Power transmission and transformation equipment abnormity early warning method and device

Abstract

The invention provides a power transmission and transformation equipment abnormity early warning method and device, and relates to the technical field of power transmission and transformation equipment monitoring. The method comprises the steps of firstly obtaining a plurality of running state data queues; performing sample similarity extraction on each running state data queue according to the plurality of standard state data queues, and constructing an extraction result into a sample feature vector; classifying each sample feature vector into the nearest standard sample feature class, and taking the state of the standard sample feature class as the state of the sample feature vector to obtain a plurality of states; and finally, according to the plurality of states, the probability of occurrence of the abnormity and the probability of occurrence of the plurality of states, determining the probability of the abnormity of the current power transmission and transformation equipment. According to the method, state division is carried out on multiple dimensions of the power transmission and transformation equipment according to the sample feature vectors, finally, the probability of anomaly is calculated by adopting the Bayesian formula according to the state division, and anomaly analysis and judgment are carried out based on dimension data, so that the accuracy is relatively high.

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission and transformation equipment monitoring, and in particular to a method and device for abnormal warning of power transmission and transformation equipment. Background Art

[0002] The electric energy produced by the power plant is supplied to the distribution system and users through the substation network constructed by various transmission and transformation equipment. The equipment of the general transmission system includes: equipment for voltage conversion, switching devices for connecting and disconnecting circuits, electrical appliances for overvoltage protection and fault current limitation, reactive power compensation equipment, current-carrying conductors, grounding devices, etc. Among them, the equipment for voltage conversion includes: transformers; switching devices for connecting and disconnecting circuits include: circuit breakers, disconnectors, fuses; electrical appliances for overvoltage protection and fault current limitation include: lightning arresters, lightning rods, lightning conductors, reactors; reactive power compensation equipment includes: power capacitors, synchronous phase regulators, static compensators; current-carrying conductors include: busbars, leads, cables, overhead lines; grounding devices include: transformer neutral point grounding, equipment casing grounding, lightning protection grounding, etc.

[0003] Most of the existing research is based on the fault diagnosis of partial equipment information of a single system. In the actual operation and maintenance of equipment, simple threshold judgment methods are mostly used to detect anomalies. The equipment information utilization rate and the accuracy of status evaluation are both low.

[0004] At present, the big data technologies that have been studied in the field of online monitoring data anomaly detection include time series analysis, Markov model, genetic programming algorithm, classification algorithm, etc., but most of these studies are still tentative and have not formed a reasonable and complete evaluation model. There is a lack of correlation analysis of the scattered fault points in the power grid.

[0005] Based on this, it is necessary to develop an abnormal early warning method for power transmission and transformation equipment. Summary of the invention

[0006] The embodiments of the present invention provide a method and device for abnormal warning of power transmission and transformation equipment, so as to solve the problem of low accuracy in abnormal state evaluation of power transmission and transformation equipment in the prior art.

[0007] In a first aspect, an embodiment of the present invention provides a method for abnormal warning of power transmission and transformation equipment, comprising:

[0008] Acquire multiple operation status data queues, wherein the operation status data queues represent the fluctuation of the power transmission and transformation voltage or current of the power transmission and transformation equipment over time;

[0009] Extracting sample similarity from each running state data queue according to the multiple standard state data queues, and constructing the extraction results into a sample feature vector, thereby obtaining multiple sample feature vectors;

[0010] Classify each sample feature vector into the nearest standard sample feature class, and use the state of the standard sample feature class as the state of the sample feature vector, thereby obtaining multiple states;

[0011] The probability of the current power transmission and transformation equipment being abnormal is determined based on the multiple states, the probability of the abnormality occurring, and the probability of the multiple states occurring.

