LS-SVM power distribution fault monitoring method and system based on edge node information

Abstract

The invention relates to an LS-SVM power distribution fault monitoring method and system based on edge node information, and belongs to the technical field of power distribution fault monitoring. According to the method, data collected by terminal equipment is processed and analyzed in real time by mainly utilizing edge nodes in a power distribution network, and faults of a power distribution system are monitored through an LS-SVM algorithm. The method mainly comprises the steps that an edge node collects operation data of a power distribution system, data information is transmitted to a cloud end, an LS-SVM algorithm is utilized to conduct learning training on standardized data, an LS-SVM fault prediction model is constructed, the trained model is deployed to the cloud end, data collected in real time is obtained, and therefore power distribution system fault monitoring is conducted. And an alarm can be timely sent to a user or operation and maintenance personnel, so that fault processing can be efficiently realized. The power distribution fault can be quickly and accurately identified and accurately positioned, the intelligent level and the operation efficiency of a power grid are improved, the manual inspection and maintenance cost is reduced, and the method has important significance on safe and stable operation of a power distribution system.

Description

Technical Field

[0001] The invention belongs to the technical field of power distribution fault monitoring, and in particular to an LS-SVM power distribution fault monitoring method and system based on edge node information. Background Art

[0002] With the increasing demand for electricity and the continuous expansion of the power grid, the distribution system, as the end of the power system, has a direct impact on the user's power demand if its operating status is stable. However, due to various reasons, the distribution system often fails, such as switchgear failure, distribution line short circuit or grounding fault, over-tripping events, etc.

[0003] At present, traditional power distribution fault location methods often rely on manual inspections, which have the problems of long fault location time and low accident handling efficiency, seriously affecting the user's power experience. Existing monitoring and positioning devices often have a certain lag when monitoring the operating status of distribution lines, making it difficult to achieve real-time monitoring, and it is also difficult to effectively detect and accurately diagnose transient line faults and high-impedance faults with relatively unclear characteristics.

[0004] Therefore, developing an efficient and intelligent distribution fault monitoring method is of great significance to improving the operating efficiency and safety of the distribution system. Summary of the invention

[0005] In view of the above problems, the present invention provides a LS-SVM distribution fault monitoring method and system based on edge node information. The method mainly uses the edge nodes in the distribution network to process and analyze the data collected by the terminal equipment in real time, and monitors the faults of the distribution system through the LS-SVM algorithm. It mainly includes the edge node collecting the operation data of the distribution system, transmitting the data information to the cloud, using the LS-SVM algorithm to learn and train the standardized data, constructing the LS-SVM fault prediction model, deploying the trained model to the cloud, obtaining the real-time collected data, so as to monitor the distribution system faults, and timely send alarms to users or operation and maintenance personnel, so as to efficiently handle the faults. The present invention can quickly and accurately identify distribution faults and accurately locate faults, improve the intelligence level and operation efficiency of the power grid, reduce the cost of manual inspection and maintenance, and is of great significance to the safe and stable operation of the distribution system.

[0006] The technical solution of the present invention is as follows: a LS-SVM distribution fault monitoring method based on edge node information, comprising the following steps: S1 installs sensors at the edge nodes of the power distribution system to collect parameter data such as voltage, current, power factor, temperature, etc. in real time, and uploads the collected data to the cloud; The data collected by the S2 sensor can be transmitted wirelessly to a database in the cloud, which has a storage function and stores the collected data in the database; S3 selects the power distribution system fault data and the power distribution system normal operation data from the stored data, converts the parameters therein into a standardized form, and then performs labeling processing on the data; S4: The labeled data is divided into a training set and a test set for training the LS-SVM model. The LS-SVM distribution fault monitoring model has the following classification output function: Among them, h(x) is the xth monitoring value obtained by monitoring; a=[a1,a2,...a n ] T is the Lagrange multiplier; K(x i ,x) is the kernel function; n is the number of samples in the training set; b is the bias vector; S5 deploys the LS-SVM power distribution fault diagnosis model to the cloud, receives edge node information data transmitted to the cloud, and then realizes real-time monitoring of faults.

[0007] Furthermore, in step S3, the conversion into a standardized form is specifically performed by the formula: The data is processed by Z-score standardization, where x* is the data after standardization, x is the data before standardization, and u is the mean of this type of samples. is the standard deviation of the sample.

[0008] Furthermore, in step S3, the labeling process is specifically to label the data according to the category to which it belongs, with normal operation data being labeled as 0 and fault data being labeled as 1.

