Intelligent substation merging unit fault diagnosis and self-healing method based on tensor decomposition

Abstract

The invention discloses an intelligent substation merging unit fault diagnosis and self-healing method based on tensor decomposition, and the method comprises the steps: collecting all operation data of a merging unit from a sensor network of an intelligent substation in real time, carrying out the processing, organizing the processed data into a tensor form, and outputting multi-dimensional data; decomposing the multi-dimensional data by using a tensor decomposition method, extracting fault features, and detecting the fault features to obtain a detection result; and determining the position and the type of the fault based on the detection result, starting a self-healing control strategy, and automatically repairing the fault by taking corresponding measures. Through the adaptive control technology, automatic repair of faults is realized, the self-healing capability of the intelligent substation is enhanced, and the reliability and stability of the system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis, and in particular to a fault diagnosis and self-healing method for a smart substation merging unit based on tensor decomposition. Background Art

[0002] The merging unit (MU) in a smart substation is responsible for collecting electrical quantity data from the transformer and converting it into digital signals for transmission to protection devices, measuring instruments and other smart devices. The performance and reliability of the MU are crucial to ensure the stable operation of the smart substation system. However, in actual applications, the MU may face a variety of faults such as data collection errors, time synchronization misalignment, and data transmission interruption. If these problems are not identified and solved in a timely and effective manner, they are likely to cause the protection device to malfunction or refuse to operate, thereby threatening the safety and stability of the entire smart substation system.

[0003] In actual operation, MU fault diagnosis faces many challenges. For example, during a routine maintenance inspection, the operation and maintenance team found that the current data of a specific line showed abnormal fluctuations, but it was difficult to determine whether this was caused by a problem with the MU's data collection or a fault in the line itself. In addition, the smart substation system has also recorded a significant increase in packet loss, but it is not certain whether this is due to a data transmission failure of the MU or a result of network congestion. Furthermore, the MU has been subjected to external interference, resulting in deviations in time synchronization, which further led to inaccurate data timestamps. These not only highlight the specific problems that MU may encounter in actual operation, but also reflect the limitations of existing MU fault diagnosis - that is, it mainly relies on preset rules and experience, and it is difficult to adapt to the complex and changing power grid environment, especially when facing new and unknown fault types, its accuracy and efficiency need to be improved.

[0004] The prior art discloses a state evaluation calculation method applicable to the combined unit of the intelligent substation. The method can evaluate the state of the combined unit and the system and calculate the evaluation index according to the requirements of the state maintenance work of the relay protection system of the intelligent substation. The method adopts a method combining expert analysis and actual engineering needs to determine the state quantities such as mercury delay drift, large discreteness of the time interval of the sample value message sending, sample value sequence error and frame loss, zero drift offset, sample value failure, ambient temperature and humidity, insulation condition and dust prevention, infrared temperature measurement, sampling accuracy, insulation condition and dust prevention, ambient temperature and humidity, infrared temperature measurement, etc. for monitoring and evaluation, and establishes corresponding calculation methods and indicators, which provides important guidance for the state maintenance work of the relay protection system of the intelligent substation. However, there is no disclosure of how to solve the technical problem of the limitations of the existing fault diagnosis method in processing multi-dimensional and nonlinear data, nor is there any disclosure of using tensor decomposition technology to solve the technical problem.

[0005] A method for monitoring the state of a merging unit of an intelligent substation, which monitors the operation state and synchronization state of the merging unit through a fault recording device, comprises the following steps: step S1, the fault recording device receives the amount of sampled value messages sent by the merging unit in each unit time and the signal of the merging unit sending the sampled value message; and the sampled value message amount is formed into a first recording data, and the signal of the merging unit sending the sampled value message is formed into a second recording data; step S2, the first recording data and the second recording data are respectively formed into a first waveform diagram and a second waveform diagram after analysis and processing, and the first waveform diagram is used to judge whether the operation of the merging unit is normal, and the second waveform diagram is used to judge whether the signal of the merging unit sending the sampled value message is synchronous, so as to monitor the operation state and synchronization state of the merging unit in advance, and the status of the merging unit can be understood in time, avoiding the blind spots of operation, maintenance, debugging and accident repair. However, it still does not disclose how to solve the technical problem of the limitations of the existing fault diagnosis method in processing multi-dimensional and nonlinear data, nor does it disclose the use of tensor decomposition technology to solve the technical problem.

