Abnormality monitoring method for substation equipment and related equipment

Abstract

The invention discloses an anomaly monitoring method for substation equipment and related equipment, relates to the technical field of anomaly monitoring of a power system, and solves the problem of large misjudgment rate difference during anomaly monitoring of different types of substation equipment in the prior art. According to the method, feature weight distribution is independently performed on various types of substation equipment, so that the spectral clustering algorithm can extract key features of each type of equipment and perform clustering based on the key features, and misjudgment caused by feature distribution difference is avoided; and in combination with unified computational logic, the cross-device type consistency of the abnormal monitoring result can be ensured. By adopting the embodiment provided by the invention, the abnormal conditions of different types of substation equipment can be accurately detected.

Description

Technical Field

[0001] The present invention relates to the technical field of abnormal monitoring of power systems, and particularly to an abnormal monitoring method for substation equipment and related devices. Background Art

[0002] Currently, due to the frequent operation of the power system, the abnormality of substation equipment has become one of the important causes of safety accidents. If the status and performance data of substation equipment are not detected in time, it may lead to equipment failures, electrical accidents, fire and other safety risks, affecting the power supply reliability and continuity. Therefore, it is necessary to monitor the abnormal conditions of substation equipment, timely detect the abnormal conditions of the equipment, so as to ensure the safety of personnel and equipment and ensure the stable operation of the power system.

[0003] In traditional technologies, the clustering method is often used to process the status data of substation equipment to determine whether the operation of the substation equipment is abnormal. However, the existing state data clustering uses a fixed similarity threshold, which cannot adapt to the characteristic distribution differences of different equipment types, resulting in a large difference in the misjudgment rate of the same clustering strategy for different substation equipment.

[0004] In view of this, an abnormal monitoring method for substation equipment and related devices is needed. Summary of the Invention

[0005] Aiming at the problem that the existing abnormal monitoring methods have a large difference in the misjudgment rate for different types of substation equipment, the present invention provides an abnormal monitoring method for substation equipment and related devices, which can improve the cross-device type detection consistency during abnormal detection. The specific technical solutions are as follows:

[0006] In a first aspect, an embodiment of the present application provides an abnormal monitoring method for substation equipment, including:

[0007] Obtain the operation data set of the substation equipment, where the operation data set includes operation variables and fault variables; calculate the mutual information between the fault variable and the target operation variable, and assign a feature weight to the target operation variable based on the mutual information; where the target operation variable is the operation variable of substation equipment of the same equipment type; construct a dynamic similarity matrix of the substation equipment based on the feature weight; perform clustering on the dynamic similarity matrix based on the spectral clustering algorithm to obtain a clustering result; calculate an abnormal score based on the clustering result; and output the abnormal level of the substation equipment based on the abnormal score.

[0008] Preferably, the expression for assigning the feature weight is:

[0009]

[0010] where type is the equipment type, j and k are the operation variable numbers, is the characteristic weight of the j-th operating variable of equipment type type; MI() represents mutual information, and f j and f k respectively represent the target operating variable j and the target operating variable k, y represents the fault variable, and d represents the feature dimension.

[0011] Preferably, the element expression representing the similarity of the same type of equipment in the dynamic similarity matrix is:

[0012]

[0013] where S ij represents the similarity between the substation equipment i and the substation equipment j of the same type, the exp function is the exponential function, and w k represents the characteristic weight of the operating variable k, and f ik represents the operating variable k of the substation equipment i, and f jk represents the operating variable k of the substation equipment j; the element representing the similarity of different types of equipment in the dynamic similarity matrix is 0.

[0014] Preferably, the spectral clustering algorithm is used to cluster the dynamic similarity matrix to obtain a clustering result, including: dividing the dynamic similarity matrix into multiple sub-matrices based on the equipment type, and each sub-matrix corresponds to one equipment type; respectively performing spectral clustering on each sub-matrix to obtain corresponding multiple clustering results, and each clustering result includes K clustering clusters; the calculation formula for the number of clustering clusters K is:

[0015]

[0016] where n is the number of equipment of the equipment type corresponding to the sub-matrix, Δ is the equipment type correction factor, and Δ is determined based on the equipment type.