[0009] Furthermore, in step S4, the model training is specifically to bring the data sample into the model to generate a training set: , x i is the input vector, y i is the tag value of the data, through the mapping function Process the data, project it to obtain a higher-order data set, and construct the decision function of the LS-SVM model: in, and b are the regression parameters of the LS-SVM model, represents the weight vector and b represents the bias vector.

[0010] Furthermore, the LS-SVM model adds a slack variable e to the constraints. i Then we get its function: in, is the penalty factor, e i is the error, and the Lagrange multiplier is substituted into the equation: Where a=[a1,a2,...,a n ] T is the Lagrange multiplier.

[0011] Furthermore, the optimal b and a can be obtained through the KKT condition, { Eliminate variables and e i , the optimization problem is transformed into solving a system of linear equations: Where y=[y1,y2,...yn]T, 1=[1,1,...,1]T, E is the unit matrix, K(x i ,x) is the kernel function, , the solution of the equation system is Then the LS-SVM prediction model is: The classification output function of the LS-SVM model is obtained as: .

[0012] An LS-SVM distribution fault monitoring system based on edge node information includes the following four modules: Data acquisition module: Through sensors installed on the power distribution equipment, various operating data of the power distribution equipment are collected in real time, including current, voltage, temperature, and power factor; Data processing module: pre-data standardization processing is performed on the collected data to improve the accuracy and efficiency of data analysis; Fault diagnosis module: Analyze the processed data using the LS-SVM algorithm, monitor the faults in the power distribution system, and automatically send warning information to the system administrator, including fault time and fault location; User interface module: provides a friendly user interface to facilitate system administrators to view equipment operating status, receive warning information, and view fault location.

[0013] The beneficial effects of the present invention are: The present invention constructs a system model through the LS-SVM algorithm to monitor power distribution faults, solves the problems of long fault location time, low accident handling efficiency and inability to monitor in time in the distribution network system, improves the real-time nature of fault monitoring and the accuracy of fault location, and improves the efficiency of solving power distribution faults.

[0014] The LS-SVM distribution fault monitoring method based on edge node information of the present invention greatly reduces the impact of distribution network faults, improves the efficiency of distribution network fault resolution, effectively reduces the power outage duration caused by distribution network faults, and ensures the stable operation of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a block diagram for implementing the method of the present invention.

[0016] Figure 2 The figure is a flow chart of the method of the present invention.

[0017] Figure 3 This is a schematic diagram of the LS-SVM algorithm used in the method of the present invention.

[0018] Figure 4 This is a graph of the LS-SVM model test results of the method of the present invention. DETAILED DESCRIPTION

[0019] It should be noted that in the description of the present invention, the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", "clockwise", "counterclockwise", etc. indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction.

[0020] In the present invention, unless otherwise clearly specified and limited, the terms "set", "install", "connect", "connect" and the like should be understood in a broad sense. For example, "fixed" can be a fixed connection, a detachable connection, or an integral connection; the connection can be a mechanical connection or an electrical connection; the connection can be a direct connection or an indirect connection through an intermediate medium, and can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0021] A least squares support vector machine (LS-SVM) distribution fault monitoring method and system based on edge node information.

[0022] LS-SVM distribution fault monitoring method based on edge node information, such as Figure 1 As shown, the following steps are included: Step 1: Deploy edge nodes near the data source of the distribution system (the starting point, end point of the distribution line or the key node in the distribution system). The edge nodes are equipped with sensors that can collect parameter data such as voltage, current, power factor, temperature, etc. in real time, and upload the collected data to the cloud platform through wireless or wired transmission. The cloud platform will perform unified management and alarm push. In this way, the actual operating status of the distribution network can be monitored in real time, and the data can be uniformly transmitted to the cloud platform for processing, reducing monitoring costs and achieving real-time, fast and efficient user response.

[0023] Step 2: The real-time data collected by the edge node sensor can be efficiently transmitted to the cloud database through wired or wireless communication technology. The cloud database not only has powerful data processing capabilities, but also has high-performance storage functions. It uses advanced relational database technology to ensure that the collected data can be safely and reliably stored in the database, ensuring the integrity and traceability of the data.

[0024] Step 3: Filter out the distribution system fault data and the data of normal operation of the distribution system from the stored data. Since the collected parameter data such as voltage, current, power factor, temperature, etc. have different dimensions and their change rates are also different, in order to avoid affecting the accuracy and generalization ability of LS-SVM, the sample data is subjected to Z-score standardization: Among them, x* is the data after standardization, x is the data before standardization, and u is the mean of this type of samples. is the standard deviation of the sample.

