | |For The number of samples.

[0031] As a further limitation of the technical solution of the present invention, the system also includes a connection control server, the edge processing terminal sends registration information to the connection control server, the connection control server issues a connection credential to the edge processing terminal based on the legitimacy of the registration information, and sends the connection credential to the data processing server for storage; the connection credential includes the identity ID and authentication information of the edge processing terminal;

[0032] The edge processing terminal sends a connection request to the data processing server through the connection credential. The data processing server compares the received connection credential with the stored connection credential. When the comparison is consistent, a connection channel is established to send the request response information to the edge processing terminal. The request response information includes the channel information assigned to the edge processing terminal. The edge processing terminal returns the connectivity information to the connection control server based on the received request response information. The connection control server stores the information of the edge processing terminal that has established a connection with the data processing server in the connection list based on the connectivity information.

[0033] As a further limitation of the technical solution of the present invention, the connection control server includes a transmission monitoring module, which monitors the transmission status of data in real time during the data transmission process; when it is determined according to the transmission status information that there is an abnormality in the data transmission of the edge processing terminal, the sensor network connected to the abnormal edge processing terminal is migrated to the edge processing terminal closest to the abnormal edge processing terminal;

[0034] When the edge processing terminal returns to normal, it re-sends a connection request to the data processing server through the connection credential. After re-establishing the connection with the data processing server, the connection control server migrates the corresponding sensor network back to the edge processing terminal that has returned to normal.

[0035] As a further limitation of the technical solution of the present invention, two queues are created inside the edge processing terminal receiving the migration to obtain the smart substation data transmitted by the corresponding sensor network, and each queue corresponds to a channel connected to the data processing server.

[0036] As a further limitation of the technical solution of the present invention, the transmission monitoring module continuously monitors the interaction process between the edge processing terminal and the data processing server, performs protocol compliance analysis, traffic analysis, interaction frequency analysis port monitoring, and performs abnormality monitoring by listening to the edge processing terminal data bus.

[0037] In a second aspect, the technical solution of the present invention provides a smart substation data processing and control method, comprising:

[0038] The edge processing terminal receives the operating data of each device in the smart substation collected in real time through the sensor network; the collected data is pre-processed and transmitted to the data processing server;

[0039] The data processing server inputs the preprocessed data into the fault diagnosis model to output the fault diagnosis result; the fault diagnosis model is a model for predicting the fault type of substation equipment obtained by training the decision tree model based on the historical fault sample data of the substation; and transmits the output fault diagnosis result to the decision control terminal;

[0040] The decision control terminal matches the protection processing method corresponding to the fault diagnosis result according to the acquired fault diagnosis result and generates a control instruction and sends it to the corresponding equipment in the substation, and outputs fault warning information based on the fault diagnosis result.

[0041] As a further limitation of the technical solution of the present invention, the edge processing terminal receives the operation data and status information of each device in the smart substation collected in real time through the sensor network; and the step of transmitting the collected data to the data processing server specifically includes:

[0042] The processing terminal includes a data receiving module, a data preprocessing module and a data transmission module;

[0043] The data receiving module receives the operation monitoring data from each device in the smart substation transmitted through the sensor network in real time;

[0044] The data preprocessing module preprocesses the received data, including data noise reduction and normalization;

[0045] The data transmission module transmits the preprocessed data to the data processing server according to the set protocol.

[0046] As a further limitation of the technical solution of the present invention, the method further includes:

[0047] The edge processing terminal sends registration information to the connection control server. The connection control server issues a connection credential to the edge processing terminal based on the legitimacy of the registration information, and sends the connection credential to the data processing server for storage. The connection credential includes the identity ID and authentication information of the edge processing terminal.

[0048] The edge processing terminal sends a connection request to the data processing server through the connection credential. The data processing server compares the received connection credential with the stored connection credential. When the comparison is consistent, a connection channel is established to send the request response information to the edge processing terminal. The request response information includes the channel information assigned to the edge processing terminal. The edge processing terminal returns the connectivity information to the connection control server based on the received request response information. The connection control server stores the information of the edge processing terminal that has established a connection with the data processing server in the connection list based on the connectivity information.

