A method, device and system for detecting faults in a relay protection circuit

Abstract

The present invention relates to the technical field of measuring electrical variables, and specifically relates to a method, device and system for detecting faults in a relay protection circuit. The method includes: obtaining multi-dimensional data of a plurality of current transformers in a target relay protection circuit; analyzing the change trend of the multi-dimensional data to obtain initial fault possibilities corresponding to a plurality of preset fault types; analyzing the time performance of the characteristic distance corresponding to the multi-dimensional data, adjusting the characteristic distance corresponding to the multi-dimensional data according to the analysis result of the time performance, and adjusting the initial fault possibility according to the adjusted characteristic distance corresponding to the multi-dimensional data to obtain target fault possibilities corresponding to a plurality of preset fault types; determining a target fault category corresponding to each current transformer according to the target fault possibility. This application measures the electrical variables of current transformers to obtain multi-dimensional data, and analyzes the change trend and time performance of the multi-dimensional data to improve the accuracy of fault detection in the relay protection circuit.

Description

Technical Field

[0001] The present invention relates to the technical field of measuring electrical variables, and particularly to a method, device and system for detecting faults in a relay protection circuit. Background Art

[0002] In the context of increasing electricity demand, the security of the power system also needs to be improved. The relay protection circuit composed of a relay protection device and a secondary circuit is an important protection device in the power system. The relay protection device is used to detect faults in the power system and take protection measures, and the secondary circuit is a circuit system to achieve this function. The secondary circuit is an electrical circuit formed by connecting secondary equipment in the power system and is responsible for monitoring and protecting primary equipment. The secondary circuit consists of four parts: an AC secondary current and voltage circuit, a power supply, a signal circuit, and a control circuit. At the same time, the substation has changed from the traditional mode to an intelligent substation, and through the secondary circuit, a comprehensive detection of the operating status of the automation equipment in the intelligent substation can be realized.

[0003] In the prior art, there is a relay protection circuit fault detection method based on the multi-dimensional Hausdorff distance (a distance used to quantify the distance between two subsets). By calculating whether the Hausdorff distances of current, zero-sequence current, and resistance are normal, it is determined whether the circuit is faulty and the type and location of the fault. However, this method only calculates the fault situation at a single moment and the Hausdorff distance of a single feature, which will lead to deviations in the calculation results, and further lead to insufficient accuracy of fault detection due to the deviation values.

[0004] Therefore, how to improve the accuracy of relay protection circuit fault detection is an urgent problem to be solved at present. Summary of the Invention

[0005] In order to solve the technical problem of how to improve the accuracy of relay protection circuit fault detection, the purpose of the present invention is to provide a method, device and system for detecting faults in a relay protection circuit. The specific technical solutions adopted are as follows:

[0006] An embodiment of the present application provides a method for detecting faults in a relay protection circuit, the method comprising:

[0007] Obtain multi-dimensional data of a plurality of current transformers in a target relay protection circuit, the multi-dimensional data including current time series data, zero-sequence current time series data, and resistance time series data;

[0008] For any one of the current transformers, analyze the change trend of the multi-dimensional data to obtain initial fault probabilities corresponding to a plurality of preset fault types;

[0009] Analyze the time performance of the feature distances corresponding to the multi-dimensional data, adjust the feature distances corresponding to the multi-dimensional data according to the analysis results of the time performance, and adjust the initial fault probabilities according to the adjusted feature distances corresponding to the multi-dimensional data to obtain the target fault probabilities corresponding to the multiple preset fault types;

[0010] Determine the target fault categories corresponding to the current transformers according to the target fault probabilities, and perform fault repairs on the current transformers according to the target fault categories;

[0011] The analysis of the time performance of the feature distances corresponding to the multi-dimensional data includes:

[0012] Obtain the current distance, zero-sequence current distance, and resistance distance according to the multi-dimensional data of any two current transformers, where the current distance refers to the distance between the current values of two current transformers within the same time window, the zero-sequence current distance refers to the distance between the zero-sequence current values of two current transformers within the same time window, and the resistance distance refers to the distance between the resistance values of two current transformers within the same time window;

[0013] Use the current distance, zero-sequence current distance, and resistance distance as the feature distances;

[0014] Obtain the time performance of the feature distances of each current transformer according to the numerical magnitudes and durations of the feature distances;

[0015] The adjustment of the feature distances corresponding to the multi-dimensional data according to the analysis results of the time performance includes:

[0016] Adjust the current distance according to the time performance of the current distance to obtain the adjusted current distance;

[0017] Adjust the zero-sequence current distance according to the time performance of the zero-sequence current distance to obtain the adjusted zero-sequence current distance;

[0018] Adjust the resistance distance according to the time performance of the resistance distance to obtain the adjusted resistance distance.

[0019] In some embodiments, after obtaining the multi-dimensional data of multiple current transformers in the target relay protection circuit, it further includes:

[0020] Perform discrete wavelet transform on the multi-dimensional data according to a preset wavelet basis function and a preset decomposition level to obtain initial approximation coefficients and initial detail coefficients;

[0021] Process the initial approximation coefficients and initial detail coefficients according to a preset soft threshold function to obtain a target approximation function and a target detail function;

[0022] Perform an inverse discrete wavelet transform on the target approximation function and the target detail function to obtain the denoised multi-dimensional data, and the denoised multi-dimensional data is used for change trend analysis.

[0023] In some embodiments, the multiple preset fault types include loop short circuit, loop ground short circuit, loop open circuit, and short circuit of a specific numbered loop. Analyzing the change trend of the multi-dimensional data to obtain the initial fault probabilities corresponding to the multiple preset fault types includes:

[0024] Obtain a time window according to a preset neighborhood monitoring time.

[0025] Analyze the change trend of the current time series data, and obtain the initial fault probability corresponding to the loop short circuit according to the average current within the time window.

[0026] Analyze the change trends of the current time series data and the zero-sequence current time series data, and obtain the initial fault probability of the loop ground short circuit according to the average current and zero-sequence current within the time window.

[0027] Analyze the change trend of the resistance time series data, and obtain the initial fault probability corresponding to the loop open circuit according to the average resistance within the time window.

[0028] Analyze the change trend of the resistance time series data and the phase change of the current time series data, and obtain the initial fault probability corresponding to the short circuit of the specific numbered loop according to the average resistance and phase change within the time window.

[0029] In some embodiments, adjusting the initial fault probabilities according to the characteristic distances corresponding to the adjusted multi-dimensional data to obtain the target fault probabilities corresponding to the multiple preset fault types includes:

[0030] Obtain an adjusted multi-dimensional Hausdorff distance according to the adjusted current distance, the adjusted zero-sequence current distance, and the adjusted resistance distance.

[0031] Determine an adjustment index of the characteristic distance according to the adjusted multi-dimensional Hausdorff distance.

[0032] Adjust the initial fault probabilities according to the adjustment index of the characteristic distance to obtain the target fault probabilities corresponding to the multiple preset fault types.

[0033] In some embodiments, determining the adjustment index of the characteristic distance according to the adjusted multi-dimensional Hausdorff distance includes:

[0034] Obtain the period of the adjusted multi-dimensional Hausdorff distance.

[0035] Based on the variance of the characteristic distances corresponding to multiple time windows within the period, a periodic pattern is obtained;

[0036] Based on the periodic pattern, as well as the number of the time windows and the range of the characteristic distances within each time window, an adjustment index of the characteristic distances is determined.

[0037] In some embodiments, the fault repair of each current transformer according to the target fault category includes:

[0038] Based on the target fault category, a fault loop is determined;

[0039] The states of the lines and equipment connected to the current transformer in the fault loop are checked, and the lines and equipment with abnormal states are repaired for faults.