[0006] The main problem with existing technologies is that they rely mainly on preset rules and experience, and are difficult to adapt to the complex and changing power grid environment. Especially when facing new and unknown fault types, their accuracy and efficiency need to be improved. Summary of the invention

[0007] In order to solve the above technical problems, the present invention proposes a smart substation merging unit fault diagnosis and self-healing method based on tensor decomposition, which can effectively extract the merging unit fault characteristics, improve the accuracy and efficiency of fault diagnosis, and realize automatic repair of faults through adaptive control technology, thereby improving the overall reliability and stability of the power grid.

[0008] To achieve the above object, the present invention provides a fault diagnosis and self-healing method for a smart substation merging unit based on tensor decomposition, comprising:

[0009] Collect various operation data of the merging unit from the sensor network of the smart substation in real time, process them, organize the processed data into tensor form, and output multi-dimensional data;

[0010] Decomposing the multidimensional data by using a tensor decomposition method, extracting fault features, and detecting the fault features to obtain detection results;

[0011] The location and type of the fault are determined based on the detection result, a self-healing control strategy is initiated, and corresponding measures are taken to achieve automatic repair of the fault.

[0012] Preferably, performing the processing includes standardizing the operating data; wherein the operating data includes current, voltage, data transmission rate, packet loss rate, time synchronization error and temperature.

[0013] Preferably, the method of organizing the processed data into a tensor form is:

[0014] The current, voltage, data transmission rate, packet loss rate, time synchronization error and temperature collected every second are organized into a three-dimensional tensor, specifically:

[0015] T=(t ijk )

[0016] Where T represents a tensor, t ijk Represents the elements in the tensor, i is the first dimension, j is the second dimension, and k is the third dimension;

[0017] Among them, the dimensions of the three-dimensional tensor are:

[0018] Time: indicates the time point of data collection;

[0019] Data type: including current, voltage, data transmission rate, packet loss rate, time synchronization error and temperature;

[0020] Collection point: indicates different data collection points.

[0021] Preferably, the method of decomposing the multidimensional data using the tensor decomposition method is:

[0022] T≈G× 1 A× 2 B× 3 C

[0023] In the formula, G is the core tensor, which represents the low-rank structure of the data; A is the factor matrix of the first dimension, which represents the characteristics of the time dimension; B is the factor matrix of the second dimension, which represents the characteristics of the data type; C is the factor matrix of the third dimension, which represents the characteristics of the collection point; 1 represents the product along the first dimension; × 2 represents the product along the second dimension; × 3 Represents the product along the third dimension.

[0024] Preferably, detecting the fault feature includes:

[0025] Determine whether there is a fault by calculating the residual tensor norm between the original tensor and the reconstructed tensor, and if the residual tensor norm exceeds a preset threshold, determine that there is a fault;

[0026] The reconstructed tensor is obtained by recombining the core tensor G and factor matrices A, B, and C of Tucker decomposition;

[0027] The method for calculating the residual tensor norm is:

[0028]

[0029] In the formula, ||Residual|| F is the residual tensor norm, ijk is the element of the residual tensor Residual at time point i, data type j, and collection point k;

[0030] The calculation method of the residual tensor Residual is:

[0031]

[0032] In the formula, t ijk is the element of the original tensor T at time point i, data type j, and collection point k, is the reconstructed tensor obtained by Tucker decomposition Elements at the same position.

[0033] Preferably, determining the location of the fault based on the detection result includes:

[0034] Calculate the residual tensor norm of each time point and acquisition point combination, and determine the initial fault location by finding the time point and acquisition point combination with the largest norm;

[0035] The contribution of the time dimension and the acquisition point dimension is introduced, the comprehensive residual norm is weighted, and the weighted final fault location is determined.

[0036] Preferably, determining the type of fault based on the detection result includes:

[0037] Determine the data type dimension, calculate the residual tensor norm of each data type dimension, and determine the data type with the largest norm as the initial fault type;

[0038] By introducing the contribution of the data type dimension, the data type norm is weighted and the final fault type is output.