[0017] Preferably, the clustering result includes the eigenvector of the substation equipment; the anomaly score includes the first anomaly score and the second anomaly score; calculating the anomaly score based on the clustering result includes: calculating the number of neighbor nodes of the substation equipment based on the clustering result, as well as the class center and the within-class standard deviation of the clustering cluster to which the substation equipment belongs; calculating the first anomaly score based on the number of neighbor nodes of the substation equipment and the number of within-class nodes in the clustering cluster to which the substation equipment belongs; calculating the second anomaly score based on the distance between the eigenvector and the class center, and the within-class standard deviation; outputting the anomaly level of the substation equipment based on the anomaly score, including: outputting the anomaly level based on the first anomaly score and the second anomaly score.

[0018] Preferably, the operating variables include voltage, current harmonic distortion rate, vibration spectrum energy, and temperature gradient.

[0019] Preferably, obtaining the operation dataset of the substation equipment includes: obtaining the operation data of the substation equipment within the window time of the sliding window with a preset length as the dataset.

[0020] In a second aspect, an embodiment of the present application provides an abnormal monitoring system for substation equipment, which is applied to the method as described in the first aspect, and includes:

[0021] An acquisition module, configured to acquire an operation dataset of substation equipment, where the operation dataset includes operation variables and fault variables;

[0022] A weight module, configured to calculate the mutual information between the fault variable and the target operation variable, and assign a feature weight to the target operation variable based on the mutual information; where the target operation variable is the operation variable of substation equipment of the same equipment type;

[0023] A construction module, configured to construct a dynamic similarity matrix of the substation equipment based on the feature weight;

[0024] A clustering module, configured to perform clustering on the dynamic similarity matrix based on the spectral clustering algorithm to obtain a clustering result;

[0025] A scoring module, configured to calculate an abnormal score based on the clustering result;

[0026] An output module, configured to output the abnormal level of the substation equipment based on the abnormal score.

[0027] In a third aspect, an embodiment of the present application provides a computing device, including: a memory, configured to store a program; a processor, configured to load the program to execute the method as described in the first aspect.

[0028] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, where the computer-readable storage medium includes a stored program, and when the program runs, it controls the device where the computer-readable storage medium is located to execute the method as described in the first aspect.

[0029] Compared with the prior art, the beneficial effects of the present invention are as follows: By independently assigning feature weights to various substation devices, the spectral clustering algorithm can extract the key features of each type of device and perform clustering based on these key features, avoiding misjudgment caused by differences in feature distributions; combined with a unified calculation logic, the consistency of the anomaly monitoring results across device types can be ensured; based on the feature weights, a dynamic similarity matrix is constructed to reflect the similarity degree of device features, which can amplify the local outlier of the anomaly node; then, through the spectral clustering method, the similarity information is converted into a matrix form and the high-dimensional data is mapped to a low-dimensional space, and the more obvious clustering structure makes the clustering results representing normal nodes and anomaly nodes more accurate; then, based on the clustering results, an anomaly score is calculated, and the anomaly level can be accurately obtained. By using the method provided by the present invention, the anomaly conditions of different types of substation devices can be accurately detected. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. In all the drawings, similar elements or parts are generally denoted by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to actual scale.

[0031] Figure 1 It is a schematic flowchart of an anomaly monitoring method for a substation device provided by an embodiment of the present application;

[0032] Figure 2 It is a schematic structural diagram of an anomaly monitoring system for a substation device provided by an embodiment of the present application;

[0033] Figure 3 It is a schematic structural diagram of a computing device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0034] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0035] It should be understood that when used in this specification and the appended claims, the terms "comprises" and "comprising" indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their groups.

[0036] It should also be understood that the terms used in the specification of the present invention are merely for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include the plural forms.