[0040] An embodiment of the present application further provides a relay protection loop fault detection device, and the device includes:

[0041] A data acquisition module, configured to acquire multi-dimensional data of multiple current transformers in a target relay protection loop, where the multi-dimensional data includes current time series data, zero-sequence current time series data, and resistance time series data;

[0042] An initial fault analysis module, configured to analyze the change trend of the multi-dimensional data for any one of the current transformers to obtain the initial fault probabilities corresponding to multiple preset fault types;

[0043] A target fault analysis module, configured to analyze the time performance of the characteristic distances corresponding to the multi-dimensional data, adjust the characteristic distances corresponding to the multi-dimensional data according to the analysis result of the time performance, and adjust the initial fault probabilities according to the adjusted characteristic distances corresponding to the multi-dimensional data to obtain the target fault probabilities corresponding to the multiple preset fault types;

[0044] A fault category determination module, configured to determine the target fault category corresponding to each current transformer according to the target fault probabilities, and perform fault repair on each current transformer according to the target fault category.

[0045] An embodiment of the present application further provides a relay protection loop fault detection system, and the system includes a computer-readable storage medium, where the computer-readable storage medium stores multiple instructions, and the instructions are suitable for being loaded by a processor to execute the steps in the relay protection loop fault detection method as described above.

[0046] The present invention has the following beneficial effects:

[0047] First, obtain the multi-dimensional data of multiple current transformers in the target relay protection circuit. The multi-dimensional data includes current time-series data, zero-sequence current time-series data, and resistance time-series data. Then, for any one of the current transformers, analyze the change trend of the multi-dimensional data to obtain the initial fault probabilities corresponding to multiple preset fault types. Then, analyze the time performance of the characteristic distance corresponding to the multi-dimensional data, adjust the characteristic distance corresponding to the multi-dimensional data according to the analysis result of the time performance, and adjust the initial fault probability according to the adjusted characteristic distance corresponding to the multi-dimensional data to obtain the target fault probabilities corresponding to the multiple preset fault types. Finally, determine the target fault categories corresponding to each current transformer according to the target fault probabilities, and perform fault repair on each current transformer according to the target fault categories. In this application, obtaining current time-series data, zero-sequence current time-series data, and resistance time-series data provides multi-dimensional information, which can more comprehensively reflect the state of the relay protection circuit. Different types of time-series data can complement each other, reduce the limitations of a single data source, and improve the reliability of fault detection. By analyzing the change trend of multi-dimensional data, the trends and patterns of the data can be identified, which helps to detect potential fault signs early. Dynamically adjusting the characteristic distance according to the time performance of the characteristic distance can more accurately reflect the actual fault situation. Static characteristic distances cannot adapt to different fault scenarios, while dynamic adjustment can improve adaptability and accuracy. Combining the adjusted characteristic distance and re-evaluating the initial fault probability to obtain a more accurate target fault probability can eliminate misjudgments and improve the accuracy of fault detection. According to the target fault probability, the specific fault category can be accurately located, avoiding blind repair and improving the repair efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0049] Figure 1 Schematic diagram of the implementation environment of a relay protection circuit fault detection method provided by an embodiment of the present invention;

[0050] Figure 2 Flowchart of a relay protection circuit fault detection method provided by an embodiment of the present invention;

[0051] Figure 3 Schematic diagram of the structure of a relay protection circuit fault detection device provided by an embodiment of the present invention. Detailed implementation manners

[0052] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following specifically describes, in conjunction with the accompanying drawings and preferred embodiments, a relay protection circuit fault detection method, device and system proposed according to the present invention, including its specific implementation manners, structures, features and effects. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0053] It should be noted that the terms "first", "second", etc. in the specification of this application and the above-mentioned accompanying drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0055] The following specifically describes the specific solutions of a relay protection circuit fault detection method, device and system provided by the present invention with reference to the accompanying drawings.

[0056] Please refer to Figure 1 , Figure 1 which is a schematic diagram of the implementation environment of a relay protection circuit fault detection method provided by an embodiment of the present invention. As Figure 1As shown in the figure, the implementation environment includes a fault detection terminal 101 and a relay protection circuit 102. The fault detection terminal 101 may be a terminal device configured with a relay protection circuit fault detection system, including but not limited to a laptop computer, a tablet computer, a personal digital assistant (PDA), a PAD, a desktop computer, etc. with local computing capabilities; the relay protection circuit fault detection system may be implemented in the form of a target client, and the target client may be a video client, an instant messaging client, a browser client, etc. that support relay protection circuit fault detection; the fault detection terminal 101 may communicate with the relay protection circuit 102 through a network, which may include but not limited to: a wired network, a wireless network, where the wired network includes: a local area network, a metropolitan area network, and a wide area network, and the wireless network includes: Bluetooth, WIFI, and other networks that implement wireless communication. The above-mentioned fault detection terminal 101 may include but not limited to a human-computer interaction screen, a processor, and a memory. The above-mentioned human-computer interaction screen may be used to display the target fault category. The above-mentioned processor may be used to respond to human-computer interaction operations, execute corresponding operations, or generate corresponding instructions.

[0057] As an optional method, multi-dimensional data of multiple current transformers in the relay protection circuit may be collected through the relay protection circuit 102, and the relay protection circuit 102 may include multiple current transformers.

[0058] As an optional method, the above-mentioned fault detection terminal 101 may also be a server, which may be a single server, a server cluster composed of multiple servers, or a cloud server. The above is only an example, and no limitation is made in this embodiment.

[0059] As an optional method, the following steps of the relay protection circuit fault detection method may be executed on the fault detection terminal 101:

[0060] Obtain multi-dimensional data of multiple current transformers in the target relay protection circuit, where the multi-dimensional data includes current time series data, zero-sequence current time series data, and resistance time series data;

[0061] For any one of the current transformers, analyze the change trend of the multi-dimensional data to obtain the initial fault probabilities corresponding to multiple preset fault types;

[0062] Analyze the time performance of the characteristic distance corresponding to the multi-dimensional data, adjust the characteristic distance corresponding to the multi-dimensional data according to the analysis result of the time performance, and adjust the initial fault probability according to the adjusted characteristic distance corresponding to the multi-dimensional data to obtain the target fault probabilities corresponding to the multiple preset fault types;

[0063] Based on the target fault possibility, determine the target fault category corresponding to each current transformer, and perform fault repair on each current transformer according to the target fault category.

[0064] Based on the above method, obtaining current time series data, zero-sequence current time series data, and resistance time series data provides multi-dimensional information, which can more comprehensively reflect the state of the relay protection circuit. Different types of time series data can complement each other, reduce the limitations of a single data source, and improve the reliability of fault detection; by analyzing the changing trends of multi-dimensional data, the trends and patterns of the data can be identified, which helps to detect potential fault signs at an early stage; dynamically adjusting the characteristic distance according to the time performance of the characteristic distance can more accurately reflect the actual fault situation. Static characteristic distances cannot adapt to different fault scenarios, while dynamic adjustment can improve adaptability and accuracy; combining the adjusted characteristic distance to re-evaluate the initial fault possibility, a more accurate target fault possibility can be obtained, which can eliminate misjudgment and improve the accuracy of fault detection; according to the target fault possibility, the specific fault category can be accurately located, avoiding blind repair and improving the repair efficiency.

[0065] As an optional example, the execution subject of the above relay protection circuit fault detection method is not limited in this embodiment. The above relay protection circuit fault detection method can be executed on the fault detection terminal 101. For example, when the fault detection terminal 101 is a desktop computer, some or all of the steps of the above relay protection circuit fault detection method can be executed on the desktop computer.

[0066] The above part introduced the content of the exemplary implementation environment applying the technical solution of the present application. Next, the relay protection circuit fault detection method of the present application will be continued to be introduced.

[0067] To solve the problem of how to improve the accuracy of relay protection circuit fault detection in the prior art, embodiments of the present application respectively propose a relay protection circuit fault detection method, a relay protection circuit fault detection device, an electronic device, and a relay protection circuit fault detection system. These embodiments will be described in detail below.