[0039] Preferably, the self-healing control strategy is:

[0040] Increase data redundancy and adjust time synchronization mechanism;

[0041] If the fault is still not resolved after adjusting the time synchronization mechanism, switch to the backup merging unit to isolate the faulty component and try to repair the faulty component and restore the component function at the same time.

[0042] Preferably, the automatic repair of the fault is achieved by taking corresponding measures, including:

[0043] Increase the bandwidth of data transmission, optimize the data transmission protocol, add a data retransmission mechanism, optimize the network topology, use a more accurate time synchronization protocol, check the power supply line, add current protection devices, check the power supply equipment, add a voltage stabilizer, check the cooling system and increase environmental monitoring.

[0044] Compared with the prior art, the present invention has the following advantages and technical effects:

[0045] (1) The present invention can improve the accuracy of fault detection. By organizing multidimensional data into tensor form and using Tucker decomposition technology, it can comprehensively capture the characteristics of data in multiple dimensions such as time, type and acquisition point, thereby significantly improving the accuracy of fault detection. In addition, by calculating the residual tensor between the original tensor and the reconstructed tensor, the specific location of the fault can be accurately located. By analyzing the values ​​of the residual tensor on different data types, the fault type can be accurately identified, including data transmission fault, packet loss rate fault, time synchronization fault, current fault, voltage fault and temperature fault, providing a scientific basis for subsequent fault processing.

[0046] (2) The present invention improves the timeliness of fault detection. Through real-time data collection and processing mechanisms, the system can quickly detect abnormalities when a fault occurs and issue an alarm in a timely manner to avoid further expansion of the fault and ensure stable operation of the system.

[0047] (3) The present invention uses an efficient self-healing control strategy to automatically adjust data transmission parameters, time synchronization mechanisms, etc. according to the fault type and location, thereby reducing data loss and time synchronization errors and ensuring data accuracy and integrity. When a merging unit failure is detected, the system can automatically switch to a backup merging unit to ensure the continuity of data collection and transmission and avoid system downtime due to single point failures. Through the fault isolation and repair mechanism, the system can quickly isolate the faulty component and attempt to repair it, reducing maintenance time and cost and improving system reliability and availability. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0049] Figure 1 The present invention is a flowchart of a method for fault diagnosis and self-healing of a smart substation merging unit based on tensor decomposition according to an embodiment of the present invention. DETAILED DESCRIPTION

[0050] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0051] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0052] The present invention proposes a fault diagnosis and self-healing method for intelligent substation merging unit based on tensor decomposition, such as Figure 1 ,include:

[0053] Collect various operation data of the merging unit from the sensor network of the smart substation in real time, process them, organize the processed data into tensor form, and output multi-dimensional data;

[0054] Decomposing the multidimensional data by using a tensor decomposition method, extracting fault features, and detecting the fault features to obtain detection results;

[0055] The specific location and type of the fault are determined based on the detection result, a self-healing control strategy is initiated, and corresponding measures are taken to achieve automatic repair of the fault.

[0056] The present invention effectively solves the limitations of existing fault diagnosis methods in processing multidimensional and nonlinear data by introducing tensor decomposition technology, and improves the accuracy and efficiency of fault diagnosis. At the same time, through adaptive control technology, automatic repair of faults is achieved, the self-healing ability of smart substations is enhanced, and the reliability and stability of the system are improved.

[0057] Further, processing is performed, including standardization of the operating data; wherein the operating data includes current, voltage, data transmission rate, packet loss rate, time synchronization error and temperature.

[0058] Specifically, various operating data of the merging unit are collected in real time from the sensor network of the smart substation, including current, voltage, data transmission rate, packet loss rate, time synchronization error and temperature. These data are standardized to ensure data consistency and accuracy.

[0059] Current data normalization:

[0060]

[0061] Among them, u I and σ I are the mean and standard deviation of the current data, respectively.

[0062] Voltage data normalization:

[0063]

[0064] Among them, u V and σ V are the mean and standard deviation of the voltage data respectively.

[0065] Similarly, we can calculate: data transmission rate, packet loss rate, time synchronization error, and temperature data standardization.