[0037] It should be further understood that the term "and / or" used in the specification of the present invention and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0038] To solve the problem that the misjudgment rate varies greatly when the prior art monitors abnormalities of different types of substation equipment, the present invention provides an abnormal monitoring method for substation equipment and related equipment, which can improve the cross-device type detection consistency during abnormal detection.

[0039] Please refer to Figure 1 , Figure 1 which is a schematic flow chart of an abnormal monitoring method for substation equipment provided by an embodiment of the present application. This method is applied to a computing device; the method specifically includes the following steps:

[0040] Step 101, the computing device obtains the operation data set of the substation equipment.

[0041] Among them, the substation equipment includes primary equipment such as transformers, generators, circuit breakers and switch lights, as well as secondary equipment such as measuring instruments, relay protection devices, measurement and control devices and communication equipment; the operation data set includes operation variables and fault variables.

[0042] Among them, the computing device can be communicatively connected to the communication equipment in the substation system by wired or wireless means to obtain the operation data of the substation equipment in real time.

[0043] Preferably, the operation variables include voltage, current harmonic distortion rate, power factor, vibration spectrum energy, partial discharge and temperature gradient. It can be understood that in order to accurately detect abnormalities of substation equipment, the computing device can obtain as many operation variables related to substation equipment failures as possible for calculation. For example, transformer abnormalities are dominated by temperature, and circuit breaker abnormalities are dominated by vibration energy.

[0044] Among them, when collecting electrical parameters such as voltage, current harmonic distortion rate, and power factor in the operation variables, the sampling frequency can be 1 kHz; when collecting non-electrical parameters such as vibration spectrum energy, partial discharge and temperature gradient, the sampling frequency can be 100 Hz.

[0045] Optionally, the fault variable can be used to indicate various different fault types or whether a fault has occurred.

[0046] Among them, the operation data set can be a set of historical data of substation equipment or a set of operation data of substation equipment collected in real time.

[0047] After obtaining the above data, the computing device can preliminarily process these data, including data cleaning, outlier detection and compensation. Among them, data cleaning is mainly to remove obviously wrong data points, such as values beyond the physically possible range; outlier detection can use the standard deviation method, that is, data points exceeding the range of plus or minus 3 times the standard deviation of the mean are marked as outliers and removed; data compensation is to perform interpolation processing on missing or abnormal data points, such as linear interpolation and spline interpolation.

[0048] Exemplarily, for a continuously monitored process, the computing device can, when first implementing the method of this embodiment, obtain the historical data set of substation equipment for subsequent calculation and evaluation; after that, the computing device can obtain real-time operation data based on a preset time interval, and combine the real-time operation data and historical calculation data to perform iterative calculation and evaluation.

[0049] Preferably, the computing device can obtain the operation data within the window time of the sliding window with a preset length. Exemplarily, the window length can be 1 minute and the step size can be 10 seconds.

[0050] After completing the preprocessing of the data, the computing device can extract the features of each substation equipment based on the preprocessed operation data set. The features of each substation equipment include target operation variables, specifically including electrical variables (such as voltage, current harmonic distortion rate and power factor), non-electrical variables (such as vibration spectral energy, partial discharge and temperature gradient) and time variables (such as load change rate and operation duration); these target operation variables can form a feature vector corresponding to the substation equipment, and each target operation variable is the feature of the corresponding substation equipment.

[0051] Step 102: The computing device calculates the mutual information between the fault variable and the target operation variable, and assigns a feature weight to the target operation variable based on the mutual information.

[0052] Among them, the target operation variable is the operation variable of substation equipment of the same equipment type. That is to say, the computing device can independently assign feature weights to each type of substation equipment, and the feature weights of the same operation variable in different types of substation equipment may be different. For example, transformer anomalies are dominated by temperature, and circuit breaker anomalies are dominated by vibration energy. Therefore, the feature weight of the temperature gradient of the transformer can be 0.8, and the feature weight of the temperature gradient of the circuit breaker can be 0.2.