[0068] Please refer to Figure 2 , Figure 2 is a schematic flowchart of a relay protection circuit fault detection method provided by an embodiment of the present invention. This method can be applied to Figure 1 the implementation environment shown. It should be understood that this method can also be applicable to other exemplary implementation environments and be specifically executed by devices in other implementation environments. This embodiment does not limit the implementation environment applicable to this method.

[0069] As Figure 2As shown, in an exemplary embodiment, the relay protection circuit fault detection method at least includes steps S210 to S240, which are introduced in detail as follows:

[0070] In step S210, multi-dimensional data of multiple current transformers in the target relay protection circuit is acquired, and the multi-dimensional data includes current time series data, zero-sequence current time series data, and resistance time series data.

[0071] Among them, a current transformer is a device used to measure alternating current. By converting large current into small current, the current can be safely measured and monitored. Current time series data refers to the sequence of current values continuously collected by the current transformer over a period of time. By analyzing the current time series data, the change trend of the current in the circuit can be understood, current anomalies can be identified, and whether there is a fault can be judged. Zero-sequence current time series data refers to the sequence of zero-sequence current values collected by the zero-sequence current transformer in a three-phase system. The zero-sequence current is the vector sum of the three-phase currents and is mainly used to detect grounding faults. Resistance time series data refers to the sequence of loop resistance values collected by a resistance sensor or other measuring devices over a period of time, which is obtained by dividing the secondary sampling circuit voltage of the current transformer by the phase current data. By analyzing the resistance time series data, faults such as poor contact and resistance change in the circuit can be detected.

[0072] Exemplarily, there is a relay protection circuit, which includes three current transformers, a zero-sequence current transformer, and a resistance sensor. The above current transformers are respectively installed at different positions in the circuit to collect three-phase current data; the above zero-sequence current transformer is installed on the neutral line of the three-phase system to collect zero-sequence current data; the above resistance sensor is installed at the key nodes of the circuit to collect loop resistance data.

[0073] In step S220, for any one of the current transformers, the change trend of the multi-dimensional data is analyzed to obtain the initial fault probabilities corresponding to multiple preset fault types.

[0074] Among them, the change trend of the multi-dimensional data refers to the change trend and pattern of the multi-dimensional data (such as current time series data, zero-sequence current time series data, and resistance time series data) over a period of time. By analyzing the change trend of the multi-dimensional data, abnormal changes in the data can be identified, and then possible fault types can be inferred. The change trend can include features such as data fluctuations, mutations, and trends.

[0075] In step S230, analyze the time performance of the feature distance corresponding to the multi-dimensional data, adjust the feature distance corresponding to the multi-dimensional data according to the analysis result of the time performance, and adjust the initial fault possibility according to the adjusted feature distance corresponding to the multi-dimensional data, so as to obtain the target fault possibility corresponding to the multiple preset fault types.

[0076] Among them, the feature distance is a measurement method used to quantify the similarity between multi-dimensional data. In the fault detection of relay protection circuits, the feature distance can be used to compare the multi-dimensional data in the normal state and the current state, and evaluate the similarity and difference of the data. By calculating the feature distance, abnormal changes in the data can be identified, and then possible fault types can be inferred. Commonly used feature distance methods include multi-dimensional Hausdorff distance, Euclidean distance, etc. The time performance refers to the change trend and pattern of the feature distance over time. By analyzing the time performance of the feature distance, the development process and dynamic changes of the fault can be understood. The time performance analysis can help dynamically adjust the feature distance and improve the accuracy and real-time performance of fault detection. The time performance can include features such as the fluctuation, trend, and mutation of the feature distance.

[0077] In step S240, determine the target fault category corresponding to each current transformer according to the target fault possibility, and perform fault repair on each current transformer according to the target fault category.

[0078] Exemplarily, there is a relay protection circuit, which includes three current transformers, a zero-sequence current transformer, and a resistance sensor. The multi-dimensional data of each current transformer are obtained through the previous steps, and the target fault possibility is calculated. Now it is necessary to determine the target fault category of each current transformer according to the target fault possibility and perform fault repair. When the target fault category is an open circuit fault, check and replace the connecting wire, and test and verify after reconnecting. When the target fault category is a single-phase grounding fault, check and repair the grounding, and test and verify after reconnecting. When the target fault category is poor contact, clean and tighten the contact point, and test and verify after reconnecting.

[0079] As can be seen from the above steps S210 to S240, the solution proposed in this embodiment obtains current time-series data, zero-sequence current time-series data, and resistance time-series data, providing multi-dimensional information that can more comprehensively reflect the state of the relay protection circuit. Different types of time-series data can complement each other, reducing the limitations of a single data source and improving the reliability of fault detection. By analyzing the changing trends of multi-dimensional data, the trends and patterns of the data can be identified, which helps to detect potential fault signs at an early stage. Dynamically adjusting the characteristic distance according to the time performance of the characteristic distance can more accurately reflect the actual fault situation. Static characteristic distances cannot adapt to different fault scenarios, while dynamic adjustment can improve adaptability and accuracy. Combining the adjusted characteristic distance and re-evaluating the initial fault possibility can obtain a more accurate target fault possibility, eliminating misjudgments and improving the accuracy of fault detection. Based on the target fault possibility, the specific fault category can be accurately located, avoiding blind maintenance and improving the maintenance efficiency.

[0080] In an embodiment of the present application, after obtaining the multi-dimensional data of multiple current transformers in the target relay protection circuit, the method further includes:

[0081] Performing discrete wavelet transform on the multi-dimensional data according to a preset wavelet basis function and a preset decomposition level to obtain initial approximation coefficients and initial detail coefficients;

[0082] Processing the initial approximation coefficients and the initial detail coefficients according to a preset soft threshold function to obtain a target approximation function and a target detail function;

[0083] Performing inverse discrete wavelet transform on the target approximation function and the target detail function to obtain the denoised multi-dimensional data, and the denoised multi-dimensional data is used for change trend analysis.

[0084] Among them, the wavelet basis function is a set of mathematical functions used for multi-scale analysis of signals in discrete wavelet transform. Selecting an appropriate wavelet basis function can better extract the features of the signal and improve the denoising effect.

[0085] Among them, discrete wavelet transform (also known as DWT) is a method of decomposing a signal into different frequency components. Through discrete wavelet transform, multi-dimensional data can be decomposed into approximation coefficients and detail coefficients. Through DWT, a signal can be decomposed into high-frequency and low-frequency parts of different scales, which helps to remove noise and retain the main features of the signal.

[0086] Among them, the initial approximation coefficients represent the low-frequency part of the signal, and the initial detail coefficients represent the high-frequency part of the signal. The approximation coefficients reflect the overall trend of the signal, and the detail coefficients reflect the local changes of the signal. By analyzing these coefficients, the noise in the signal can be identified and processed.

[0087] Among them, the soft threshold function is a method for removing the part of the detail coefficients smaller than the threshold after wavelet transform. Through the soft threshold function, the noise in the signal can be effectively removed while retaining the main features of the signal. Processing the initial approximation coefficients and initial detail coefficients using the soft threshold function can remove noise, retain the main features of the signal, and improve the quality of the signal.

[0088] Among them, the inverse discrete wavelet transform is the inverse process of the discrete wavelet transform, which is used to recombine the processed approximation coefficients and detail coefficients into the denoised signal. It can restore the denoised approximation coefficients and detail coefficients to the denoised multi-dimensional data for further analysis and processing.