[0049] As a further limitation of the technical solution of the present invention, the data transmission status is monitored in real time during the data transmission process; when it is determined according to the transmission status information that there is an edge processing terminal data transmission abnormality, the sensor network connected to the abnormal edge processing terminal is migrated to the edge processing terminal closest to the abnormal edge processing terminal;

[0050] When the edge processing terminal returns to normal, it re-sends a connection request to the data processing server through the connection credential. After re-establishing the connection with the data processing server, the connection control server migrates the corresponding sensor network back to the edge processing terminal that has returned to normal.

[0051] As a further limitation of the technical solution of the present invention, two queues are created inside the edge processing terminal receiving the migration to obtain the smart substation data transmitted by the corresponding sensor network, and each queue corresponds to a channel connected to the data processing server.

[0052] As a further limitation of the technical solution of the present invention, the transmission monitoring module continuously monitors the interaction process between the edge processing terminal and the data processing server, performs protocol compliance analysis, traffic analysis, interaction frequency analysis port monitoring, and performs abnormality monitoring by listening to the edge processing terminal data bus.

[0053] It can be seen from the above technical solutions that the present invention has the following advantages: the edge processing terminal collects and processes the operating data of each device in the substation in real time, greatly reducing the delay of data transmission and improving the real-time performance and response speed of the system. At the same time, the introduction of the edge processing terminal enables the data preprocessing tasks to be distributed to multiple terminals, avoiding the processing bottleneck of the data processing server and improving the overall processing efficiency of the system.

[0054] The fault diagnosis model preset in the data processing server is obtained by training the decision tree model based on the historical fault sample data of the substation, and can accurately predict the fault conditions of the substation equipment. Compared with the traditional fault diagnosis method based on expert experience and manual judgment, the present invention has a higher level of intelligence and accuracy. The data-driven fault diagnosis method adopted does not depend on specific equipment or systems, and has strong versatility and scalability. By updating the training data of the fault diagnosis model, it can adapt to the specific needs of different substations. The decision control terminal matches the corresponding protection processing method according to the fault diagnosis results and generates control instructions, which can realize timely protection and early warning of faulty equipment.

[0055] By introducing the connection control server, the system can realize the registration and authentication management of edge processing terminals, ensuring that only legitimate terminals can access the system, improving the security of the system. At the same time, the transmission monitoring module can monitor the data transmission status in real time, detect and handle abnormal situations in time, and ensure the stability and reliability of data transmission.

[0056] In addition, the invention has a reliable design principle, a simple structure and a very broad application prospect. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Figure 1 is a schematic block diagram of a system according to an embodiment of the present invention.

[0059] Figure 2 is a schematic flow chart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not 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 should fall within the scope of protection of the present invention.

[0061] like Figure 1 As shown, an embodiment of the present invention provides a smart substation data processing and control system, including an edge processing terminal, a data processing server and a decision control terminal; the number of edge processing terminals is several;

[0062] It includes edge processing terminals, data processing servers and decision control terminals; the number of edge processing terminals is several;

[0063] The edge processing terminal receives the operating data of each device in the smart substation collected in real time through the sensor network; the collected data is pre-processed and transmitted to the data processing server;

[0064] A fault diagnosis model is preset in the data processing server, and the pre-processed data is input into the fault diagnosis model to output a fault diagnosis result; the fault diagnosis model is a model for predicting substation equipment failures obtained by training a decision tree model based on historical fault sample data of the substation; and the output fault diagnosis result is transmitted to the decision control terminal;

[0065] The decision control terminal is preset with protection processing methods corresponding to different fault types; the decision control terminal matches the protection processing method corresponding to the fault diagnosis result according to the acquired fault diagnosis result and generates a control instruction to send to the corresponding equipment in the substation, and outputs fault warning information based on the fault diagnosis result.

[0066] In the embodiment of the present invention, the fault type is diagnosed by inputting the monitoring data acquired in real time into the fault diagnosis model, and the location of the fault is determined by the location of the acquisition device in the sensor network that provides the monitoring data. Fault warning information is output according to the fault diagnosis result, and the warning information includes the fault type and the fault location. If a protection processing method is matched, the warning information also includes the matched protection processing method.