[0053] By independently assigning feature weights to various substation devices, the spectral clustering algorithm can extract the key features of each type of device and perform clustering based on these key features, avoiding misjudgments caused by differences in feature distributions.

[0054] It can be understood that the computing device calculates the mutual information between the target operating variables and the fault variables indicating the occurrence of faults, and does not calculate the mutual information related to the fault variables indicating "no fault" or "normal device state".

[0055] Among them, the mutual information method is a correlation analysis method that can analyze the correlation between different variables. Specifically, the computing device can first calculate the univariate entropy of each target operating variable, and the calculation formula is:

[0056] H(f) = -∑ a∈A p(a)logp(a)

[0057] Among them, f is the target operating variable, a is the data corresponding to the target operating variable, H(f) is the univariate entropy of f, and p(a) is the probability distribution of A. Then, the computing device can calculate the joint entropy of the target operating variable and the fault variable under different combinations, and the calculation formula is:

[0058] H(f|y) = -∑ a∈A，b∈B p(a|b)logp(a|b)

[0059] Among them, y is the fault variable, b is the data corresponding to the fault variable, and H(f, y) is the joint entropy of f and y.

[0060] Then, the computing device can calculate the mutual information between the fault variable and the target operating variable, and the calculation formula is:

[0061] I(f; y) = H(f) + H(y) - H(f|y)

[0062] Among them, I(f; y) is the mutual information between the fault variable y and the target operating variable f.

[0063] Then, the computing device can normalize the mutual information to better represent the correlation. The calculation formula is:

[0064]

[0065] Among them, MI(f; y) is the mutual information between the normalized fault variable y and the target operating variable f.

[0066] The computing device can determine the correlation between the target operating variable and the occurrence of faults based on the mutual information value indicating the correlation, and then assign feature weights to the operating variables. Preferably, the expression for assigning the feature weights is:

[0067]

[0068] Among them, type is the device type, and j and k are the running variable numbers. is the characteristic weight of the j-th running variable of the device type type; MI() represents mutual information, and f j and f k represent the target running variable j and the target running variable k respectively, y represents the fault variable, and d represents the characteristic dimension.

[0069] Step 103: The computing device constructs a dynamic similarity matrix of the substation devices based on the characteristic weights.

[0070] Among them, the computing device can construct a total dynamic similarity matrix based on the characteristic weights of all substation devices.

[0071] Preferably, the element expression representing the similarity of devices of the same type in the dynamic similarity matrix is:

[0072]

[0073] Among them, S ij represents the similarity between the substation device i and the substation device j of the same type, the exp function is the exponential function, and w k represents the characteristic weight of the running variable k, and f ik represents the running variable k of the substation device i, and f jk represents the running variable k of the substation device j; the elements representing the similarity of devices of different types in the dynamic similarity matrix are 0.

[0074] Specifically, calculating the weighted Euclidean distance as the similarity can pay more attention to the feature differences of high weights and amplify the local outlier of abnormal nodes.

[0075] In some other possible implementations, the computing device can also assign different weights to the elements representing the similarity of devices of the same type and the elements representing the similarity of devices of different types, so as to further distinguish the characteristics of different types of devices and avoid misjudgment.

[0076] Step 104: The computing device performs clustering on the dynamic similarity matrix based on the spectral clustering algorithm to obtain a clustering result.

[0077] Spectral clustering is a clustering algorithm based on graph theory. It realizes clustering by performing eigen-decomposition on the similarity matrix between data points and using the spectral properties of the matrix.

[0078] Among them, the computing device can partition the dynamic similarity matrix based on the device type and then perform clustering, or directly classify based on the overall dynamic similarity matrix; it can be understood that when using different clustering methods, a reasonable K value needs to be calculated to obtain K clustering clusters through clustering.