[0089] Exemplarily, the preset wavelet basis function is the dbN wavelet function, and the preset decomposition level is 4 levels. The denoised multi-dimensional data is denoted as , , . Among them, represents the multi-dimensional data at the i-th current transformer, n represents the number of current transformers, respectively represent the current time series data, zero-sequence current time series data, and resistance time series data at the i-th current transformer, represents the current data at the T-th moment at the i-th current transformer, and T is the total number of monitoring moments. In this embodiment, through wavelet transform and soft threshold processing, the noise in the multi-dimensional data is effectively removed, the quality of the signal is improved, and the subsequent analysis is more accurate. The denoised multi-dimensional data can more clearly reflect the change trend and pattern of the signal, which helps to more accurately identify and locate faults, improve the sensitivity and accuracy of fault detection. The denoising process can reduce misjudgments caused by noise, ensure the reliability of the fault detection results, and reduce the false alarm rate and missed alarm rate.

[0090] In an embodiment of the present application, the multiple preset fault types include loop short circuit, loop ground short circuit, loop open circuit, and short circuit of a specific numbered loop. Analyzing the change trend of the multi-dimensional data to obtain the initial fault probabilities corresponding to the multiple preset fault types includes:

[0091] Obtain a time window according to the preset neighborhood monitoring moment;

[0092] Analyze the change trend of the current time series data, and obtain the initial fault probability corresponding to the loop short circuit according to the average current within the time window;

[0093] Analyze the change trends of the current time series data and the zero-sequence current time series data, and obtain the initial fault probability of the loop ground short circuit according to the average current and zero-sequence current within the time window;

[0094] Analyze the change trend of the resistance time-series data, and obtain the initial fault possibility corresponding to the circuit open circuit according to the average resistance within the time window;

[0095] Analyze the change trend of the resistance time-series data and the phase change of the current time-series data, and obtain the initial fault possibility corresponding to the short circuit of a specific numbered circuit according to the average resistance and phase change within the time window.

[0096] Among them, there is a correlation between current, zero-sequence current, and resistance characteristics. By analyzing the change trend of multi-dimensional data, the initial fault possibility of any fault can be obtained. In addition, in order to reduce the false detection rate, the time of data anomaly should also be analyzed. When the anomaly manifestation is small, amplify the anomaly characteristics, adaptively adjust the anomaly manifestation of the characteristics, and count the time of the anomaly manifestation to comprehensively obtain the target fault possibility of any fault.

[0097] Among them, the common fault types are as follows: when the circuit is short-circuited, the current will decrease; when the circuit is grounded and short-circuited, the current value will decrease and the zero-sequence current will appear at the same time; when the circuit is open-circuited, the resistance will increase; when a specific numbered circuit is short-circuited, the resistance will increase and the phase of the current waveform will change at the same time. Therefore, the faults can be divided into two types: short circuit and open circuit. It can be found that under the two faults of short circuit and open circuit, the circuit short circuit and the circuit ground short circuit can be distinguished by whether the zero-sequence current appears, and the circuit open circuit and the short circuit of a specific numbered circuit can be distinguished by whether the phase of the current waveform changes.

[0098] Among them, the preset neighborhood monitoring moment refers to several key monitoring points preset within a period of time for collecting and analyzing data. By collecting data at these critical moments, the change trend of the signal can be captured more accurately, and the sensitivity of fault detection can be improved.

[0099] Among them, the time window refers to a period of time before and after the preset neighborhood monitoring moment for analyzing the change trend of data. By setting the time window, the data within a certain period of time can be analyzed intensively, noise interference can be reduced, and the accuracy of analysis can be improved.

[0100] Among them, by statistically analyzing the change trend of the current time-series data, the fluctuations, mutations, and trends of the data are identified. The analysis of the change trend can help identify the abnormal changes of the current, thereby inferring the possible fault types.

[0101] Among them, calculate the average value of the current time-series data within the time window. The average current can reflect the overall level of the current and is used to judge whether the current is within the normal range.

[0102] Among them, by statistically analyzing the changing trend of the resistance timing data, fluctuations, mutations, and trends in the data are identified. The analysis of the changing trend can help identify abnormal changes in the resistance, thereby inferring possible types of faults.

[0103] Among them, the average value of the resistance timing data is calculated within a time window. The average resistance can reflect the overall level of the resistance and is used to determine whether the resistance is within the normal range.

[0104] Among them, the phase change refers to the phase change of the current timing data within a time window. The phase change can help identify the phase relationship of the current and is used to determine the short - circuit fault of a specific numbered loop.

[0105] Exemplarily, an arbitrary moment, 5 monitoring moments in the left neighborhood, and 5 monitoring moments in the right neighborhood are combined as the time window for the said moment; the fault manifestation of a loop short - circuit is a decrease in current. Therefore, the changing trend of the current change can be analyzed. As the time window extends, if the average current decreases instead, it can indicate that the changing trend of the current is decreasing. Thus, a loop short - circuit fault can be detected, and the more obvious the degree of current decrease, the more serious the loop short - circuit fault manifestation. The representation of the initial fault possibility corresponding to the loop short - circuit can be:

[0106]

[0107] Among them, represents the initial fault possibility corresponding to the loop short - circuit of any current transformer at any moment; represents the size of the time window of the current transformer; 、 respectively represent the average currents at the r and s moments before the time window; is the maximum - minimum normalization function.

[0108] Among them, represents the sum of the differences between the average currents at all relatively few moments and the average currents at relatively many moments within the time window, indicating the changing trend of the current. The larger this difference, the greater the initial fault possibility corresponding to the loop short - circuit.

[0109] Exemplarily, on the basis of a loop short - circuit fault, if there is also zero - sequence current and the zero - sequence current is larger, then the fault possibility of loop ground short - circuit is greater. The representation of the initial fault possibility of loop ground short - circuit can be:

[0110]

[0111] Among them, represents the initial fault possibility of loop ground short - circuit of any current transformer at any moment; represents the initial fault possibility corresponding to the short - circuit of the loop at any time of any current transformer; is the zero - sequence current at the r - th moment before the time window; is the maximum - minimum normalization function.

[0112] Among them, represents the summation of the zero - sequence current within the time window. The larger this value is, the more likely it indicates the occurrence of zero - sequence current, and the greater the initial fault possibility of loop ground short - circuit.

[0113] Exemplarily, the loop open - circuit fault is manifested as an increase in resistance. It can be judged by the trend of resistance change. As the time window extends, if the average resistance increases, it can indicate that the trend of resistance change is rising. Therefore, the loop open - circuit fault can be detected, and the more obvious the degree of resistance increase, the more serious the manifestation of the loop open - circuit fault. The representation method of the initial fault possibility corresponding to the loop open - circuit can be:

[0114]

[0115] Among them, represents the initial fault possibility corresponding to the loop open - circuit of any current transformer at any time; represents the size of the time window of the current transformer; 、 respectively represent the average resistances at the r - th and s - th moments before the time window; is the maximum - minimum normalization function.

[0116] Among them, represents the summation of the differences between the average resistances of all relatively few moments and the average resistances of relatively many moments within the time window, representing the trend of resistance change. The larger this value is, the greater the initial fault possibility corresponding to the loop open - circuit.

[0117] Exemplarily, on the basis of the loop open - circuit fault, if there is also a phase change of the current, then the loop open - circuit fault and the short - circuit fault of a specific - numbered loop can be distinguished. The phase change can be distinguished by the period. As time changes, if the current representative data (extremum) no longer maintains the original periodic law, then it can be explained that the phase has changed. The representation method of the initial fault possibility corresponding to the short - circuit of a specific - numbered loop can be:

[0118]

[0119] Among them, represents the initial fault possibility corresponding to the short - circuit of a specific - numbered loop of any current transformer at any time; represents the initial fault possibility corresponding to the loop open - circuit of any current transformer at any time; Represents the number of phase maxima within a time window; Represents the , distance between the Represents the , distance between the is the maximum - minimum normalization function.

[0120] Among them, represents adding up the distance differences between all adjacent maxima, indicating phase change. The larger this value, the greater the degree of phase change. The greater the degree of phase change, the greater the initial fault probability corresponding to the short - circuit of a specific - numbered loop.