[0012] In a possible implementation, performing sample similarity extraction on each running state data queue according to a plurality of standard state data queues and constructing the extraction result as a sample feature vector includes:

[0013] Get multiple standard status data queues;

[0014] For each of the multiple running status data queues, the following steps are performed respectively:

[0015] Starting from the first position of the running state data queue, continuously take out data with the same total amount of data as the standard state data queue to construct a queue to be processed;

[0016] Calculating the similarities between the queue to be processed and the plurality of standard state data queues respectively, and taking the obtained plurality of similarities as column vectors and adding them into a similarity matrix;

[0017] If the traversal is not completed, the first position is offset, and the process jumps to the step of continuously taking out data having the same total amount of data as the standard state data queue from the first position of the running state data queue to construct a queue to be processed;

[0018] Otherwise, the average value of the elements in each row of the similarity matrix is ​​extracted, and the extracted multiple average values ​​are constructed as a frequency feature vector.

[0019] In a possible implementation, respectively calculating the similarities between the queue to be processed and the plurality of standard status data queues includes:

[0020] The similarity between the queue to be processed and the plurality of standard state data queues is calculated according to a first formula, wherein the first formula is:

[0021]

[0022] Where SM is the similarity between the queue to be processed and the standard state data queue, Q(n) is the nth data in the queue to be processed, QS(n) is the nth data in the standard state data queue, and nmax is the total number of data in the standard state data queue.

[0023] In a possible implementation manner, acquiring the multiple standard status data queues includes:

[0024] Get multiple sample status data queues and cluster numbers;

[0025] Clustering the plurality of sample state data queues according to the clustering quantity, thereby obtaining a plurality of sample classes;

[0026] Extracting the class centers of the multiple sample classes to obtain multiple sample class centers;

[0027] Unitizing the multiple sample class centers to obtain multiple unit class centers;

[0028] Performing orthogonalization tests on the multiple unit class centers in pairs to obtain multiple test results;

[0029] If the average value of the plurality of test results is less than the orthogonalization threshold, the plurality of unit class centers are used as a plurality of standard state data queues;

[0030] Otherwise, the clustering quantity is adjusted and the process jumps to the step of clustering the plurality of sample status data queues according to the clustering quantity to obtain a plurality of sample classes.

[0031] In a possible implementation, performing orthogonalization tests on the multiple unit class centers in pairs to obtain multiple test results includes:

[0032] According to the second formula, orthogonalization tests are performed on the multiple unit class centers in pairs to obtain multiple test results, wherein the second formula is:

[0033]

[0034] Where QN i (n) is the nth data of the center of the i-th unit class, QN j (n) is the nth data of the jth unit class center, Orth(i,j) is the orthogonalization test result of the ith unit class center and the jth unit class center, and nmax is the total number of data in the unit class center.

[0035] In a possible implementation, classifying each sample feature vector into the nearest standard sample feature class and using the state of the standard sample feature class as the state of the sample feature vector includes:

[0036] For each sample feature vector, perform the following steps:

[0037] Obtain the class centers of multiple standard sample feature classes;

[0038] Calculating the distances between the sample feature vectors and the class centers of the plurality of standard sample feature classes respectively as a plurality of target distances;

[0039] Select the standard sample feature class corresponding to the smallest target distance as the target class;

[0040] The state of the target class is taken as the state of the sample feature vector.

[0041] In a possible implementation, after selecting the standard sample feature class corresponding to the target distance with the smallest value as the target class, the method includes:

[0042] Get the radius of the target class;

[0043] Taking the distance between the class center of the target class and the sample feature vector as the first distance;

[0044] If the ratio of the first distance to the radius of the target class is greater than a ratio threshold, a message indicating that the power transmission and transformation equipment is abnormal is output.

[0045] In a possible implementation, determining the probability of abnormality of the current power transmission and transformation equipment according to the multiple states, the probability of abnormality occurrence, and the probability of multiple states occurring includes:

[0046] According to the third formula, the multiple states, the probability of abnormal occurrence and the probability of multiple states occurring, the probability of abnormality of the current power transmission and transformation equipment is determined, wherein the third formula is:

[0047]

[0048] In the formula, P(E│S 1 S 2 …S k ) is the probability of abnormality of the current power transmission and transformation equipment, P(S k │E) When an abnormality occurs, the power transmission and transformation equipment is in S k The probability of the state, P(E) is the probability of abnormal state of the power transmission and transformation equipment, P(S k ) is the power transmission and transformation equipment in S k The probability of the state.