[0079] Preferably, the computing device can partition the dynamic similarity matrix into multiple sub-matrices based on the device type, and each sub-matrix corresponds to a device type; spectral clustering is performed on each sub-matrix respectively to obtain corresponding multiple clustering results, and each clustering result includes K clustering clusters; the calculation formula for the number K of clustering clusters is:

[0080]

[0081] where n is the number of devices of the device type corresponding to the sub-matrix, Δ is the device type correction factor, and Δ is determined based on the device type. The specific clustering process can be as follows:

[0082] The computing device can calculate the adjacency matrix and the degree matrix based on the sub-matrix. The elements in the adjacency matrix are the connection weights between two substation devices; then, based on the adjacency matrix and the degree matrix, the normalized Laplacian matrix is calculated, and the corresponding calculation formula is:

[0083] L = D -1 / 2 (D - W)D -1 / 2 ;

[0084] where L is the normalized Laplacian matrix, W is the adjacency matrix, and the elements in the adjacency matrix are the connection weights between two substation devices; D is the degree matrix, and the elements in D are the sum of the connection weights corresponding to a substation device. Then the computing device can extract the eigenvectors corresponding to the first K smallest eigenvalues in the normalized Laplacian matrix, and then construct the eigenmatrix based on the eigenvectors corresponding to the first K smallest eigenvalues. It can be understood that each row of the eigenmatrix corresponds to the eigenvector of a substation device.

[0085] Then, the computing device can perform K-means clustering on the eigenmatrix to obtain the clustering result. For the sub-matrices corresponding to multiple device types, the computing device obtains multiple clustering results, each clustering result includes K clustering clusters, and each clustering cluster includes the eigenvectors of one or more substation devices.

[0086] In spectral clustering, the eigenvector corresponding to the smallest eigenvalue of the normalized Laplacian matrix is a constant vector (all elements are equal), representing the global connectivity of the submatrix; other eigenvalues correspond to the local structural information of the submatrix, and their corresponding eigenvectors can be used to divide the submatrix into K community structures. Therefore, spectral clustering can be considered as embedding the nodes of the submatrix into a low-dimensional space composed of the first K eigenvectors; in this low-dimensional space, since similar nodes will be geometrically closer, k-means clustering can effectively divide the communities.

[0087] Step 105, the computing device calculates an anomaly score based on this clustering result.

[0088] Among them, the computing device can calculate the degree of deviation of each substation device based on this clustering result, so as to determine the anomaly level of the substation device in step 106. Specifically, the computing device can quantify this degree of deviation by calculating the anomaly score of the substation.

[0089] Preferably, the computing device calculates the number of neighbor nodes of this substation device based on this clustering result, as well as the class center and within-class standard deviation of the clustering cluster to which this substation device belongs; calculates this first anomaly score based on the number of neighbor nodes of this substation device and the number of nodes within the clustering cluster to which this substation device belongs; calculates this second anomaly score based on the distance between this eigenvector and the class center, as well as the within-class standard deviation.

[0090] Among them, the computing device can regard the eigenvector of the substation device as a node and calculate the number of neighbor nodes of this substation device through the DBSCAN algorithm.

[0091] Specifically, in the DBSCAN algorithm, for a given node p, calculating its number of neighbors is to count the number of other nodes within a range not exceeding Eps from p. Usually, distance metrics such as Euclidean distance are used to calculate the distance between nodes. When specifically implemented, all nodes in the dataset can be traversed, the distance between each node and p is calculated, and the number of nodes with a distance less than or equal to Eps is counted.

[0092] It can be understood that the computing device can also use other algorithms to calculate the number of neighbors, and no specific limitation is made here.

[0093] Among them, the computing device can calculate the mean of all nodes in the clustering cluster to obtain the class center, and calculate the standard deviation of all nodes as the within-class standard deviation.

[0094] The formula for calculating the first anomaly score can be as follows:

[0095]

[0096] Among them, S strucRepresents the first anomaly score, N neighbor Represents the number of neighboring samples, N cluster Represents the number of intra-class nodes.