[0066] Furthermore, the standardized data is organized into a tensor form. The current, voltage, data transmission rate, packet loss rate, time synchronization error, and temperature collected every second are organized into a three-dimensional tensor. The first dimension represents time, the second dimension represents the data type (current, voltage, data transmission rate, packet loss rate, time synchronization error, temperature), and the third dimension represents the collection point.

[0067] Specifically, the tensor definition:

[0068] T=(t ijk )

[0069] Among them, T represents the tensor, t ijk Represents the elements in the tensor, i is the first dimension representing time, j is the second dimension representing the data type (current, voltage, data transmission rate, packet loss rate, time synchronization error, temperature), and k is the third dimension representing the acquisition point. For example, t 123 Indicates the data of the second data type (such as voltage) collected in the first second at the third collection point.

[0070] Tensor structure:

[0071] The first dimension (time): represents the time point when data is collected, for example, data is collected once every second.

[0072] The second dimension (data type): includes current, voltage, data transmission rate, packet loss rate, time synchronization error and temperature.

[0073] The third dimension (collection point): represents different data collection points, for example, sensors at different locations.

[0074] Through the above steps, data collection, standardization and organization are completed, providing a high-quality data foundation for subsequent tensor decomposition and fault detection.

[0075] Furthermore, the method of decomposing the multidimensional data using the tensor decomposition method is:

[0076] T≈G× 1 A× 2B× 3 C

[0077] In the formula, G is the core tensor, which represents the low-rank structure of the data; A is the factor matrix of the first dimension, which represents the characteristics of the time dimension; B is the factor matrix of the second dimension, which represents the characteristics of the data type; C is the factor matrix of the third dimension, which represents the characteristics of the collection point; 1 represents the product along the first dimension; × 2 represents the product along the second dimension; × 3 Represents the product along the third dimension.

[0078] Specifically, tensor decomposition technology is used to decompose the collected multidimensional data, extract fault features, and detect whether there is a fault by calculating the residual tensor. Specifically, this step uses tensor decomposition technology (Tucker decomposition) to decompose the tensor data and extract fault features.

[0079] Through the above decomposition, a 3D tensor T can be decomposed into a core tensor G and three factor matrices A, B, and C.

[0080] Furthermore, the fault characteristics are detected, including:

[0081] Determine whether there is a fault by calculating the residual tensor norm between the original tensor and the reconstructed tensor, and if the residual tensor norm exceeds a preset threshold, determine that there is a fault;

[0082] Specifically, the reconstructed tensor is obtained by recombining the core tensor G and factor matrices A, B, and C of Tucker decomposition;

[0083] The reconstruction process is:

[0084]

[0085] Reconstructing Tensors It is obtained by recombining the core tensor G and factor matrices A, B, and C of Tucker decomposition.

[0086] Reconstructing Tensors It is obtained by low-rank approximation of the original tensor T, so some high-order information will be lost. This loss leads to the reconstructed tensor With the difference between the original tensor T, we can calculate the residual tensor below.

[0087] The calculation method of the residual tensor Residual is:

[0088]

[0089] Where, t ijkis the element of the original tensor T at time point i, data type j, and collection point k, is the reconstructed tensor obtained by Tucker decomposition Elements at the same position.

[0090] By adding each element t in the original tensor T ijk Subtract the reconstructed tensor The elements at the corresponding positions in Get the residual tensor Residual ijk The residual tensor reflects the differences between the original data and the low-rank approximation, and these differences may be caused by failures.

[0091] The method to calculate the norm of the residual tensor is:

[0092]

[0093] In the formula, ||Residual|| F is the residual tensor norm, ijk is the element of the residual tensor Residual at time point i, data type j, and collection point k;

[0094] The norm result of the residual tensor is used to determine whether there is a fault. If ||Residual|| F If it exceeds the preset threshold ∈, it is considered that there is a fault. In this embodiment, if ∈ == 0.1, then when ||Residual|| F When >=0.1, it is judged as a fault.

[0095] This step extracts fault features and detects potential faults through Tucker decomposition and residual calculation, providing a basis for subsequent fault location and classification.