[0067] It should be noted in the embodiments of the present invention that the decision control terminal will continuously monitor the diagnostic results sent by the data processing server. These results usually include the type of equipment operating status fault. The decision control terminal is provided with a preset protection processing method. Most of the fault protection measures are concentrated on the input and output sides of the substation equipment, such as phase loss protection, overcurrent protection and overvoltage protection on the input side of the equipment, overcurrent protection, overvoltage protection, and overtemperature protection on the output side of the equipment. The preset protection processing method can only play a basic protection role. The decision control terminal also outputs early warning information to prompt the staff to handle it.

[0068] After receiving the diagnosis results, the decision control terminal generates corresponding control instructions. These instructions may include specific operations such as equipment start and stop, parameter adjustment, fault isolation, etc. The decision control terminal will send the control instructions to the corresponding equipment in the smart substation through the communication network. This communication network usually uses high-speed, reliable wired or wireless methods to ensure real-time transmission and accurate execution of instructions. When sending instructions, the decision control terminal will use technical means such as encryption and verification to ensure the security and integrity of the instructions.

[0069] After the command is sent, the decision control terminal will continue to monitor the execution of the equipment and receive feedback information from the equipment. This information includes key data such as the execution results of the command and the real-time status of the equipment. By analyzing these data, the decision control terminal can grasp the operating status and control effect of the equipment in real time, providing important reference for subsequent operations. If the equipment encounters an abnormality or failure during the execution of the command, the decision control terminal will immediately activate the emergency response mechanism. This includes re-sending the command, switching to backup equipment, and starting the fault isolation program to ensure the stable operation of the power grid and the safety and reliability of the equipment.

[0070] In some embodiments, the edge processing terminal includes a data receiving module, a data preprocessing module, and a data transmission module;

[0071] The data receiving module receives the operation monitoring data from each device in the smart substation transmitted through the sensor network in real time;

[0072] The data preprocessing module preprocesses the received data, including data noise reduction and normalization processing; the normalization processing in the embodiment of the present invention is implemented by adopting the existing normalization processing method.

[0073] The data transmission module transmits the preprocessed data to the data processing server according to the set protocol.

[0074] In some embodiments, the process of data denoising by the data preprocessing module is as follows:

[0075] The substation equipment operation monitoring data is divided into M modal components, and the energy density of the nth modal component is calculated. and average period To calculate the modal component noise threshold ;

[0076]

[0077] D is the sampling density; V is the sampling volume; k is the amplitude coefficient; In order to remove the time required for noise contained in the substation equipment operation monitoring data, when the change in the modal component noise threshold is greater than the set threshold, the n-1th modal component is noise. The modal component noise threshold is replaced by the n-1th modal component to reconstruct the modal component of the substation equipment operation monitoring data to obtain noise-free substation equipment operation monitoring data.

[0078] In some embodiments, the process of generating a fault diagnosis model by a data processing server includes: obtaining historical operating status data of a substation, performing data preprocessing, extracting fault features, and generating a fault sample data set; dividing the fault sample data set into a training data set and a test data set; using the C5.0 decision tree algorithm to calculate the information gain rate of the fault features of the training data set, and selecting the feature attribute with the largest information gain value as the root node partitioning attribute of the decision tree model; in the branches generated after the root node partition, recursively performing the same method as the root node partitioning for each branch node; when all samples of a branch node after the parent node partition are of the same category, stop further partitioning and use the node as a leaf node to generate a trained decision tree; input the test set data into the trained decision tree, and the nodes and edges in the decision tree divide each test set data into corresponding categories to obtain the final fault diagnosis result, and evaluate the model performance by analyzing the amount of data diagnosed correctly and incorrectly in the test set, and the model with a parameter value greater than the set value is the optimal model.

[0079] In some embodiments, the data processing server calculates the gain rate of each fault feature as follows:

[0080] Calculate the empirical entropy H(R) of the fault sample data set:

[0081]

[0082] Calculate the empirical conditional entropy H(R|A) of fault feature A for the training data set:

[0083]

[0084] Calculate information gain:

[0085]

[0086] Calculate the entropy of fault feature A:

[0087]

[0088] Calculate the information gain rate of fault feature A for the sample data set:

[0089]

[0090] In the formula, R is the fault sample data set, |R| is the number of samples, and F types of fault type samples are ,| | is the number of samples of each type of fault, k=1,2,…,F; F | |The sum is equal to the number of samples |R|; R is divided into N subsets according to the value of the fault feature, and the i-th subset is ,| |For The number of samples, for Fault type A sample collection of | |For The number of samples.