[0121] Exemplarily, the initial fault probabilities corresponding to multiple preset fault types can be integrated into , among which, represents the current - fault probability, represents the resistance - fault probability. When , it indicates loop short - circuit. When , it indicates loop ground short - circuit. When , it indicates loop open - circuit. When , it indicates short - circuit of a specific - numbered loop.

[0122] In this embodiment, by presetting the neighborhood monitoring time and time window, the data within the key time period can be centrally analyzed, reducing noise interference and improving the accuracy of fault detection. By analyzing the change trends of different types of multi - dimensional data (current time - series data, zero - sequence current time - series data, resistance time - series data), different fault types can be more accurately identified, improving the ability to identify fault types. By calculating the average value and change trend within the time window, misjudgments caused by instantaneous noise can be reduced, ensuring the reliability of the fault - detection results. Through an efficient analysis method, fault detection can be completed in a relatively short time, improving the real - time performance of the system and promptly discovering and handling faults. This method has strong adaptability and can be applied to multi - dimensional data of different types and complexities, and is applicable to the fault detection of various relay protection loops.

[0123] In an embodiment of the present application, the analysis of the time performance of the characteristic distance corresponding to the multi - dimensional data includes:

[0124] Based on the multi - dimensional data of any two of the current transformers, obtain the current distance, zero - sequence current distance, and resistance distance;

[0125] Take the current distance, zero - sequence current distance, and resistance distance as characteristic distances;

[0126] Based on the numerical value and duration of the characteristic distance, the time performance of the characteristic distance of each current transformer is obtained.

[0127] Among them, the current distance refers to the distance between the current values of two current transformers within the same time window. Common distance measurement methods include multi-dimensional Hausdorff distance, Euclidean distance, etc. By calculating the current distance, the similarity and difference of the current values of two current transformers in time can be evaluated, which is used to identify abnormal changes in current.

[0128] Among them, the zero-sequence current distance refers to the distance between the zero-sequence current values of two current transformers within the same time window. Common distance measurement methods include multi-dimensional Hausdorff distance, Euclidean distance, etc. By calculating the zero-sequence current distance, the similarity and difference of the zero-sequence current values of two current transformers in time can be evaluated, which is used to identify abnormal changes in zero-sequence current.

[0129] Among them, the resistance distance refers to the distance between the resistance values of two current transformers within the same time window. Common distance measurement methods include multi-dimensional Hausdorff distance, Euclidean distance, etc. By calculating the resistance distance, the similarity and difference of the resistance values of two current transformers in time can be evaluated, which is used to identify abnormal changes in resistance.

[0130] Among them, the time performance of the characteristic distance refers to the change trend and pattern of the characteristic distance in time, including the numerical value and duration of the characteristic distance. By analyzing the time performance of the characteristic distance, the fluctuations, trends and durations of the characteristic distance can be identified, so as to infer the possible fault types and the fault development process.

[0131] Exemplarily, some faults are relatively minor, such as poor contact, etc. The fault manifestation at a single moment may not be obvious enough. At this time, the fault size is adjusted, and by statistically analyzing the regularity and persistence in time, the possibility of the fault occurrence can be further adjusted to make the fault occurrence more significant and easier to be detected. Calculate the multi-dimensional Hausdorff distance based on time series of the current, zero-sequence current, and resistance of any two current transformers:

[0132]

[0133]

[0134] Among them, represents the multi-dimensional Hausdorff distance based on time series between the i-th current transformer and the j-th current transformer at the t-th moment; are respectively the current distance, zero-sequence current distance, and resistance distance between the i-th current transformer and the j-th current transformer at the t-th moment; represents the average Euclidean distance function; and respectively represent the currents of the i-th current transformer and the j-th current transformer at the t-th moment; and respectively represent the currents of the j-th current transformer at the (t - 1)-th moment and the (t + 1)-th moment. and The calculation method of is the same as

[0135] Next, the current distance, zero-sequence current distance, and resistance distance are denoted as characteristic distances, and the time performance of the characteristic distances is analyzed. When the characteristic distances are not very significant, if the characteristic distances persist, it can still be considered that a fault exists, and the longer the continuous existence time and the larger the overall proportion of existence, the more prominent the time performance. For example, on a connection line of a primary device, due to line aging, etc., the line is unstable. Due to weather reasons, the line current distance appears frequently within a certain period of time, but it is not completely disconnected. Therefore, the value of the current distance is small. At this time, it should still be determined as an open-circuit fault.

[0136] The representation method of the time performance can be:

[0137]

[0138] where represents the time performance of any characteristic distance of any current transformer; represents the number of monitoring moments when the characteristic distance of the current transformer is not 0; represents the total number of monitoring moments; represents the number of time periods when the characteristic distance is 0; and represent the durations of the (u + 1)-th and u-th time periods when the characteristic distance is 0; is the maximum-minimum normalization function.

[0139] where represents the continuous existence time length. The longer this time length, the more prominent the time performance and the larger the time performance.

[0140] In this embodiment, by calculating the current distance, zero-sequence current distance, and resistance distance, the similarity and difference between different current transformers can be evaluated more accurately, which helps to identify abnormal changes in current, zero-sequence current, and resistance. By analyzing the time performance of the characteristic distances, the fluctuations, trends, and durations of the characteristic distances can be identified, thereby inferring possible fault types and fault development processes. By comprehensively analyzing the numerical magnitudes and durations of the characteristic distances, misjudgments caused by instantaneous noise can be reduced, ensuring the reliability of the fault detection results.

[0141] In an embodiment of the present application, adjusting the characteristic distance corresponding to the multi-dimensional data according to the analysis result of the time performance includes:

[0142] Adjusting the current distance according to the time performance of the current distance to obtain an adjusted current distance;

[0143] Adjusting the zero-sequence current distance according to the time performance of the zero-sequence current distance to obtain an adjusted zero-sequence current distance;

[0144] Adjusting the resistance distance according to the time performance of the resistance distance to obtain an adjusted resistance distance.

[0145] Among them, if the characteristic distance itself is larger, the adjustable part is less, and it can be maintained. If the characteristic distance itself is smaller, the characteristic distance is adjusted according to the time performance. The stronger the time performance, the larger the adjustment amount of the characteristic distance, and the easier it is to be recognized as a fault. Therefore, adjusting the characteristic distance corresponding to the multi-dimensional data according to the analysis result of the time performance, the adjustment method can be:

[0146]

[0147]

[0148] Among them, 、 、 respectively represent the adjusted current distance, adjusted zero-sequence current distance, and adjusted resistance distance of the i-th current transformer and the j-th current transformer at the t-th moment; is the current distance between the i-th current transformer and the j-th current transformer at the t-th moment; represents the time performance of the current distance of any current transformer; represents the adjusted multi-dimensional Hausdorff distance based on the time series between the i-th current transformer and the j-th current transformer at the t-th moment; represents an exponential function, specifically an exponential function with the natural constant e as the base, 、 The calculation methods of are the same as .

[0149] In this embodiment, by dynamically adjusting the feature distance, the similarity and difference of multi-dimensional data can be more accurately evaluated according to the change trend and pattern of the actual data, thereby improving the accuracy of fault detection. Dynamically adjusting the feature distance can reduce misjudgments caused by instantaneous noise and ensure the reliability of the fault detection results. By analyzing the time performance of the current distance, zero-sequence current distance, and resistance distance, the fault type can be identified from multiple perspectives, improving the ability to identify the fault type. Combining the time performance of multiple feature distances can comprehensively evaluate the possibility and severity of the fault, providing more comprehensive fault diagnosis information.