[0096] Further, determining the specific location of the fault based on the detection result includes:

[0097] Calculate the residual tensor norm of each time point and acquisition point combination, and determine the initial fault location by finding the time point and acquisition point combination with the largest norm;

[0098] The contribution of the time dimension and the acquisition point dimension is introduced to weight the comprehensive residual norm and determine the weighted final fault location;

[0099] Specifically, calculate the contribution of each factor vector:

[0100] Contribution calculation:

[0101]

[0102] Among them, ar It is the rth column of the factor matrix A, representing the eigenvector of the time dimension, and i represents the index of the time dimension.

[0103]

[0104] Among them, b r It is the rth column of the factor matrix B, representing the eigenvector of the data type dimension, and j represents the index of the data type dimension.

[0105]

[0106] Among them, c r It is the rth column of the factor matrix C, representing the eigenvector of the acquisition point dimension, and k represents the index of the acquisition point dimension.

[0107] The role of contribution is that it can highlight more important factors in fault detection. If the contribution of a certain time point is very high, then even if the residual norm of this time point is not the largest, it may become the final fault location after weighting; similarly, if the contribution of a certain data type is very high, then even if its residual norm is not the largest, it may become the final fault type after weighting.

[0108] When determining the fault location, the contribution of the time dimension and the acquisition point dimension are introduced when calculating the comprehensive residual norm of each time point and acquisition point combination; when determining the fault type, the contribution of the data type dimension is also introduced to weight the data type norm.

[0109] Specifically, determine the fault location, including:

[0110] The final fault location is determined by calculating the residual tensor norm for each combination of time point and acquisition point.

[0111] ① Calculate the residual tensor norm of each combination of time point i and acquisition point k:

[0112]

[0113] Among them, J represents the size of the data type dimension, that is, the number of data types.

[0114] The above formula calculates the square root of the sum of the squares of the residual tensor norms of all data types j for each time point i and acquisition point k, and obtains the comprehensive residual norm. This norm value reflects the possibility of failure at a specific time point and acquisition point.

[0115] ② Determine the maximum norm to obtain the final fault location:

[0116]

[0117] The final fault location is determined by finding the time point and acquisition point combination with the largest norm. That is, the time dimension and acquisition point dimension are comprehensively considered to determine the fault location. The final fault location is determined by calculating the residual tensor norm of each time point and acquisition point combination and finding the combination with the largest norm. This multi-dimensional comprehensive analysis method ensures the comprehensiveness and accuracy of fault detection and provides more detailed and reliable guidance for subsequent self-healing control strategies.

[0118] ③Determine the final fault location after weighting:

[0119] By introducing the contribution of the time dimension and the acquisition point dimension and weighting the comprehensive residual norm, the importance of each time point and acquisition point in fault detection can be more accurately reflected.

[0120] Weighted Combined Residual Norm(i,k)

[0121] =Combined Residual Norm(i,k)×(Contribution(a r )

[0122] ×Contribution(c r ))

[0123] Among them, Weighted Combined Residual Norm(i,k) is the weighted residual tensor norm of each time point and acquisition point combination. Combined Residual Norm(i,k) is the residual tensor norm of each time point and acquisition point combination, reflecting the possibility of failure at a specific time point and acquisition point. r ) is the factor vector a of the time dimension r Contribution(c r ) is the factor vector c of the collection point dimension r The contribution of represents the importance of the acquisition point dimension in fault detection.

[0124] The above formula weights the comprehensive residual norm by multiplying the contribution of the time dimension and the acquisition point dimension, which can highlight those time points and acquisition points that are more important in fault detection. Time points and acquisition points with high contribution will occupy a larger weight in the weighted norm, thus playing a greater role in the final determination of the fault location and improving the accuracy of fault location determination. For example, if a time point has a high contribution, even if its residual norm is not the largest, it may become the final fault location after weighting.

[0125] Further, determining the type of the fault based on the detection result includes:

[0126] Determine the data type dimension, calculate the residual tensor norm of each data type dimension, and determine the data type with the largest norm as the initial fault type;

[0127] By introducing the contribution of the data type dimension, the data type norm is weighted and the final fault type is output.