[0097] The formula for calculating the second anomaly score can be as follows:

[0098]

[0099] Where, C conf Represents the second anomaly score, x represents the feature vector of the substation node, μ c Represents the class center, σ c Represents the intra-class standard deviation.

[0100] Step 106, calculate the anomaly level of the substation equipment based on this anomaly score and output it.

[0101] Among them, the calculation device can determine the anomaly level of the substation equipment based on the first anomaly score, or based on the second anomaly score, or based on both the first anomaly score and the second anomaly score.

[0102] Preferably, the calculation device can perform a weighted calculation based on the preset weight, based on the first anomaly score and the second anomaly score to obtain a comprehensive score, then determine the anomaly level based on the comprehensive score, and then output the anomaly level to the output device that can be perceived by maintenance personnel or management personnel.

[0103] Specifically, the first anomaly score can capture the outlier property of the node in the topological space, and the second anomaly score can quantify the degree of deviation of the node features from the class center; through the first anomaly score representing structural anomalies and the second anomaly score representing confidence, maintenance personnel can quickly locate the root cause of the anomaly.

[0104] Exemplarily, the weighted value of the first anomaly score is 0.6, and the weighted value of the second anomaly score is 0.4.

[0105] Preferably, the calculation device can determine the actions to be performed based on the comprehensive score and a preset score threshold.

[0106] Exemplarily, when the comprehensive score is greater than or equal to 0.9, the calculation device can determine that the corresponding substation equipment is abnormal and needs to be immediately stopped for inspection; when the comprehensive score is less than 0.9 and greater than 0.7, the calculation device can determine that the corresponding substation equipment is suspected of being abnormal and needs to strengthen detection and plan maintenance; when the comprehensive score is less than or equal to 0.7, the calculation device can determine that the corresponding substation equipment is normal and only requires routine inspection.

[0107] In the embodiments of the present application, by independently assigning feature weights to various types of substation equipment, the spectral clustering algorithm can extract the key features of each type of equipment and perform clustering based on these key features, avoiding misjudgment caused by differences in feature distributions. Then, combined with a unified calculation logic, the cross-device type consistency of the abnormal monitoring results can be ensured. Based on this feature weight, a dynamic similarity matrix is constructed to reflect the similarity degree of equipment features, which can amplify the local outlier of abnormal nodes. Then, through the spectral clustering method, the similarity information is converted into a matrix form and the high-dimensional data is mapped to a low-dimensional space, and the more obvious clustering structure makes the clustering results representing normal nodes and abnormal nodes more accurate. Then, based on the clustering results, an abnormal score is calculated, and the abnormal level can be accurately obtained. By using the method provided by the present invention, the abnormal conditions of different types of substation equipment can be accurately detected.

[0108] The method provided in the embodiments of the present application has been described above. The system provided in the embodiments of the present application will be described below.

[0109] Please refer to Figure 2 , Figure 2 which is a schematic structural diagram of an abnormal monitoring system for substation equipment provided in the embodiments of the present application. As Figure 2 shown, the system 200 includes:

[0110] An acquisition module 201, configured to acquire an operation data set of substation equipment, where the operation data set includes operation variables and fault variables;

[0111] A weight module 202, configured to calculate the mutual information between the fault variable and the target operation variable, and assign a feature weight to the target operation variable based on the mutual information; where the target operation variable is the operation variable of substation equipment of the same equipment type;

[0112] A construction module 203, configured to construct a dynamic similarity matrix of the substation equipment based on the feature weight;

[0113] A clustering module 204, configured to perform clustering on the dynamic similarity matrix based on the spectral clustering algorithm to obtain a clustering result;

[0114] A scoring module 205, configured to calculate an abnormal score based on the clustering result;

[0115] An output module 206, configured to output the abnormal level of the substation equipment based on the abnormal score.

[0116] Preferably, the expression for assigning the feature weight is:

[0117]

[0118] where type is the equipment type, j and k are the operation variable numbers, is the characteristic weight of the j-th running variable of device type type; MI() represents mutual information, f j and f k respectively represent the target running variable j and the target running variable k, y represents the fault variable, and d represents the feature dimension.