[0150] In an embodiment of the present application, adjusting the initial fault possibility according to the feature distance corresponding to the adjusted multi-dimensional data to obtain the target fault possibility corresponding to the multiple preset fault types includes:

[0151] Obtaining an adjusted multi-dimensional Hausdorff distance according to the adjusted current distance, the adjusted zero-sequence current distance, and the adjusted resistance distance;

[0152] Determining an adjustment index of the feature distance according to the adjusted multi-dimensional Hausdorff distance;

[0153] Adjusting the initial fault possibility according to the adjustment index of the feature distance to obtain the target fault possibility corresponding to the multiple preset fault types.

[0154] Among them, the multi-dimensional Hausdorff distance is a measurement method for measuring the maximum distance between two multi-dimensional data sets. It not only considers the distance between data points but also the overall structure of the data set. By calculating the multi-dimensional Hausdorff distance, the similarity and difference between multi-dimensional data can be more comprehensively evaluated for identifying abnormal changes in the data.

[0155] Among them, the adjustment index refers to a parameter or standard for adjusting the feature distance determined according to the adjusted multi-dimensional Hausdorff distance. By determining the adjustment index, the feature distance can be adjusted more precisely, improving the accuracy and reliability of fault detection.

[0156] Among them, according to the adjustment index, the initial fault possibility is adjusted to obtain the final target fault possibility. By adjusting the initial fault possibility, the probability of different fault types can be more accurately evaluated, improving the accuracy and reliability of fault detection.

[0157] Exemplarily, when the characteristic distance corresponding to the adjusted multi-dimensional data belongs to a regular value range and changes periodically, the characteristic distance corresponding to any adjusted multi-dimensional data is relatively regular, and the adjustment index is small. The more irregular the numerical performance of the characteristic distance in the period, the more adjustment should be made. The regularity of the numerical performance of the characteristic distance can be reflected in the numerical difference between time windows. Through the difference between windows represented by representative values, that is, extreme values, the regularity of the numerical performance of the characteristic distance can be obtained. Add up the differences in the ranges between all time windows. The larger this value is, the more concentrated the range distribution is, so the more irregular the numerical performance of the characteristic distance in the period, and the more adjustment should be made. Thus, according to the change of the adjusted multi-dimensional Hausdorff distance based on the time series, the adjustment index of the characteristic distance. According to the adjustment index of the distance, adjust the initial fault possibility to obtain the target fault possibility corresponding to multiple preset fault types. The representation of the target fault possibility can be:

[0158]

[0159] wherein, represents the target fault possibility corresponding to multiple preset fault types; 、 respectively represent the target fault possibility corresponding to the current fault and the target fault possibility corresponding to the resistance fault; 、 respectively represent the initial fault possibility corresponding to the current fault and the initial fault possibility corresponding to the resistance fault; 、 、 respectively represent the adjustment indexes of the current characteristic distance, the zero-sequence current characteristic distance, and the resistance characteristic distance.

[0160] In this embodiment, by calculating the multi-dimensional Hausdorff distance, the similarity and difference between multi-dimensional data can be evaluated more comprehensively, and the accuracy of fault detection can be improved. By determining the adjustment index, the characteristic distance can be adjusted more precisely, reducing misjudgment and improving the reliability of fault detection. Through the adjusted multi-dimensional Hausdorff distance, fault types can be identified from multiple angles, improving the ability to identify fault types. Combining the time performance of multiple characteristic distances, the possibility and severity of faults can be comprehensively evaluated, providing more comprehensive fault diagnosis information.

[0161] In an embodiment of the present application, the determining the adjustment index of the characteristic distance according to the adjusted multi-dimensional Hausdorff distance includes:

[0162] Obtain the period of the adjusted multi-dimensional Hausdorff distance;

[0163] Obtain a periodic pattern based on the variance of the characteristic distances corresponding to multiple time windows within the period.

[0164] Determine an adjustment index for the characteristic distance according to the periodic pattern, the number of the time windows, and the range of the characteristic distances within each time window.

[0165] Among them, variance is a statistical concept used to measure the dispersion degree of a set of data. The periodic pattern refers to the regularity shown by the variances of the characteristic distances corresponding to multiple time windows within a certain period. By calculating the variances of the characteristic distances corresponding to multiple time windows within the period, the change trend and periodic pattern of the characteristic distances can be identified, so as to better understand the dynamic characteristics of the data.

[0166] Among them, the range is the difference between the maximum value and the minimum value in a set of data. The range of the characteristic distances within a time window refers to the difference between the maximum value and the minimum value of the characteristic distances within a certain time window. By calculating the range of the characteristic distances within the time window, the fluctuation range of the characteristic distances within the time window can be evaluated, which is used to judge the stability of the data.

[0167] Exemplarily, obtain the period of the adjusted multi-dimensional Hausdorff distance: for the characteristic distances corresponding to the adjusted multi-dimensional data, with 3 as the preset time window, calculate the variance of the average characteristic distances of any time window of the characteristic distances corresponding to any adjusted multi-dimensional data for the adjusted multi-dimensional Hausdorff distance based on the time series; increase the time window by a step of 1, with 15 as the termination time window, and calculate the variance of the average characteristic distances of time windows of any length; record the time window corresponding to the minimum value of the variances of the characteristic distances of all time windows of all lengths as the period of the characteristic distances, and the periods of three different characteristic distances constitute the period of the adjusted multi-dimensional Hausdorff distance; record the difference between the maximum value and the minimum value within the time window within the period of the adjusted multi-dimensional Hausdorff distance as the range.

[0168] Exemplarily, the representation method of the adjustment index of the characteristic distance can be:

[0169]

[0170] Among them, 、 、 represent the adjustment indexes of the current characteristic distance, the zero-sequence current characteristic distance, and the resistance characteristic distance respectively; 、 、 respectively represent the variances of the average characteristic distances within the time windows of the adjusted current distance, the adjusted zero-sequence current distance, and the adjusted resistance distance in chronological order, indicating the periodic pattern. If the average characteristic distances of each period are similar, then the variance is relatively small, the period is more regular, and the adjustment index is smaller; 、 、 respectively represent the numbers of time windows with the cycle lengths of the adjusted current distance, the adjusted zero-sequence current distance, and the adjusted resistance distance; 、 respectively represent the maximum and minimum values within the time window of the adjusted current distance in one cycle; 、respectively represent the maximum and minimum values within the time window of the adjusted zero-sequence current distance; 、 respectively represent the maximum and minimum values within the time window of the adjusted resistance distance; is the maximum-minimum normalization function.

[0171] Among them, represents the range within the time window of the adjusted current distance in one cycle. The larger this value is, the larger the adjustment index of the current distance; represents the range within the time window of the adjusted zero-sequence current distance in one cycle. The larger this value is, the larger the adjustment index of the zero-sequence current distance; represents the range within the time window of the adjusted resistance distance in one cycle. The larger this value is, the larger the adjustment index of the resistance distance.

[0172] In this embodiment, by identifying the periodic pattern, the change trend of the characteristic distance can be evaluated more accurately, and misjudgment can be reduced. By calculating the range within the time window, the fluctuation range of the characteristic distance can be evaluated, and the accuracy of fault detection can be improved. By analyzing the periodic pattern and the range, the fault type can be identified from multiple perspectives, and the ability to identify the fault type can be improved. Combining the variances and ranges of multiple time windows can comprehensively evaluate the possibility and severity of the fault, and provide more comprehensive fault diagnosis information.

[0173] In an embodiment of the present application, the fault repair of each of the current transformers according to the target fault category includes:

[0174] Determine the fault loop according to the target fault category;

[0175] Check the states of the lines and equipment connected to the current transformer in the fault loop, and repair the faults of the lines and equipment with abnormal states.

[0176] Among them, obtained from the above steps , the target fault probability corresponding to the current fault Obtained from the current and zero-sequence current, indicating a short-circuit fault. When it indicates a loop short circuit, and when it indicates a loop ground short-circuit fault ( , The value range is 0 - 1, the value range is 0 - 2), and within the value range, the larger the value, the more serious the fault. The target fault probability corresponding to the resistance fault Obtained from the current and resistance, indicating an open-circuit fault. When it indicates a loop open-circuit fault, and when it indicates a short-circuit fault of a loop with a specific number. And within the value range, the larger the value, the more serious the fault.