[0128] Specifically, ① determine the data type dimension:

[0129]

[0130] Among them, I represents the size of the time dimension, that is, the number of time points, and K represents the size of the collection point dimension, that is, the number of collection points.

[0131] The above formula calculates the square root of the sum of the squares of the residual tensor norms of all time points i and acquisition points k for each data type j, and obtains the data type residual norm to determine the data type with the largest norm as the fault type.

[0132] ② Determine the maximum norm to obtain the final fault type:

[0133] Fault Type = argm j ax(Type Residual Norm(j))

[0134] By calculating the norm of the residual tensor of each data type, the data type with the largest norm is determined as the fault type.

[0135] In this embodiment, specific fault types include:

[0136] Data transmission failure: If the residual tensor has the highest value in the data transmission rate dimension, it is considered a data transmission failure.

[0137] Packet loss rate failure: If the residual tensor has the highest value in the packet loss rate dimension, it is considered a packet loss rate failure.

[0138] Time synchronization failure: If the residual tensor has the highest value in the time synchronization error dimension, it is considered a time synchronization failure.

[0139] Current fault: If the residual tensor has the highest value in the current dimension, it is considered a current fault.

[0140] Voltage fault: If the residual tensor has the highest value in the voltage dimension, it is considered a voltage fault.

[0141] Temperature failure: If the residual tensor has the highest value in the temperature dimension, it is considered a temperature failure.

[0142] ③Determine the final weighted fault type:

[0143] By introducing the contribution of the data type dimension and weighting the data type norm, the importance of each data type in fault detection can be more accurately reflected.

[0144] Weighted Combined Residual Norm(j)

[0145] =Combined Residual Norm(j)×Contribution(b r )

[0146] Among them, Weighted Combined Residual Norm(i,k) is the weighted residual tensor norm of each data type. Combined Residual Norm(j) is the residual tensor norm of each data type, reflecting the possibility of failure on a specific data type. Contribution(b r ) is a factor vector of data type dimension b r The contribution of represents the importance of the data type dimension in fault detection.

[0147] The above formula weights the data type norm by multiplying the contribution of the data type dimension, which can highlight those data types that are more important in fault detection. Data types with high contribution will occupy a larger weight in the weighted norm, thus playing a greater role in the final determination of the fault type and improving the accuracy of fault type determination. For example, if a data type has a high contribution, even if its residual norm is not the largest, it may become the final fault type after weighting.

[0148] Furthermore, the self-healing control strategy is:

[0149] Increase data redundancy and adjust time synchronization mechanism;

[0150] If the fault is not resolved after adjusting the time synchronization mechanism, switch to the backup merging unit to isolate the faulty component and try to repair the faulty component and restore its function at the same time.

[0151] Specifically, after confirming the fault location and type, the self-healing control strategy is started to automatically repair the fault by adjusting data transmission parameters, time synchronization mechanism, switching standby merging units, and fault isolation and repair.

[0152] The method to increase data redundancy is:

[0153] New Redundancy=Current Redundancy+δ

[0154] Where δ is the increased redundancy. Increasing data redundancy can improve data reliability and integrity and reduce the risk of data loss.

[0155] In this embodiment, if the current redundancy is 1.25 and δ==0.1, then the new redundancy is 1.25+0.1=1.35.

[0156] After increasing data redundancy, the system will proceed to the next step, which is to adjust the time synchronization mechanism to reduce the time synchronization error.

[0157] The methods to adjust the time synchronization mechanism are:

[0158] t new =t measured -α·Δt

[0159] Among them, α is the adjustment coefficient and Δt is the time deviation.

[0160] In this embodiment, if the measurement time t measured is 10.5 seconds, the reference time is 10.0 seconds, and the adjustment coefficient is 0.8, then Δt = 10.5-10 = 0.5 seconds, and the new time t new =10.5-0.8x0.5=10.1 seconds. Adjusting the clock can ensure the accuracy of time synchronization and reduce failures caused by time errors.

[0161] Furthermore, corresponding measures are taken to achieve automatic repair of the fault, including:

[0162] Increase the bandwidth of data transmission, optimize the data transmission protocol, add a data retransmission mechanism, optimize the network topology, use a more accurate time synchronization protocol, check the power supply line, add current protection devices, check the power supply equipment, add a voltage stabilizer, check the cooling system and increase environmental monitoring.