[0025] After converting the parameters into a standardized form, the data is labeled, that is, the data is labeled according to its category, where normal operation data is marked as 0 and fault data is marked as 1.

[0026] Step 4: Divide the labeled sample data set into 10 subsets, and select one of the subsets as the test set, and the other subsets as the training set, train the LS-SVM model, calculate and save the LS-SVM model accuracy, repeat the above process to make each subset a training set, and after the training, select the model with the highest accuracy as the optimal fault diagnosis model. The specific operations are as follows: Bring the data sample into the model to generate a training set: x i is the s-dimensional input vector, y i is the tag value of the data. Through the mapping function Process the data, project it to obtain a higher-order data set, and construct the decision function of the LS-SVM model: in, and b are the regression parameters of the LS-SVM model, represents the weight vector and b represents the bias vector.

[0027] In order to minimize the structural risk, the LS-SVM model adds a slack variable e to the constraints. i Then we get its function: in, is the penalty factor, e i is the error, and the Lagrange multiplier is substituted into the equation: Where a=[a1,a2,...,a n ] T is the Lagrange multiplier. The optimal b and a can be obtained by the KKT condition, that is, Eliminate variables and e i , the optimization problem is transformed into solving a system of linear equations: Where y=[y1,y2,...yn]T, 1=[1,1,...,1]T, E is the unit matrix, K(x i ,x) is the kernel function, , the solution of the equation system is Then the LS-SVM prediction model is: The classification output function of the LS-SVM model is obtained as: , where h(x) is the xth monitoring value obtained by monitoring; a=[a1,a2,...a n ] T is the Lagrange multiplier; K(x i ,x) is the kernel function; n is the number of samples in the training set; b is the bias vector; Step 5. In order to verify the effectiveness of the LS-SVM distribution fault monitoring method based on edge node information, the model accuracy is tested using actual data, and the model parameters are set as follows: the radial basis function is used as the kernel function of the LS-SVM; the value range of the penalty parameter C is set in [1,1000]; the value range of the kernel parameter is set to [0.1,1000]; the inertia weight w is set to 0.9; and the cross-validation fold is 10.

[0028] This paper selects 200 sets of information data from the actual edge nodes of a city's power distribution system as test data, including 164 sets of normal operation data and 36 sets of fault data. The test results are as follows: Figure 4 As shown in the figure, the fault diagnosis accuracy of the obtained model is as high as 98.5%, indicating that the model can accurately identify the faults of the monitored distribution system.

[0029] Step 6: Deploy the optimal LS-SVM distribution fault diagnosis model to the cloud, receive the edge node information data transmitted to the cloud, and then realize real-time monitoring of faults, which mainly includes the following parts: Model deployment: In the LS-SVM-based power distribution fault monitoring method described in the present invention, the trained and verified optimal LS-SVM power distribution fault diagnosis model needs to be deployed to a cloud server. The cloud server needs to have strong computing and data storage capabilities to ensure the efficient operation of the LS-SVM model and the stability of real-time data processing.

[0030] Data reception: The cloud server communicates with each edge node wirelessly and receives the transmitted information data in real time. These edge nodes include sensors and data acquisition devices installed at key locations in the power distribution system.

[0031] Data processing: After the cloud server receives the data from the edge node, it should have a data standardization step and can input the standardized data into the deployed LS-SVM model.

[0032] Fault monitoring: The LS-SVM model identifies faults based on the input data. If the model determines that a fault exists, it will immediately trigger the alarm system and send the fault information to the terminal devices of relevant personnel so that timely measures can be taken to troubleshoot and handle the fault and reduce the impact of the fault on the operation of the power grid.

[0033] Model optimization and update: As the distribution system operates and data continues to accumulate, the LS-SVM model needs to be optimized and updated. This can be achieved by regularly collecting new data or adjusting model parameters. The optimized and updated model will be deployed again on the cloud server to provide better fault monitoring results.

[0034] The distribution fault monitoring method based on LS-SVM can realize real-time monitoring and early warning of the operating status of the distribution system, and improve the safety and stability of the power grid operation. At the same time, through the centralized management and data processing capabilities of the cloud server, the edge node data can be quickly processed and efficiently utilized.