[0177] Exemplarily, through the previous multi-dimensional data analysis and feature distance adjustment, the target fault categories of each current transformer have been determined. According to the target fault categories, the loops involved in the faults are determined. For example, if the target fault category of a certain current transformer is "loop short circuit", then the loop where the current transformer is located is determined as the fault loop. Check all the connecting lines in the fault loop, including cables, connectors, connection points, etc. Check all the relevant devices in the fault loop, including switches, circuit breakers, relays, etc. If any damage, breakage, or poor contact is found in the lines, repair or replace them. If any fault or abnormality is found in the devices, repair or replace them.

[0178] In this embodiment, by determining the target fault category, the fault loop can be accurately identified, unnecessary inspections and repairs can be avoided, and the accuracy of fault detection can be improved. Through detailed inspections of the lines and devices, ineffective repairs caused by misjudgment can be reduced, ensuring the reliability of the fault detection results. By promptly repairing the lines and devices with abnormal states, the normal operation of the system can be quickly restored, improving the reliability of the system. Through regular inspections and maintenance, potential faults can be prevented, and the service life of the devices can be extended. With clear fault categories and fault loops, maintenance personnel can quickly locate problems, shorten the repair time, and improve the repair efficiency. Through efficient repairs, the downtime of the system can be reduced, ensuring the continuity of production activities. By accurately locating faults and avoiding unnecessary repairs, the maintenance cost can be reduced. Through regular inspections and maintenance, potential faults can be prevented, and the costs of major repairs and equipment replacements can be reduced.

[0179] Figure 3 The structural schematic diagram of a relay protection loop fault detection device provided by an embodiment of the present invention. This device can be applied to Figure 1The implementation environment shown. This device can also be applicable to other exemplary implementation environments and can be specifically configured in other devices. This embodiment does not limit the implementation environment applicable to this device.

[0180] As Figure 3 shown, this exemplary relay protection circuit fault detection device includes:

[0181] A data acquisition module 301, configured to acquire multi-dimensional data of a plurality of current transformers in a target relay protection circuit, where the multi-dimensional data includes current time series data, zero-sequence current time series data, and resistance time series data;

[0182] An initial fault analysis module 302, configured to analyze the change trend of the multi-dimensional data for any one of the current transformers to obtain initial fault probabilities corresponding to a plurality of preset fault types;

[0183] A target fault analysis module 303, configured to analyze the time performance of the characteristic distance corresponding to the multi-dimensional data, adjust the characteristic distance corresponding to the multi-dimensional data according to the analysis result of the time performance, and adjust the initial fault probability according to the adjusted characteristic distance corresponding to the multi-dimensional data to obtain target fault probabilities corresponding to the plurality of preset fault types;

[0184] A fault category determination module 304, configured to determine a target fault category corresponding to each current transformer according to the target fault probability, and perform fault repair on each current transformer according to the target fault category.

[0185] In this exemplary relay protection circuit fault detection device, acquiring current time series data, zero-sequence current time series data, and resistance time series data provides multi-dimensional information, which can more comprehensively reflect the state of the relay protection circuit. Different types of time series data can complement each other, reduce the limitations of a single data source, and improve the reliability of fault detection; by analyzing the change trend of the multi-dimensional data, the trend and pattern of the data can be identified, which helps to detect potential fault signs at an early stage; dynamically adjusting the characteristic distance according to the time performance of the characteristic distance can more accurately reflect the actual fault situation. Static characteristic distances cannot adapt to different fault scenarios, while dynamic adjustment can improve adaptability and accuracy; combining the adjusted characteristic distance and re-evaluating the initial fault probability to obtain a more accurate target fault probability can eliminate misjudgment and improve the accuracy of fault detection; according to the target fault probability, the specific fault category can be accurately located, avoiding blind repair and improving the repair efficiency.

[0186] It should be noted that the relay protection circuit fault detection device provided in the above embodiments and the relay protection circuit fault detection method provided in the above embodiments belong to the same concept. The specific manners in which each module and unit perform operations have been described in detail in the method embodiments, and will not be elaborated here. In practical applications, the relay protection circuit fault detection device provided in the above embodiments can, according to needs, allocate the above functions to different functional modules, that is, divide the internal structure of the device into different functional modules to complete all or part of the functions described above. This is not limited herein either.

[0187] An embodiment of the present application also provides an electronic device, including: one or more processors; a storage device for storing one or more programs, which, when executed by the one or more processors, cause the electronic device to implement the relay protection circuit fault detection method provided in each of the above embodiments.

[0188] It should be noted that the computer-readable medium shown in the embodiments of the present application can be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable computer program. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The computer program contained on a computer-readable medium can be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0189] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. Among them, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the above-mentioned module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, as well as the combination of blocks in the block diagram or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0190] The units involved in the embodiments described in the present application can be implemented in software or in hardware, and the described units can also be provided in a processor. Among them, the names of these units do not necessarily constitute a limitation on the unit itself in some cases.

[0191] Another aspect of the present application also provides a relay protection circuit fault detection system, where the system includes a computer-readable storage medium, and the computer-readable storage medium stores multiple instructions, and the instructions are suitable for being loaded by a processor to execute the steps in any one of the relay protection circuit fault detection methods provided by the embodiments of the present application. This computer-readable storage medium can be included in the electronic device described in the above embodiments, or can exist alone without being assembled into the electronic device.

[0192] It should be noted that: the above-mentioned sequence of the embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or consecutive order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0193] Each embodiment in this specification is described in a progressive manner, and the same or similar parts among the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A method for detecting faults in a relay protection circuit, characterized in that: The method comprises: Acquire multi-dimensional data of multiple current transformers in a target relay protection circuit, wherein the multi-dimensional data includes current time series data, zero-sequence current time series data, and resistance time series data; For any of the current transformers, analyzing the change trend of the multi-dimensional data to obtain initial fault probabilities corresponding to a plurality of preset fault types; Analyzing the time performance of the characteristic distance corresponding to the multidimensional data, adjusting the characteristic distance corresponding to the multidimensional data according to the analysis result of the time performance, and adjusting the initial fault probability according to the adjusted characteristic distance corresponding to the multidimensional data to obtain the target fault probability corresponding to the multiple preset fault types; Determine a target fault category corresponding to each of the current transformers according to the target fault possibility, and perform fault repair on each of the current transformers according to the target fault category; The analyzing the time performance of the characteristic distance corresponding to the multi-dimensional data includes: According to the multidimensional data of any two current transformers, a current distance, a zero-sequence current distance and a resistance distance are obtained, wherein the current distance refers to the distance between the current values ​​of the two current transformers in the same time window, the zero-sequence current distance refers to the distance between the zero-sequence current values ​​of the two current transformers in the same time window, and the resistance distance refers to the distance between the resistance values ​​of the two current transformers in the same time window; Taking the current distance, zero-sequence current distance and resistance distance as characteristic distances; According to the numerical value and duration of the characteristic distance, a time representation of the characteristic distance of each current transformer is obtained; The step of adjusting the characteristic distance corresponding to the multidimensional data according to the analysis result of the time performance includes: adjusting the current distance according to the time performance of the current distance to obtain an adjusted current distance; According to the time performance of the zero-sequence current distance, the zero-sequence current distance is adjusted to obtain an adjusted zero-sequence current distance; adjusting the resistance distance according to the time performance of the resistance distance to obtain an adjusted resistance distance; The time expression is as follows: Where P represents the time performance of any characteristic distance of any current transformer; T V represents the number of monitoring moments when the characteristic distance of the current transformer is not 0; T represents the total number of monitoring moments; U represents the number of time periods when the characteristic distance is 0; H u+1 , H u Indicates the duration of the time period where the u+1th and uth feature distances are 0; norm is the maximum and minimum value normalization function; The characteristic distance corresponding to the multi-dimensional data is adjusted according to the analysis result of the time performance, and the adjustment method may be: TMHD ′ (U i,t ,U j,t )={A ′ i,j,t ,B i ′ ,j,t ,C i ′ ,j,t }; A ′ i,j,t =A i,j,t ×[1+exp(-A i,j,t )×P A ]; A ′ i,j,t , B i ′ ,j,t , C i ′ ,j,t They represent the adjusted current distance, adjusted zero-sequence current distance, and adjusted resistance distance of the i-th current transformer and the j-th current transformer at the t-th moment, respectively; A i,j,t is the current distance between the i-th current transformer and the j-th current transformer at the t-th moment; P A Represents the time representation of the current distance of any current transformer; TMHD ′ (U i,t ,U j,t ) represents the time series-based regulated multidimensional Hausdorff distance between the i-th current transformer and the j-th current transformer at the t-th time; exp represents an exponential function, specifically an exponential function with the natural constant e as the base, B i ′ ,j,t , C i ′ ,j,t The calculation method is the same as A ′ i,j,t .