[0119] Preferably, the element expression representing the similarity of devices of the same type in the dynamic similarity matrix is:

[0120]

[0121] where S ij represents the similarity between the substation device i and the substation device j of the same type, the exp function is the exponential function, w k represents the characteristic weight of the running variable k, f ik represents the running variable k of the substation device i, f jk represents the running variable k of the substation device j; the elements representing the similarity of devices of different types in the dynamic similarity matrix are 0.

[0122] Preferably, the clustering module 204 is specifically configured to divide the dynamic similarity matrix into multiple sub-matrices based on the device type, and each such sub-matrix corresponds to a device type; spectral clustering is respectively performed on each such sub-matrix to obtain corresponding multiple clustering results, and each clustering result includes K clustering clusters; the calculation formula for the number K of clustering clusters is:

[0123]

[0124] where n is the number of devices of the device type corresponding to the sub-matrix, Δ is the device type correction factor, and Δ is determined based on the device type.

[0125] Preferably, the clustering result includes the feature vector of the substation device; the anomaly score includes a first anomaly score and a second anomaly score; the scoring module 205 is specifically configured to calculate the number of neighbor nodes of the substation device, and the class center and within-class standard deviation of the clustering cluster to which the substation device belongs based on the clustering result; calculate the first anomaly score based on the number of neighbor nodes of the substation device and the number of within-class nodes in the clustering cluster to which the substation device belongs; calculate the second anomaly score based on the distance between the feature vector and the class center and the within-class standard deviation; the output module 206 is specifically configured to output the anomaly level based on the first anomaly score and the second anomaly score.

[0126] Preferably, the running variables include voltage, current harmonic distortion rate, vibration spectrum energy, and temperature gradient.

[0127] Preferably, the obtaining module 201 is specifically configured to obtain the operation data of the substation equipment within the window time of a sliding window with a preset length as the operation data set.

[0128] The abnormal monitoring system for substation equipment provided by the embodiments of the present application can be understood by referring to the corresponding content in the foregoing method embodiment section, and will not be repeated here.

[0129] As Figure 3 shown, Figure 3 FIG. is a possible schematic logical structure diagram of a computing device provided by an embodiment of the present application. The computing device 300 includes: a processor 301, a communication interface 302, a memory 303, and a bus 304. The processor 301, the communication interface 302, and the memory 303 are interconnected through the bus 304. In the embodiment of the present application, the processor 301 is used to control and manage the actions of the computing device 300. For example, the processor 3 = 01 is used to execute Figure 1 the steps in the embodiment and / or other processes for the technologies described herein. The communication interface 302 is used to support the computing device 300 to communicate. The memory 303 is used to store the program code and data of the computing device 300.

[0130] Among them, the processor 301 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logic blocks, modules, and circuits described in combination with the disclosure of the present application. The processor may also be a combination of computing functions, such as a combination of one or more microprocessors, a combination of a digital signal processor and a microprocessor, and so on. The bus 304 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 only a thick line is shown in FIG., but it does not mean that there is only one bus or one type of bus.

[0131] In another embodiment of the present application, a computer-readable storage medium is further provided. The computer-readable storage medium includes instructions, and when the instructions are run on a computer, the computer is caused to execute the above Figure 1 method described in the embodiment.

[0132] Those of ordinary skill in the art can realize that the units of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the components of each example have been generally described according to their functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0133] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0134] In several embodiments provided by the embodiments of the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0135] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0136] In addition, the functional units in various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0137] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs that can store program codes.

[0138] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the various embodiments of the present invention, and they should all be covered within the scope of the claims and the description of the present invention.