[0091] In the embodiment of the present invention, the process of evaluating the model includes: calculating the accuracy, recall and F1 score; the accuracy Ass is the ratio between the number of samples correctly classified by the model and the total number of samples;

[0092] The recall rate Rss measures the proportion of all true positive examples that are correctly identified by the model;

[0093] The F1 score is the harmonic mean of precision and recall.

[0094]

[0095]

[0096]

[0097] Wherein, TN represents the number of true negative examples, TP represents the number of true positive examples, FN represents the number of false negative examples, and FP represents the number of false positive examples. In the embodiment of the present invention, the model whose F1 score of the validation set of the model is greater than the maximum value of the current F1 score is the fault diagnosis model.

[0098] In some embodiments, the system further includes a connection control server, the edge processing terminal sends registration information to the connection control server, the connection control server issues a connection credential to the edge processing terminal based on the legitimacy of the registration information, and sends the connection credential to the data processing server for storage; the connection credential includes an identity ID and authentication information of the edge processing terminal;

[0099] The edge processing terminal sends a connection request to the data processing server through the connection credential. The data processing server compares the received connection credential with the stored connection credential. When the comparison is consistent, a connection channel is established to send the request response information to the edge processing terminal. The request response information includes the channel information assigned to the edge processing terminal. The edge processing terminal returns the connectivity information to the connection control server based on the received request response information. The connection control server stores the information of the edge processing terminal that has established a connection with the data processing server in the connection list based on the connectivity information.

[0100] In some embodiments, the connection control server includes a transmission monitoring module, which monitors the transmission status of data in real time during the data transmission process; when it is determined according to the transmission status information that there is an abnormality in the data transmission of the edge processing terminal, the sensor network connected to the abnormal edge processing terminal is migrated to the edge processing terminal closest to the abnormal edge processing terminal;

[0101] When the edge processing terminal returns to normal, it re-sends a connection request to the data processing server through the connection credential. After re-establishing the connection with the data processing server, the connection control server migrates the corresponding sensor network back to the edge processing terminal that has returned to normal.

[0102] In some embodiments, two queues are created inside the edge processing terminal receiving the migration to obtain the smart substation data transmitted by the corresponding sensor network, and each queue corresponds to a channel connected to the data processing server.

[0103] In some embodiments, the transmission monitoring module continuously monitors the interaction process between the edge processing terminal and the data processing server, performs protocol compliance analysis, traffic analysis, interaction frequency analysis, port monitoring, and abnormality monitoring by listening to the edge processing terminal data bus.

[0104] like Figure 2 As shown, an embodiment of the present invention provides a smart substation data processing and control method, including:

[0105] Step 1: The edge processing terminal receives the operating data of each device in the smart substation collected in real time through the sensor network; the collected data is pre-processed and transmitted to the data processing server;

[0106] Specifically include:

[0107] The data receiving module receives the operation monitoring data from each device in the smart substation transmitted through the sensor network in real time;

[0108] The data preprocessing module preprocesses the received data, including data noise reduction and normalization;

[0109] The data transmission module transmits the preprocessed data to the data processing server according to the set protocol.

[0110] Step 2: A fault diagnosis model is preset in the data processing server, and the preprocessed data is input into the fault diagnosis model to output a fault diagnosis result; the fault diagnosis model is a model for predicting substation equipment failures obtained by training a decision tree model based on historical fault sample data of the substation; and the output fault diagnosis result is transmitted to the decision control terminal.

[0111] Step 3: The decision control terminal is preset with protection processing methods corresponding to different fault types; the decision control terminal matches the protection processing method corresponding to the fault diagnosis result according to the acquired fault diagnosis result and generates a control instruction to send to the corresponding equipment in the substation, and outputs fault warning information based on the fault diagnosis result.