2. The relay protection circuit fault detection method according to claim 1, characterized in that: After the multi-dimensional data of multiple current transformers in the target relay protection circuit are obtained, the method further includes: According to a preset wavelet basis function and a preset number of decomposition layers, a discrete wavelet transform is performed on the multidimensional data to obtain an initial approximation coefficient and an initial detail coefficient; According to a preset soft threshold function, the initial approximation coefficient and the initial detail coefficient are processed to obtain a target approximation function and a target detail function; The target approximation function and the target detail function are subjected to inverse discrete wavelet transform to obtain the denoised multidimensional data, and the denoised multidimensional data is used for change trend analysis.

3. The relay protection circuit fault detection method according to claim 1, characterized in that: The multiple preset fault types include loop short circuit, loop ground short circuit, loop open circuit and specific number loop short circuit, and the analysis of the change trend of the multi-dimensional data to obtain the initial fault probability corresponding to the multiple preset fault types includes: According to the preset neighborhood monitoring time, a time window is obtained; Analyze the change trend of the current time series data, and obtain the initial fault possibility corresponding to the loop short circuit according to the average current in the time window; Analyze the change trends of the current time series data and the zero-sequence current time series data, and obtain the initial fault possibility of the loop ground short circuit according to the average current and the zero-sequence current in the time window; Analyze the change trend of the resistance time series data, and obtain the initial fault possibility corresponding to the circuit break according to the average resistance in the time window; The change trend of the resistance time series data and the phase change of the current time series data are analyzed, and the initial fault possibility corresponding to the short circuit of the specific numbered loop is obtained according to the average resistance and phase change in the time window.

4. The relay protection circuit fault detection method according to claim 1, characterized in that: The adjusting the initial fault probability according to the adjusted characteristic distance corresponding to the multi-dimensional data to obtain the target fault probability corresponding to the multiple preset fault types includes: Obtaining an adjusted multi-dimensional Hausdorff distance according to the adjusted current distance, the adjusted zero-sequence current distance, and the adjusted resistance distance; Determining an adjustment index of the feature distance according to the adjusted multi-dimensional Hausdorff distance; The initial fault probability is adjusted according to the adjustment index of the characteristic distance to obtain the target fault probability corresponding to the multiple preset fault types.

5. The relay protection circuit fault detection method according to claim 4, characterized in that: Determining the adjustment index of the feature distance according to the adjusted multi-dimensional Hausdorff distance includes: Obtaining the period of the adjusted multi-dimensional Hausdorff distance; Obtaining a periodic law according to the variance of the characteristic distances corresponding to the multiple time windows within the period; An adjustment index of the characteristic distance is determined according to the periodic law, the number of the time windows, and the range of the characteristic distance in each time window.

6. The relay protection circuit fault detection method according to claim 1, characterized in that: The performing fault repair on each current transformer according to the target fault category includes: Determining a fault circuit according to the target fault category; Check the status of the lines and devices connected to the current transformer in the fault loop, and perform fault repair on the lines and devices with abnormal status.

7. A relay protection circuit fault detection device, characterized in that: The device comprises: A data acquisition module, used to acquire multi-dimensional data of multiple current transformers in a target relay protection circuit, wherein the multi-dimensional data includes current time series data, zero-sequence current time series data and resistance time series data; An initial fault analysis module, for analyzing the change trend of the multi-dimensional data for any current transformer, and obtaining initial fault possibilities corresponding to a plurality of preset fault types; a target fault analysis module, configured to analyze the time performance of the characteristic distance corresponding to the multidimensional data, adjust the characteristic distance corresponding to the multidimensional data according to the analysis result of the time performance, and adjust the initial fault possibility according to the adjusted characteristic distance corresponding to the multidimensional data, so as to obtain the target fault possibility corresponding to the multiple preset fault types; The analyzing the time performance of the characteristic distance corresponding to the multi-dimensional data includes: According to the multidimensional data of any two current transformers, a current distance, a zero-sequence current distance and a resistance distance are obtained, wherein the current distance refers to the distance between the current values ​​of the two current transformers in the same time window, the zero-sequence current distance refers to the distance between the zero-sequence current values ​​of the two current transformers in the same time window, and the resistance distance refers to the distance between the resistance values ​​of the two current transformers in the same time window; Taking the current distance, zero-sequence current distance and resistance distance as characteristic distances; According to the numerical value and duration of the characteristic distance, a time representation of the characteristic distance of each current transformer is obtained; The step of adjusting the characteristic distance corresponding to the multidimensional data according to the analysis result of the time performance includes: adjusting the current distance according to the time performance of the current distance to obtain an adjusted current distance; According to the time performance of the zero-sequence current distance, the zero-sequence current distance is adjusted to obtain an adjusted zero-sequence current distance; adjusting the resistance distance according to the time performance of the resistance distance to obtain an adjusted resistance distance; The time expression is as follows: Where P represents the time performance of any characteristic distance of any current transformer; T V represents the number of monitoring moments when the characteristic distance of the current transformer is not 0; T represents the total number of monitoring moments; U represents the number of time periods when the characteristic distance is 0; H u+1 , H u Indicates the duration of the time period where the u+1th and uth feature distances are 0; norm is the maximum and minimum value normalization function; The characteristic distance corresponding to the multi-dimensional data is adjusted according to the analysis result of the time performance, and the adjustment method may be: TMHD ′ (U i,t ,U j,t )={A ′ i,j,t ,B i ′ ,j,t ,C i ′ ,j,t }; A ′ i,j,t =A i,j,t ×[1+exp(-A i,j,t )×P A ]; A ′ i,j,t , B i ′ ,j,t , C i ′ ,j,t They represent the adjusted current distance, adjusted zero-sequence current distance, and adjusted resistance distance of the i-th current transformer and the j-th current transformer at the t-th moment, respectively; A i,j,t is the current distance between the i-th current transformer and the j-th current transformer at the t-th moment; P A Represents the time representation of the current distance of any current transformer; TMHD ′ (U i,t ,U j,t ) represents the time series-based regulated multidimensional Hausdorff distance between the i-th current transformer and the j-th current transformer at the t-th time; exp represents an exponential function, specifically an exponential function with the natural constant e as the base, B i ′ ,j,t , C i ′ ,j,t The calculation method is the same as A ′ i,j,t ; The fault category determination module is used to determine the target fault category corresponding to each of the current transformers according to the target fault possibility, and perform fault repair on each of the current transformers according to the target fault category.

8. A relay protection circuit fault detection system, characterized in that: The system includes a computer-readable storage medium storing a plurality of instructions, wherein the instructions are suitable for being loaded by a processor to execute the steps in the relay protection circuit fault detection method according to any one of claims 1 to 6.

Images