[0049] In a second aspect, an embodiment of the present invention provides an abnormal warning device for power transmission and transformation equipment, comprising:

[0050] A data acquisition module is used to acquire a plurality of operation status data queues, wherein the operation status data queues represent the fluctuation of the power transmission and transformation voltage or current of the power transmission and transformation equipment over time;

[0051] A feature extraction module is used to extract sample similarity from each running state data queue according to multiple standard state data queues, and construct the extraction results into a sample feature vector, thereby obtaining multiple sample feature vectors;

[0052] A state analysis module is used to classify each sample feature vector into the nearest standard sample feature class, and use the state of the standard sample feature class as the state of the sample feature vector, thereby obtaining multiple states;

[0053] as well as,

[0054] The abnormality analysis module is used to determine the probability of abnormality of the current power transmission and transformation equipment according to the multiple states, the probability of abnormality occurrence and the probability of multiple states occurring.

[0055] In a possible implementation, the abnormality analysis module includes:

[0056] The abnormality probability calculation module is used to determine the probability of abnormality of the current power transmission and transformation equipment according to a third formula, the multiple states, the probability of abnormality and the probability of multiple states, wherein the third formula is:

[0057]

[0058] In the formula, P(E│S 1 S 2 …S k ) is the probability of abnormality of the current power transmission and transformation equipment, P(S k │E) When an abnormality occurs, the power transmission and transformation equipment is in S k The probability of the state, P(E) is the probability of abnormal state of the power transmission and transformation equipment, P(S k ) is the power transmission and transformation equipment in S k The probability of the state.

[0059] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0060] The present invention provides a method and device for warning abnormality of power transmission and transformation equipment. The embodiment of the present invention first obtains multiple operation status data queues, wherein the operation status data queue characterizes the fluctuation of the power transmission and transformation voltage or current of the power transmission and transformation equipment over time; then, sample similarity extraction is performed on each operation status data queue according to multiple standard status data queues, and the extraction results are constructed as sample feature vectors, thereby obtaining multiple sample feature vectors; then, each sample feature vector is classified into the nearest standard sample feature class, and the state of the standard sample feature class is used as the state of the sample feature vector, thereby obtaining multiple states; finally, according to the multiple states, the probability of abnormal occurrence and the probability of multiple states occurring, the probability of abnormality of the current power transmission and transformation equipment is determined. The embodiment of the present invention extracts the sample feature vector of the operation status data queue by comparing the similarity with the standard status data queue, and then divides the state of the power transmission and transformation equipment into multiple dimensions according to the sample feature vector, and finally calculates the probability of abnormality according to the state division using the Bayesian formula. Since the abnormal analysis and judgment is performed based on the dimensional data, the accuracy is high, and compared with the artificial intelligence algorithm, the required sample volume is smaller, the modeling process is simpler, and the universality is higher. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0062] Figure 1 It is a flow chart of the implementation of the abnormal early warning method for power transmission and transformation equipment provided by the embodiment of the present invention;

[0063] Figure 2 It is a flowchart of a process of constructing a sample feature vector according to multiple standard state data queues provided by an embodiment of the present invention;

[0064] Figure 3 It is a structural schematic diagram of an abnormal warning device for power transmission and transformation equipment provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0065] In the following description, specific details such as specific system structures and techniques are provided for the purpose of illustration rather than limitation so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary details.

[0066] In order to make the purpose, technical solutions and advantages of the present invention more clear, a specific implementation method will be described below in conjunction with the accompanying drawings.

[0067] Figure 1 The implementation flow chart of the abnormal warning method for power transmission and transformation equipment provided by the embodiment of the present invention is detailed as follows:

[0068] In step 101, a plurality of operation status data queues are obtained, wherein the operation status data queues represent the fluctuation of the power transmission and transformation voltage or current of the power transmission and transformation equipment over time.

[0069] In step 102, sample similarity extraction is performed on each running state data queue according to a plurality of standard state data queues, and the extraction results are constructed as sample feature vectors, thereby obtaining a plurality of sample feature vectors.