[0112] In some embodiments, the process of data denoising by the data preprocessing module is as follows:

[0113] The substation equipment operation monitoring data is divided into M modal components, and the energy density of the nth modal component is calculated. and average period To calculate the modal component noise threshold ;

[0114]

[0115] D is the sampling density; V is the sampling volume; k is the amplitude coefficient; In order to remove the time required for noise contained in the substation equipment operation monitoring data, when the change in the modal component noise threshold is greater than the set threshold, the n-1th modal component is noise. The modal component noise threshold is replaced by the n-1th modal component to reconstruct the modal component of the substation equipment operation monitoring data to obtain noise-free substation equipment operation monitoring data.

[0116] In some embodiments, the process of generating a fault diagnosis model by a data processing server includes: obtaining historical operating status data of a substation, performing data preprocessing, extracting fault features, and generating a fault sample data set; dividing the fault sample data set into a training data set and a test data set; using the C5.0 decision tree algorithm to calculate the information gain rate of the fault features of the training data set, and selecting the feature attribute with the largest information gain value as the root node partitioning attribute of the decision tree model; in the branches generated after the root node partition, recursively performing the same method as the root node partitioning for each branch node; when all samples of a branch node after the parent node partition are of the same category, stop further partitioning and use the node as a leaf node to generate a trained decision tree; input the test set data into the trained decision tree, and the nodes and edges in the decision tree divide each test set data into corresponding categories to obtain the final fault diagnosis result, and evaluate the model performance by analyzing the amount of data diagnosed correctly and incorrectly in the test set, and output the model with the best performance as the fault diagnosis model.

[0117] In some embodiments, the data processing server calculates the gain rate of each fault feature as follows:

[0118] Calculate the empirical entropy H(R) of the fault sample data set:

[0119]

[0120] Calculate the empirical conditional entropy H(R|A) of fault feature A for the training data set:

[0121]

[0122] Calculate information gain:

[0123]

[0124] Calculate the entropy of fault feature A:

[0125]

[0126] Calculate the information gain rate of fault feature A for the sample data set:

[0127]

[0128] In the formula, R is the fault sample data set, |R| is the number of samples, and F types of fault type samples are ,| | is the number of samples of each type of fault, k=1,2,…,F; F | |The sum is equal to the number of samples |R|; R is divided into N subsets according to the value of the fault feature, and the i-th subset is ,| |For The number of samples, for Fault type A sample collection of | |For The number of samples.

[0129] In the embodiment of the present invention, the process of evaluating the model includes:

[0130] Accuracy Ass is the ratio between the number of samples correctly classified by the model and the total number of samples;

[0131] The recall rate Rss measures the proportion of all true positive examples that are correctly identified by the model;

[0132] The F1 score is the harmonic mean of precision and recall.

[0133]

[0134]

[0135]

[0136] Wherein, TN represents the number of true negative examples, TP represents the number of true positive examples, FN represents the number of false negative examples, and FP represents the number of false positive examples. In the embodiment of the present invention, the model whose F1 score of the validation set of the model is greater than the maximum value of the current F1 score is the fault diagnosis model.

[0137] In some embodiments, the method further comprises:

[0138] The edge processing terminal sends registration information to the connection control server. The connection control server issues a connection credential to the edge processing terminal based on the legitimacy of the registration information, and sends the connection credential to the data processing server for storage. The connection credential includes the identity ID and authentication information of the edge processing terminal.

[0139] The edge processing terminal sends a connection request to the data processing server through the connection credential. The data processing server compares the received connection credential with the stored connection credential. When the comparison is consistent, a connection channel is established to send the request response information to the edge processing terminal. The request response information includes the channel information assigned to the edge processing terminal. The edge processing terminal returns the connectivity information to the connection control server based on the received request response information. The connection control server stores the information of the edge processing terminal that has established a connection with the data processing server in the connection list based on the connectivity information.

[0140] In some embodiments, the data transmission status is monitored in real time during the data transmission process; when it is determined according to the transmission status information that there is an edge processing terminal data transmission abnormality, the sensor network connected to the abnormal edge processing terminal is migrated to the edge processing terminal closest to the abnormal edge processing terminal;

[0141] When the edge processing terminal returns to normal, it re-sends a connection request to the data processing server through the connection credential. After re-establishing the connection with the data processing server, the connection control server migrates the corresponding sensor network back to the edge processing terminal that has returned to normal.