[0163] After adjusting the time synchronization mechanism, if the fault is still not resolved, the system will proceed to the next step, which is to switch to the backup merging unit to ensure the continuity of data collection and transmission.

[0164] Specifically, after increasing data redundancy and adjusting the time synchronization mechanism, if the fault still exists, the system will automatically switch to the backup merging unit. The specific method is as follows:

[0165] Fault detection: When the system detects a fault in a merging unit, the switching mechanism is immediately initiated.

[0166] Switching process: The system automatically switches the data collection and transmission tasks to the standby merging unit to ensure the continuity of data collection and transmission. This measure can avoid system downtime caused by single point failure.

[0167] After switching to the standby merging unit, the system will further isolate and repair the fault to completely resolve the fault. The specific method includes isolating the faulty component to ensure that it no longer affects the normal operation of the system, and trying to repair the faulty component to restore its function to ensure that the system can quickly return to normal status.

[0168] Additional measures for specific fault types:

[0169] After completing the above general self-healing measures, take additional targeted measures according to the fault type to further improve the reliability of the system. Specific measures include increasing the bandwidth of data transmission, optimizing the data transmission protocol, adding a data retransmission mechanism, optimizing the network topology, using a more accurate time synchronization protocol, checking the power supply line, adding current protection devices, checking the power supply equipment, adding a voltage stabilizer, checking the cooling system and adding environmental monitoring, as follows:

[0170] ① Data transmission failure: Increase the bandwidth of data transmission. Optimize the data transmission protocol and reduce transmission delay and packet loss rate.

[0171] ②Packet loss rate failure: Add data retransmission mechanism to ensure data integrity. Optimize network topology and reduce network congestion.

[0172] ③ Time synchronization failure: Use a more accurate time synchronization protocol, such as PTP (Precision = Time = Protocol). Regularly calibrate the clock to ensure the accuracy of time synchronization.

[0173] ④Current fault: Check the power supply circuit to eliminate short circuit or open circuit problems. Add current protection devices to prevent overload.

[0174] ⑤ Voltage failure: Check the power supply equipment to ensure the voltage is stable. Add a voltage stabilizer to prevent voltage fluctuations.

[0175] ⑥ Temperature failure: Check the cooling system to ensure that the equipment dissipates heat well. Increase environmental monitoring to detect temperature abnormalities in a timely manner.

[0176] The self-healing control strategy in this embodiment is applicable to all types of faults mentioned above. These measures include increasing data redundancy, adjusting the time synchronization mechanism, switching to a standby merging unit, and additional measures for specific fault types.

[0177] In this embodiment, if the system detects that the current data of the third collection point has abnormal fluctuations and the time synchronization error is large. Through Tucker decomposition and residual calculation, it is determined that there is a fault at the third collection point, and the fault type is a data transmission failure. The system first increases the redundancy of data transmission from 1.25 to 1.35 to reduce data loss. At the same time, adjust the clock and adjust the time deviation from 0.5 seconds to 0.1 seconds to ensure data synchronization. If these measures still fail to solve the problem, the system will switch to the backup MU. Finally, the system isolates the third collection point and attempts to repair it.

[0178] In the above technical solution, data collection and preprocessing provide a high-quality data foundation for subsequent tensor decomposition and fault detection. Tensor decomposition and fault detection extract fault features through Tucker decomposition, and determine whether there is a fault through residual tensor calculation, which provides a basis for fault location and classification. Fault location and classification determine the specific location and type of the fault by analyzing the residual tensor, providing clear guidance for the self-healing control strategy. The self-healing control strategy takes corresponding self-healing measures according to the fault location and type to ensure the stable operation of the system and the continuity of data.

[0179] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for fault diagnosis and self-healing of intelligent substation merging units based on tensor decomposition, characterized in that: include: Collect various operation data of the merging unit from the sensor network of the smart substation in real time, process them, organize the processed data into tensor form, and output multi-dimensional data; Decomposing the multidimensional data by using a tensor decomposition method, extracting fault features, and detecting the fault features to obtain detection results; The location and type of the fault are determined based on the detection result, a self-healing control strategy is initiated, and corresponding measures are taken to achieve automatic repair of the fault.