[0070] In some implementations, performing sample similarity extraction on each running state data queue according to a plurality of standard state data queues and constructing the extraction result as a sample feature vector includes:

[0071] Get multiple standard status data queues;

[0072] For each of the multiple running status data queues, the following steps are performed respectively:

[0073] Starting from the first position of the running state data queue, continuously take out data with the same total amount of data as the standard state data queue to construct a queue to be processed;

[0074] Calculating the similarities between the queue to be processed and the plurality of standard state data queues respectively, and taking the obtained plurality of similarities as column vectors and adding them into a similarity matrix;

[0075] If the traversal is not completed, the first position is offset, and the process jumps to the step of continuously taking out data having the same total amount of data as the standard state data queue from the first position of the running state data queue to construct a queue to be processed;

[0076] Otherwise, the average value of the elements in each row of the similarity matrix is ​​extracted, and the extracted multiple average values ​​are constructed as a frequency feature vector.

[0077] In some implementations, respectively calculating the similarities between the queue to be processed and the plurality of standard state data queues includes:

[0078] The similarity between the queue to be processed and the plurality of standard state data queues is calculated according to a first formula, wherein the first formula is:

[0079]

[0080] Where SM is the similarity between the queue to be processed and the standard state data queue, Q(n) is the nth data in the queue to be processed, QS(n) is the nth data in the standard state data queue, and nmax is the total number of data in the standard state data queue.

[0081] In some implementations, acquiring the plurality of standard status data queues includes:

[0082] Get multiple sample status data queues and cluster numbers;

[0083] Clustering the plurality of sample state data queues according to the clustering quantity, thereby obtaining a plurality of sample classes;

[0084] Extracting the class centers of the multiple sample classes to obtain multiple sample class centers;

[0085] Unitizing the multiple sample class centers to obtain multiple unit class centers;

[0086] Performing orthogonalization tests on the multiple unit class centers in pairs to obtain multiple test results;

[0087] If the average value of the plurality of test results is less than the orthogonalization threshold, the plurality of unit class centers are used as a plurality of standard state data queues;

[0088] Otherwise, the clustering quantity is adjusted and the process jumps to the step of clustering the plurality of sample status data queues according to the clustering quantity to obtain a plurality of sample classes.

[0089] In some embodiments, performing pairwise orthogonalization tests on the multiple unit class centers to obtain multiple test results comprises:

[0090] According to the second formula, orthogonalization tests are performed on the multiple unit class centers in pairs to obtain multiple test results, wherein the second formula is:

[0091]

[0092] Where QN i (n) is the nth data of the center of the i-th unit class, QN j (n) is the nth data of the jth unit class center, Orth(i,j) is the orthogonalization test result of the ith unit class center and the jth unit class center, and nmax is the total number of data in the unit class center.

[0093] For example, power transmission and transformation equipment usually gathers together some of the power transmission and transformation equipment according to functional requirements to form a collection composed of multiple power transmission and transformation equipment. For example, a substation includes transformers, circuit breakers, disconnectors and various protection devices. From the big data of the input voltage, output voltage, input current and output current changes of the substation, the abnormal operation points of the substation can be mined.

[0094] To achieve the above purpose, the embodiment of the present invention classifies the states of the power transmission and transformation equipment according to the power transmission and transformation voltage and current, and determines the probability of abnormal occurrence based on the statistically obtained probability of abnormal occurrence, the probability of various states when the abnormality occurs, and the probability of state occurrence. When the probability of abnormality reaches a certain threshold, it indicates that the power transmission and transformation equipment should be inspected or repaired.

[0095] It takes a period of time to collect data on the power transmission and transformation voltage and current of the power transmission and transformation equipment. The collected data are arranged in chronological order to form an operation status data queue. For example, the input voltage signal is continuously collected to form a voltage status data queue, and the output current signal is continuously collected to form a current status data queue. The embodiments of the present invention divide the status of the power transmission and transformation equipment from different dimensions based on these data queues. For example, the input voltage status is divided according to the voltage status data queue, and the output current status is divided according to the current status data queue.