[0035] An LS-SVM distribution fault monitoring system based on edge node information includes the following four modules: Data acquisition module: Through sensors installed on the power distribution equipment, various operating data of the power distribution equipment, such as current, voltage, temperature, power factor, etc., are collected in real time.

[0036] Data processing module: perform pre-data standardization processing on the collected data to improve the accuracy and efficiency of data analysis.

[0037] Fault diagnosis module: Use the LS-SVM algorithm to analyze the processed data, monitor the faults in the distribution system, and automatically send warning information to the system administrator, including fault time, fault location, etc.

[0038] User interface module: provides a friendly user interface to facilitate system administrators to view equipment operating status, receive warning information, check fault location, etc.

[0039] The above is only a specific implementation of the present invention, but the protection scope of the present invention is not limited thereto. Any modification, equivalent replacement and improvement made by any person skilled in the art within the technical scope disclosed by the present invention and within the spirit and principle of the present invention shall be covered within the protection scope of the present invention. At the same time, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0040] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

Claims

1. A LS-SVM distribution fault monitoring method based on edge node information, characterized in that: The following steps are involved: S1 installs sensors at the edge nodes of the power distribution system to collect voltage, current, power factor, and temperature parameter data in real time, and uploads the collected data to the cloud; The data collected by the S2 sensor can be transmitted wirelessly to a database in the cloud, which has a storage function and stores the collected data in the database; S3 selects the power distribution system fault data and the power distribution system normal operation data from the stored data, converts the parameters therein into a standardized form, and then performs labeling processing on the data; S4: The labeled data is divided into a training set and a test set for training the LS-SVM model. The LS-SVM distribution fault monitoring model has the following classification output function: Among them, h(x) is the xth monitoring value obtained by monitoring; a=[a1,a2,...a n ] T is the Lagrange multiplier; K(x i ,x) is the kernel function; n is the number of samples in the training set; b is the bias vector; S5 deploys the LS-SVM power distribution fault diagnosis model to the cloud, receives edge node information data transmitted to the cloud, and then realizes real-time monitoring of faults.

2. The LS-SVM power distribution fault monitoring method based on edge node information according to claim 1 is characterized in that: In step S3, the conversion into a standardized form is specifically performed by the formula: The data is processed by Z-score standardization, where x* is the data after standardization, x is the data before standardization, and u is the mean of this type of samples. is the standard deviation of the sample.

3. The LS-SVM power distribution fault monitoring method based on edge node information according to claim 2 is characterized in that: In step S3, the labeling process is specifically to label the data according to the category to which it belongs, with normal operation data being labeled as 0 and fault data being labeled as 1.

4. The LS-SVM power distribution fault monitoring method based on edge node information according to claim 1 is characterized in that: In step S4, the model training is specifically to bring the data sample into the model to generate a training set: , x i is the input vector, y i is the tag value of the data, through the mapping function Process the data, project it to obtain a higher-order data set, and construct the decision function of the LS-SVM model: in, and b are the regression parameters of the LS-SVM model, represents the weight vector and b represents the bias vector.

5. The LS-SVM power distribution fault monitoring method based on edge node information according to claim 4 is characterized in that: The LS-SVM model adds a slack variable e to the constraints i Then we get its function: in, is the penalty factor, e i is the error, and the Lagrange multiplier is substituted into the equation: Where a=[a1,a2,...,a n ] T is the Lagrange multiplier.

6. The LS-SVM power distribution fault monitoring method based on edge node information according to claim 5 is characterized in that: The optimal b and a can be obtained through KKT conditions, { Eliminate variables and e i , the optimization problem is transformed into solving a system of linear equations: Where y=[y1,y2,...yn]T, 1=[1,1,...,1]T, E is the unit matrix, K(x i ,x) is the kernel function, , the solution of the equation system is Then the LS-SVM prediction model is: The classification output function of the LS-SVM model is obtained as: .

7. A LS-SVM power distribution fault monitoring system based on edge node information, characterized in that: It includes the following four modules: Data acquisition module: Through sensors installed on the power distribution equipment, various operating data of the power distribution equipment are collected in real time, including current, voltage, temperature, and power factor; Data processing module: pre-data standardization processing is performed on the collected data to improve the accuracy and efficiency of data analysis; Fault diagnosis module: monitors faults in the power distribution system and automatically sends warning information to the system administrator, including fault time and fault location; User interface module: provides a friendly user interface to facilitate system administrators to view equipment operating status, receive warning information, and view fault location.

Images