2. The method for fault diagnosis and self-healing of intelligent substation merging units based on tensor decomposition according to claim 1 is characterized in that: The processing includes standardizing the operating data; wherein the operating data includes current, voltage, data transmission rate, packet loss rate, time synchronization error and temperature.

3. The method for fault diagnosis and self-healing of intelligent substation merging unit based on tensor decomposition according to claim 2 is characterized in that: The method to organize the processed data into tensor form is: The current, voltage, data transmission rate, packet loss rate, time synchronization error and temperature collected every second are organized into a three-dimensional tensor, specifically: T=(t ijk ) Among them, T represents the tensor, t ijk Represents the elements in the tensor, i is the first dimension, j is the second dimension, and k is the third dimension; Among them, the dimensions of the three-dimensional tensor are: Time: indicates the time point of data collection; Data type: including current, voltage, data transmission rate, packet loss rate, time synchronization error and temperature; Collection point: indicates different data collection points.

4. The method for fault diagnosis and self-healing of intelligent substation merging unit based on tensor decomposition according to claim 1 is characterized in that: The method of decomposing the multidimensional data using the tensor decomposition method is: T≈G×1A×2B×3C In the formula, G is the core tensor, which represents the low-rank structure of the data; A is the factor matrix of the first dimension, which represents the characteristics of the time dimension; B is the factor matrix of the second dimension, which represents the characteristics of the data type; C is the factor matrix of the third dimension, which represents the characteristics of the collection point; ×1 represents the product along the first dimension; ×2 represents the product along the second dimension; and ×3 represents the product along the third dimension.

5. The method for fault diagnosis and self-healing of intelligent substation merging unit based on tensor decomposition according to claim 1 is characterized in that: Detecting the fault feature includes: Determine whether there is a fault by calculating the residual tensor norm between the original tensor and the reconstructed tensor, and if the residual tensor norm exceeds a preset threshold, determine that there is a fault; The reconstructed tensor is obtained by recombining the core tensor G and factor matrices A, B, and C of Tucker decomposition; The method for calculating the norm of the residual tensor is: In the formula, ||Residual|| F is the residual tensor norm, ijk is the element of the residual tensor Residual at time point i, data type j, and collection point k; The calculation method of the residual tensor Residual is: Where, t ijk is the element of the original tensor T at time point i, data type j, and collection point k, is the reconstructed tensor obtained by Tucker decomposition Elements at the same position.

6. The method for fault diagnosis and self-healing of intelligent substation merging unit based on tensor decomposition according to claim 1 is characterized in that: Determining the location of the fault based on the detection result includes: Calculate the residual tensor norm of each time point and acquisition point combination, and determine the initial fault location by finding the time point and acquisition point combination with the largest norm; The contribution of the time dimension and the acquisition point dimension is introduced, the comprehensive residual norm is weighted, and the weighted final fault location is determined.

7. The method for fault diagnosis and self-healing of intelligent substation merging unit based on tensor decomposition according to claim 1 is characterized in that: Determining the type of the fault based on the detection result includes: Determine the data type dimension, calculate the residual tensor norm of each data type dimension, and determine the data type with the largest norm as the initial fault type; By introducing the contribution of the data type dimension, the data type norm is weighted and the final fault type is output.

8. The method for fault diagnosis and self-healing of intelligent substation merging units based on tensor decomposition according to claim 1 is characterized in that: The self-healing control strategy is: Increase data redundancy and adjust time synchronization mechanism; If the fault is still not resolved after adjusting the time synchronization mechanism, switch to the backup merging unit to isolate the faulty component and try to repair the faulty component and restore the component function at the same time.

9. The method for fault diagnosis and self-healing of intelligent substation merging units based on tensor decomposition according to claim 8 is characterized in that: Automatically repair faults by taking appropriate measures, including: Increase the bandwidth of data transmission, optimize the data transmission protocol, add a data retransmission mechanism, optimize the network topology, use a more accurate time synchronization protocol, check the power supply line, add current protection devices, check the power supply equipment, add a voltage stabilizer, check the cooling system and increase environmental monitoring.

Images