[0096] In order to realize the division of the power transmission and transformation equipment status according to the operation status data queue, the embodiment of the present invention extracts similar features from the operation status data queue according to the standard status data queue, and constructs a sample feature vector according to the similar features.

[0097] Figure 2 A flowchart of a process of constructing a sample feature vector according to a plurality of standard state data queues provided by an embodiment of the present invention is shown. In the figure, the running state data queue 201 takes out data of the same length as the standard state data queue each time as a queue to be processed 202. The queue to be processed 202 is respectively determined to have a similarity with a plurality of standard state data queues 203 using a first formula. The first formula is:

[0098]

[0099] Where SM is the similarity between the queue to be processed and the standard state data queue, Q(n) is the nth data in the queue to be processed, QS(n) is the nth data in the standard state data queue, and nmax is the total number of data in the standard state data queue.

[0100] The calculated similarity is constructed as a column vector according to the arrangement order of multiple standard state data queues and added to the similarity matrix 205. If the running state data queue 201 has not been traversed, the running state data queue 201 is sequentially offset according to a predetermined step length, and the above steps of constructing the similarity matrix 205 are repeated after the offset. When the running state data queue 201 is traversed, the row vector can be taken out from the similarity matrix 205, and the average value 206 of the row vector is calculated. In this way, each row can obtain an average value 206, and these average values ​​206 are arranged according to the position in the similarity matrix 205 to obtain a sample feature vector for constructing similar features.

[0101] In terms of constructing a standard state data queue, the implementation method of the present invention starts from multiple sample state data queues and obtains them by clustering (for example, kmeans clustering or DBSCAN clustering). Specifically, multiple sample state data queues are clustered and the center of the cluster is extracted: the sample class center. Then, the sample class center is unitized using the following formula:

[0102]

[0103] Where CC(n) is the nth element of the sample class center, CCN(n) is the nth element of the unit class center, and nmax is the total number of sample class center elements.

[0104] Then, the obtained unit class centers are subjected to pairwise orthogonalization test using the second formula, which is:

[0105]

[0106] Where QN i (n) is the nth data of the center of the i-th unit class, QN j (n) is the nth data of the jth unit class center, Orth(i,j) is the orthogonalization test result of the ith unit class center and the jth unit class center, and nmax is the total number of data in the unit class center.

[0107] If the average value of the orthogonalization test result is less than the orthogonalization threshold, the unit class center is used as the standard state data queue, otherwise, the number of clusters is adjusted and the clustering step is repeated. Generally speaking, the number of clusters is adjusted to a smaller value.

[0108] In step 103, each sample feature vector is classified into the nearest standard sample feature class, and the state of the standard sample feature class is used as the state of the sample feature vector, thereby obtaining multiple states.

[0109] In some implementations, classifying each sample feature vector into the nearest standard sample feature class and using the state of the standard sample feature class as the state of the sample feature vector includes:

[0110] For each sample feature vector, perform the following steps:

[0111] Obtain the class centers of multiple standard sample feature classes;

[0112] Calculating the distances between the sample feature vectors and the class centers of the plurality of standard sample feature classes respectively as a plurality of target distances;

[0113] Select the standard sample feature class corresponding to the smallest target distance as the target class;

[0114] The state of the target class is taken as the state of the sample feature vector.

[0115] In some embodiments, after selecting the standard sample feature class corresponding to the target distance with the smallest value as the target class, the method includes:

[0116] Get the radius of the target class;

[0117] Taking the distance between the class center of the target class and the sample feature vector as the first distance;

[0118] If the ratio of the first distance to the radius of the target class is greater than a ratio threshold, a message indicating that the power transmission and transformation equipment is abnormal is output.

[0119] Exemplarily, in terms of state classification, the embodiment of the present invention classifies the sample feature vector into the nearest feature class, and extracts the state of the nearest feature class as the class of the sample feature vector.

[0120] For example, after the above steps, the current state data queue obtains a current sample feature vector. This vector is distanced with the class centers of multiple standard current sample feature classes (for example, Euclidean distance calculation) to obtain the distances from the current sample feature vector to the class centers of multiple standard current sample feature classes. The standard current sample feature class with the shortest distance is selected as its target class, and the state of the target class is used as the state of the current sample feature vector (we can mark the state of the current state data queue as S k ).