[0142] In some embodiments, two queues are created inside the edge processing terminal receiving the migration to obtain the smart substation data transmitted by the corresponding sensor network, and each queue corresponds to a channel connected to the data processing server.

[0143] In some embodiments, the transmission monitoring module continuously monitors the interaction process between the edge processing terminal and the data processing server, performs protocol compliance analysis, traffic analysis, interaction frequency analysis, port monitoring, and abnormality monitoring by listening to the edge processing terminal data bus.

[0144] In some instances, the method for generating a decision tree model also includes: preprocessing and extracting features from the historical data to obtain a corresponding training data set; calculating the importance of each feature in the training data set based on the contribution of the features in the training data set; selecting a target feature type based on the importance of each feature; dividing the training data set into multiple sub-training data sets based on the target feature type, and recursively constructing a decision tree based on the multiple sub-training data sets until a stopping condition is met to obtain a constructed decision tree; and optimizing the decision tree using an optimization algorithm to generate the trained decision tree model.

[0145] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0146] The embodiment of the smart substation data processing and control method provided in the embodiment of the present invention belongs to the same inventive concept as the smart substation data processing and control system of the above-mentioned embodiments. For details not described in detail in the embodiment of the smart substation data processing and control method, reference can be made to the embodiment of the above-mentioned smart substation data processing and control system.

[0147] The smart substation data processing and control system is a combination of the units and algorithm steps of each example described in the embodiments disclosed herein, which can be implemented by electronic hardware, computer software or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0148] Those skilled in the art will appreciate that various aspects of the smart substation data processing and control method can be implemented as a system, method or program product. Therefore, various aspects of the present disclosure can be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software, which can be collectively referred to as "circuit", "module" or "system".

[0149] Although the present invention has been described in detail with reference to the accompanying drawings and in combination with the preferred embodiments, the present invention is not limited thereto. Without departing from the spirit and essence of the present invention, a person of ordinary skill in the art may make various equivalent modifications or substitutions to the embodiments of the present invention, and these modifications or substitutions shall be within the scope of the present invention. Any person of ordinary skill in the art may easily think of changes or substitutions within the technical scope disclosed by the present invention, and these shall be within the scope of protection of the present invention.

Claims

1. A smart substation data processing and control system, It is characterized in that It includes edge processing terminals, data processing servers and decision control terminals; the number of edge processing terminals is several; The edge processing terminal receives the operating data of each device in the smart substation collected in real time through the sensor network; the collected data is pre-processed and transmitted to the data processing server; A fault diagnosis model is preset in the data processing server, and the pre-processed data is input into the fault diagnosis model to output a fault diagnosis result; the fault diagnosis model is a model for predicting substation equipment failures obtained by training a decision tree model based on historical fault sample data of the substation; and the output fault diagnosis result is transmitted to the decision control terminal; The decision control terminal is preset with protection processing methods corresponding to different fault types; the decision control terminal matches the protection processing method corresponding to the fault diagnosis result according to the acquired fault diagnosis result and generates a control instruction to send to the corresponding equipment in the substation, and outputs fault warning information based on the fault diagnosis result; The edge processing terminal includes a data receiving module, a data preprocessing module and a data transmission module; The data receiving module receives the operation monitoring data from each device in the smart substation transmitted through the sensor network in real time; The data preprocessing module preprocesses the received data, including data noise reduction and normalization; The data transmission module transmits the preprocessed data to the data processing server according to the set protocol; The process of data denoising by the data preprocessing module is as follows: The substation equipment operation monitoring data is divided into M modal components, and the energy density of the nth modal component is calculated. and average period To calculate the modal component noise threshold ; D is the sampling density; V is the sampling volume; k is the amplitude coefficient; In order to remove the time required for noise contained in the substation equipment operation monitoring data, when the change in the modal component noise threshold is greater than the set threshold, the n-1th modal component is noise, and the modal component noise threshold is replaced by the n-1th modal component to reconstruct the modal component of the substation equipment operation monitoring data to obtain noise-free substation equipment operation monitoring data; The system also includes a connection control server, the edge processing terminal sends registration information to the connection control server, the connection control server issues a connection credential to the edge processing terminal based on the legitimacy of the registration information, and sends the connection credential to the data processing server for storage; the connection credential includes the identity ID and authentication information of the edge processing terminal; The edge processing terminal sends a connection request to the data processing server through the connection credential. The data processing server compares the received connection credential with the stored connection credential. When the comparison is consistent, a connection channel is established to send the request response information to the edge processing terminal. The request response information includes channel information assigned to the edge processing terminal. The edge processing terminal returns connectivity information to the connection control server based on the received request response information. The connection control server stores information of the edge processing terminal connected to the data processing server into a connection list based on the connectivity information. The connection control server includes a transmission monitoring module, which monitors the transmission status of data in real time during the data transmission process; When it is determined according to the transmission status information that there is an edge processing terminal data transmission anomaly, the sensor network connected to the abnormal edge processing terminal is migrated to the edge processing terminal closest to the abnormal edge processing terminal; When the edge processing terminal returns to normal, it re-sends a connection request to the data processing server through the connection credential. After re-establishing the connection with the data processing server, the connection control server migrates the corresponding sensor network back to the edge processing terminal that has returned to normal.