Claims

1. An abnormal monitoring method for substation equipment, characterized in that, Including: Obtain the operation data set of substation equipment, where the operation data set includes operation variables and fault variables; Calculate the mutual information between the fault variables and the target operation variables, and assign feature weights to the target operation variables based on the mutual information; Where the target operation variables are the operation variables of substation equipment of the same equipment type; Construct the dynamic similarity matrix of the substation equipment based on the feature weights; Cluster the dynamic similarity matrix based on the spectral clustering algorithm to obtain a clustering result; Calculate the anomaly score based on the clustering result; Output the anomaly level of the substation equipment based on the anomaly score.

2. The method according to claim 1, wherein The expression for assigning the feature weights is: Among them, type is the device type, and j and k are the running variable numbers. is the characteristic weight of the j-th running variable of the device type type; MI() represents mutual information, f j and f k respectively represent the target running variable j and the target running variable k, y represents the fault variable, and d represents the feature dimension.

3. The method according to claim 1 or 2, characterized in that The element expression in the dynamic similarity matrix representing the similarity of the same type of equipment is: Among them, S ij represents the similarity between substation equipment i and substation equipment j of the same type. The exp function is the exponential function, and w k represents the characteristic weight of the operating variable k, and f ik represents the operating variable k of substation equipment i, and f jk represents the operating variable k of substation equipment j; The elements in the dynamic similarity matrix representing the similarity of different types of equipment are 0.

4. The method according to claim 1 or 2, characterized in that The clustering the dynamic similarity matrix based on the spectral clustering algorithm to obtain a clustering result includes: Divide the dynamic similarity matrix into multiple sub-matrices based on the equipment type, and each sub-matrix corresponds to one equipment type; Perform spectral clustering on each sub-matrix respectively to obtain corresponding multiple clustering results, and each clustering result includes K clustering clusters; the calculation formula for the number of clustering clusters K is: Where n is the number of equipment of the equipment type corresponding to the sub-matrix, and Δ is the equipment type correction factor, and Δ is determined based on the equipment type.

5. The method according to claim 1 or 2, characterized in that The clustering result includes the feature vectors of the substation equipment; the anomaly score includes the first anomaly score and the second anomaly score; calculating the anomaly score based on the clustering result includes: Calculate the number of neighbor nodes of the substation equipment, as well as the class center and within-class standard deviation of the clustering cluster to which the substation equipment belongs based on the clustering result; Calculate the first anomaly score based on the number of neighbor nodes of the substation equipment and the number of within-class nodes in the clustering cluster to which the substation equipment belongs; Calculate the second anomaly score based on the distance between the feature vector and the class center, and the within-class standard deviation; Outputting the anomaly level of the substation equipment based on the anomaly score includes: Output the anomaly level based on the first anomaly score and the second anomaly score.

6. The method according to claim 1 or 2, characterized in that, The operation variables include voltage, current harmonic distortion rate, vibration spectrum energy, and temperature gradient.

7. The method according to claim 1 or 2, characterized in that, The obtaining the operation data set of substation equipment includes: Obtain the operation data of the substation equipment within the window time corresponding to the sliding window through a sliding window with a preset length as the operation data set.

8. An abnormal monitoring system for substation equipment, characterized in that, Applied to the method according to any one of claims 1 to 7, including: An obtaining module, configured to obtain the operation data set of substation equipment, where the operation data set includes operation variables and fault variables; A weight module, configured to calculate the mutual information between the fault variables and the target operation variables, and assign feature weights to the target operation variables based on the mutual information; where the target operation variables are the operation variables of substation equipment of the same equipment type; A construction module, configured to construct the dynamic similarity matrix of the substation equipment based on the feature weights; A clustering module, configured to cluster the dynamic similarity matrix based on a spectral clustering algorithm to obtain a clustering result; A scoring module, configured to calculate an anomaly score based on the clustering result; An output module, configured to output the anomaly level of the substation equipment based on the anomaly score.

9. A computing device, characterized in that, Comprising: A memory, configured to store programs; A processor, configured to load the programs to execute the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes instructions that, when run on a computer, cause the computer to execute the method according to any one of claims 1-7.

Images