[0121] Similarly, we can extract the status of the voltage status data queue.

[0122] In addition, when calculating the distance, we may find that even the smallest distance is much larger than the radius of the target class. This means that the current current sample feature distance is far from the class center distance of multiple standard sample feature classes. Therefore, we can determine that there is an abnormality in the power transmission and transformation equipment.

[0123] Step 104, determining the probability of the current power transmission and transformation equipment being abnormal based on the multiple states, the probability of the abnormality occurring, and the probability of the multiple states occurring.

[0124] In some implementations, determining the probability of the current power transmission and transformation equipment being abnormal based on the multiple states, the probability of the abnormality occurring, and the probability of the multiple states occurring includes:

[0125] According to the third formula, the multiple states, the probability of abnormal occurrence and the probability of multiple states occurring, the probability of abnormality of the current power transmission and transformation equipment is determined, wherein the third formula is:

[0126]

[0127] In the formula, P(E│S 1 S 2 …S k ) is the probability of abnormality of the current power transmission and transformation equipment, P(S k │E) When an abnormality occurs, the power transmission and transformation equipment is in S k The probability of the state, P(E) is the probability of abnormal state of the power transmission and transformation equipment, P(S k ) is the power transmission and transformation equipment in S k The probability of the state.

[0128] For example, after obtaining the status of different dimensions of the power transmission and transformation equipment, we can use the third formula to determine the probability of abnormality of the current power transmission and transformation station equipment:

[0129]

[0130] In the formula, P(E│S 1 S 2 …S k ) is the probability of abnormality of the current power transmission and transformation equipment, P(S k │E) When an abnormality occurs, the power transmission and transformation equipment is in S k The probability of the state, P(E) is the probability of abnormal state of the power transmission and transformation equipment, P(S k ) is the power transmission and transformation equipment in S k The probability of the state.

[0131] In the above formula, the probability of abnormal state of power transmission and transformation equipment and the probability of the state of a certain dimension of power transmission and transformation equipment when the abnormality occurs can be obtained through statistics. Similarly, the probability of power transmission and transformation equipment being in a certain state can also be obtained through statistics, and can also be obtained through statistics of multiple standard sample feature classes in the previous steps.

[0132] The embodiment of the present invention first obtains multiple operation status data queues, wherein the operation status data queues characterize the fluctuation of the transmission and transformation voltage or current of the power transmission and transformation equipment over time; then, sample similarity extraction is performed on each operation status data queue according to multiple standard status data queues, and the extraction results are constructed as sample feature vectors, thereby obtaining multiple sample feature vectors; then, each sample feature vector is classified into the nearest standard sample feature class, and the state of the standard sample feature class is used as the state of the sample feature vector, thereby obtaining multiple states; finally, according to the multiple states, the probability of abnormal occurrence and the probability of multiple states occurring, the probability of abnormality of the current power transmission and transformation equipment is determined. The embodiment of the present invention extracts the sample feature vector of the operation status data queue by comparing the similarity with the standard status data queue, and then divides the state of the power transmission and transformation equipment in multiple dimensions according to the sample feature vector, and finally calculates the probability of abnormality according to the state division using the Bayesian formula. Since the abnormal analysis and judgment is based on dimensional data, the accuracy is high, and compared with the artificial intelligence algorithm, the required sample size is smaller, the modeling process is simpler, and the universality is higher.

[0133] It should be understood that the size of the serial numbers of the steps in the above implementation does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the implementation method of the present invention.

[0134] The following is an implementation of the device of the present invention. For details not described in detail, reference may be made to the corresponding method implementation described above.

[0135] Figure 3 The schematic diagram of the structure of the abnormal warning device for power transmission and transformation equipment provided by the embodiment of the present invention is shown. For the convenience of explanation, only the part related to the embodiment of the present invention is shown, which is described in detail as follows:

[0136] like Figure 3 As shown, the abnormal warning device 3 for power transmission and transformation equipment includes: a data acquisition module 301, a feature extraction module 302, a state analysis module 303 and an abnormality analysis module 304, wherein:.