2. According to claim 1, the smart substation data processing and control system, It is characterized in that The process of generating a fault diagnosis model by the data processing server includes: Obtain the historical operating status data of the substation, perform data preprocessing, extract fault features, and generate a fault sample data set; divide the fault sample data set into a training data set and a test data set; The C5.0 decision tree algorithm is used to calculate the information gain rate of the fault features of the training data set, and the feature attribute with the largest information gain value is selected as the root node partition attribute of the decision tree model; in the branches generated after the root node partition, each branch node is recursively divided using the same method as the root node; when all the samples of a branch node after the parent node partition are of the same category, the partition is stopped and the node is used as a leaf node to generate a trained decision tree; The test set data is input into the trained decision tree. The nodes and edges in the decision tree divide each test set data into corresponding categories to obtain the final fault diagnosis results. By analyzing the amount of correctly and incorrectly diagnosed data in the test set, the model performance is evaluated and the model with the best performance is output as the fault diagnosis model.

3. According to claim 2, the smart substation data processing and control system, It is characterized in that Two queues are created inside the edge processing terminal that receives the migration to obtain the smart substation data transmitted by the corresponding sensor network. Each queue corresponds to a channel connected to the data processing server.

4. According to claim 3, the smart substation data processing and control system, It is characterized in that The transmission monitoring module continuously monitors the interaction process between the edge processing terminal and the data processing server, and performs protocol compliance analysis, traffic analysis, interaction frequency analysis, port monitoring, and abnormality monitoring by listening to the edge processing terminal data bus.

5. A smart substation data processing and control method applied to the system according to any one of claims 1 to 4, It is characterized in that include: The edge processing terminal receives the operating data of each device in the smart substation collected in real time through the sensor network; The collected data is pre-processed and transmitted to the data processing server; The data processing server inputs the preprocessed data into the fault diagnosis model and outputs the fault diagnosis result; the fault diagnosis model is a model for predicting the fault type of substation equipment obtained by training the decision tree model based on the historical fault sample data of the substation; and transmits the output fault diagnosis result to the decision control terminal; The decision control terminal matches the protection processing method corresponding to the fault diagnosis result according to the acquired fault diagnosis result and generates a control instruction and sends it to the corresponding equipment in the substation, and outputs fault warning information based on the fault diagnosis result.

6. The smart substation data processing and control method according to claim 5, It is characterized in that The steps of generating a fault diagnosis model by the data processing server include: The historical operating status data of the substation is obtained for data preprocessing to extract fault features and generate a fault sample data set; the fault sample data set is divided into a training data set and a test data set; the C5.0 decision tree algorithm is used to calculate the information gain rate of the fault features of the training data set, and the feature attribute with the largest information gain value is selected as the root node partition attribute of the decision tree model; in the branches generated after the root node partition, each branch node is recursively divided using the same method as the root node partition; when all the samples of a branch node after the parent node partition are of the same category, the partition is stopped and the node is used as a leaf node to generate a trained decision tree; the test set data is input into the trained decision tree, and the nodes and edges in the decision tree divide each test set data into corresponding categories to obtain the final fault diagnosis result. By analyzing the amount of data diagnosed correctly and incorrectly in the test set, the model performance is evaluated and the model with the best performance is output as the fault diagnosis model.

Images