[0137] The data acquisition module 301 is used to acquire a plurality of operation status data queues, wherein the operation status data queues represent the fluctuation of the power transmission and transformation voltage or current of the power transmission and transformation equipment over time;

[0138] A feature extraction module 302 is used to extract sample similarity from each running state data queue according to multiple standard state data queues, and construct the extraction results into a sample feature vector, thereby obtaining multiple sample feature vectors;

[0139] The state analysis module 303 is used to classify each sample feature vector into the nearest standard sample feature class, and use the state of the standard sample feature class as the state of the sample feature vector, thereby obtaining multiple states;

[0140] The abnormality analysis module 304 is used to determine the probability of abnormality of the current power transmission and transformation equipment according to the multiple states, the probability of abnormality occurrence and the probability of occurrence of multiple states.

[0141] In the above-mentioned embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0142] Those skilled in the art will appreciate that the templates, units, and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0143] If the module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned implementation method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, it can implement the steps of the above-mentioned implementation methods of the abnormal warning method of each power transmission and transformation equipment. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory, random access memory, electrical carrier signal, telecommunication signal and software distribution medium.

[0144] The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for abnormal warning of power transmission and transformation equipment, characterized in that: include: Acquire multiple operation status data queues, wherein the operation status data queues represent the fluctuation of the power transmission and transformation voltage or current of the power transmission and transformation equipment over time; Extracting sample similarity from each running state data queue according to the multiple standard state data queues, and constructing the extraction results into a sample feature vector, thereby obtaining multiple sample feature vectors; Classify each sample feature vector into the nearest standard sample feature class, and use the state of the standard sample feature class as the state of the sample feature vector, thereby obtaining multiple states; The probability of the current power transmission and transformation equipment being abnormal is determined based on the multiple states, the probability of the abnormality occurring, and the probability of the multiple states occurring.

2. The abnormal warning method for power transmission and transformation equipment according to claim 2 is characterized in that: The method of performing sample similarity extraction on each running state data queue according to a plurality of standard state data queues and constructing the extraction result into a sample feature vector includes: Get multiple standard status data queues; For each of the multiple running status data queues, the following steps are performed respectively: Starting from the first position of the running state data queue, continuously take out data with the same total amount of data as the standard state data queue to construct a queue to be processed; Calculating the similarities between the queue to be processed and the plurality of standard state data queues respectively, and taking the obtained plurality of similarities as column vectors and adding them into a similarity matrix; If the traversal is not completed, the first position is offset, and the process jumps to the step of continuously taking out data having the same total amount of data as the standard state data queue from the first position of the running state data queue to construct a queue to be processed; Otherwise, the average value of the elements in each row of the similarity matrix is ​​extracted, and the extracted multiple average values ​​are constructed as a frequency feature vector.

3. The abnormal warning method for power transmission and transformation equipment according to claim 2 is characterized in that: The respectively calculating the similarities between the queue to be processed and the plurality of standard status data queues includes: The similarity between the queue to be processed and the plurality of standard state data queues is calculated according to a first formula, wherein the first formula is: Where SM is the similarity between the queue to be processed and the standard state data queue, Q(n) is the nth data in the queue to be processed, QS(n) is the nth data in the standard state data queue, and nmax is the total number of data in the standard state data queue.

4. The abnormal warning method for power transmission and transformation equipment according to claim 2 is characterized in that: The acquisition of the plurality of standard status data queues includes: Get multiple sample status data queues and cluster numbers; Clustering the plurality of sample state data queues according to the clustering quantity, thereby obtaining a plurality of sample classes; Extracting the class centers of the multiple sample classes to obtain multiple sample class centers; Unitizing the multiple sample class centers to obtain multiple unit class centers; Performing orthogonalization tests on the multiple unit class centers in pairs to obtain multiple test results; If the average value of the plurality of test results is less than the orthogonalization threshold, the plurality of unit class centers are used as a plurality of standard state data queues; Otherwise, the clustering quantity is adjusted and the process jumps to the step of clustering the plurality of sample status data queues according to the clustering quantity to obtain a plurality of sample classes.

5. The abnormal warning method for power transmission and transformation equipment according to claim 4 is characterized in that: The step of performing orthogonalization tests on the plurality of unit class centers in pairs to obtain multiple test results includes: The orthogonalization test is performed on the multiple unit class centers in pairs according to the second formula to obtain multiple test results, wherein the second formula is: Where QN i (n) is the nth data of the center of the i-th unit class, QN j (n) is the nth data of the jth unit class center, Orth(i,j) is the orthogonalization test result of the ith unit class center and the jth unit class center, and nmax is the total number of data in the unit class center.

6. The abnormal warning method for power transmission and transformation equipment according to claim 1 is characterized in that: The step of classifying each sample feature vector into the nearest standard sample feature class and taking the state of the standard sample feature class as the state of the sample feature vector includes: For each sample feature vector, perform the following steps: Obtain the class centers of multiple standard sample feature classes; Calculating the distances between the sample feature vectors and the class centers of the plurality of standard sample feature classes respectively as a plurality of target distances; Select the standard sample feature class corresponding to the smallest target distance as the target class; The state of the target class is taken as the state of the sample feature vector.

7. The abnormal warning method for power transmission and transformation equipment according to claim 6, characterized in that: After selecting the standard sample feature class corresponding to the target distance with the smallest value as the target class, it includes: Get the radius of the target class; Taking the distance between the class center of the target class and the sample feature vector as the first distance; If the ratio of the first distance to the radius of the target class is greater than a ratio threshold, a message indicating that the power transmission and transformation equipment is abnormal is output.

8. The abnormal warning method for power transmission and transformation equipment according to any one of claims 1 to 7, characterized in that: The determining the probability of the current power transmission and transformation equipment being abnormal according to the multiple states, the probability of the abnormality occurring, and the probability of the multiple states occurring includes: According to the third formula, the multiple states, the probability of abnormal occurrence and the probability of multiple states occurring, the probability of abnormality of the current power transmission and transformation equipment is determined, wherein the third formula is: Where, P(E│S1S2…S k ) is the probability of abnormality of the current power transmission and transformation equipment, P(S k │E) When an abnormality occurs, the power transmission and transformation equipment is in S k The probability of the state, P(E) is the probability of abnormal state of the power transmission and transformation equipment, P(S k ) is the power transmission and transformation equipment in S k The probability of the state.

9. An abnormal warning device for power transmission and transformation equipment, characterized in that: Used to implement the abnormal warning method for power transmission and transformation equipment according to any one of claims 1 to 8, the abnormal warning device for power transmission and transformation equipment comprises: A data acquisition module is used to acquire a plurality of operation status data queues, wherein the operation status data queues represent the fluctuation of the power transmission and transformation voltage or current of the power transmission and transformation equipment over time; A feature extraction module is used to extract sample similarity from each running state data queue according to multiple standard state data queues, and construct the extraction results into a sample feature vector, thereby obtaining multiple sample feature vectors; A state analysis module is used to classify each sample feature vector into the nearest standard sample feature class, and use the state of the standard sample feature class as the state of the sample feature vector, thereby obtaining multiple states; as well as, The abnormality analysis module is used to determine the probability of abnormality of the current power transmission and transformation equipment according to the multiple states, the probability of abnormality occurrence and the probability of multiple states occurring.

10. The abnormal warning device for power transmission and transformation equipment according to claim 9, characterized in that: The abnormality analysis module comprises: The abnormality probability calculation module is used to determine the probability of abnormality of the current power transmission and transformation equipment according to a third formula, the multiple states, the probability of abnormality and the probability of multiple states, wherein the third formula is: Where, P(E│S1S2…S k ) is the probability of abnormality of the current power transmission and transformation equipment, P(S k │E) When an abnormality occurs, the power transmission and transformation equipment is in S k The probability of the state, P(E) is the probability of abnormal state of the power transmission and transformation equipment, P(S k ) is the power transmission and transformation equipment in S k The probability of the state.

Images