Real-time data management method, device, equipment and storage medium for pole-mounted circuit breaker

Abstract

The present application provides a real-time data management method, apparatus, device, and storage medium for a pole-mounted circuit breaker, which are applied to edge devices. The method includes: processing electrical data of the pole-mounted circuit breaker to obtain an electrical feature vector, and processing switch data of the pole-mounted circuit breaker to obtain a switch feature vector; inputting the electrical feature vector and the switch feature vector into a pre-trained fault analysis model, outputting a fault time period, a fault cause within the fault time period, and the credibility of the fault cause; setting a credibility lower than a preset credibility threshold as a target credibility, and determining the fault cause and target fault time period corresponding to the target credibility; extracting data from recorded data according to the fault time period to obtain target recorded data, generating data to be analyzed based on the target recorded data, target fault cause, and target credibility, and encrypting the data to be analyzed and uploading it to a cloud server.

Description

Technical Field

[0001] The present application relates to the field of edge computing technology, and in particular to a real-time data management method, device, equipment and storage medium for a pole-mounted circuit breaker. Background Art

[0002] Currently, pole-mounted circuit breakers are used to ensure the stable operation of power grids and the reliable distribution of power. As key protective devices in the power grid, pole-mounted circuit breakers can rapidly interrupt fault currents when high-voltage line faults (such as short circuits and overloads) occur, limiting the fault's scope and preventing its spread. This protects other equipment in the grid from damage, ultimately maintaining overall grid stability and security. With the advancement of computer technology, pole-mounted circuit breakers are now equipped with devices for intelligent management and control, such as edge devices. Based on data collected from the pole-mounted circuit breaker and the pre-set data control range, edge devices determine whether the high-voltage line associated with the pole-mounted circuit breaker is abnormal and issue corresponding control instructions to the pole-mounted circuit breaker. However, the computing power resources of edge devices are limited. For complex faults, data must be uploaded to cloud servers for auxiliary analysis. When reporting regional power faults, if edge devices simultaneously upload data to cloud servers, the large amount of data will congest the data transmission channel, reducing the efficiency of fault analysis. Summary of the Invention

[0003] The present application provides a real-time data management method, apparatus, device, and storage medium for a pole-mounted circuit breaker, which are used to analyze the real-time data of the pole-mounted circuit breaker to obtain the cause of the fault, and upload the fault cause with low credibility and the corresponding recorded data to a cloud server for auxiliary analysis, so as to improve the analysis efficiency of complex faults and thus reduce the time of fault handling.

[0004] In a first aspect, an embodiment of the present application provides a real-time data management method for a pole-mounted circuit breaker, applied to an edge device, the method comprising:

[0005] Acquire recorded wave data of the pole-mounted circuit breaker, classify and process the recorded wave data to obtain electrical data and switch data;

[0006] Perform feature extraction based on the electrical data to obtain an electrical feature set;

[0007] Performing vector processing on the electrical feature set to obtain an electrical feature vector, and performing vector processing on the switch data to obtain a switch feature vector;

[0008] Inputting the electrical characteristic vector and the switch characteristic vector into a pre-trained fault analysis model, and outputting a fault time period, a fault cause within the fault time period, and the credibility of the fault cause;

[0009] The credibility lower than the preset credibility threshold is set as the target credibility, the fault cause corresponding to the target credibility is set as the target fault cause, and the fault time period corresponding to the target fault cause is set as the target fault time period;

[0010] Extract data from the recorded data according to the fault time period to obtain target recorded data, generate data to be analyzed according to the target recorded data, the target fault cause and the target credibility, and encrypt the data to be analyzed and upload it to the cloud server.

[0011] In a second aspect, an embodiment of the present application provides a real-time data management device for a pole-mounted circuit breaker, the real-time data management device comprising:

[0012] A data processing module is used to obtain the recorded wave data of the pole-mounted circuit breaker, classify and process the recorded wave data, and obtain electrical data and switch data;

[0013] A feature extraction module, configured to extract features based on the electrical data to obtain an electrical feature set;

[0014] a vector conversion module, configured to perform vector processing on the electrical feature set to obtain an electrical feature vector, and to perform vector processing on the switch data to obtain a switch feature vector;

[0015] a fault analysis module, configured to input the electrical characteristic vector and the switch characteristic vector into a pre-trained fault analysis model, and output a fault time period, a fault cause within the fault time period, and a credibility of the fault cause;

[0016] a fault classification module, configured to set a credibility lower than a preset credibility threshold as a target credibility, set a fault cause corresponding to the target credibility as a target fault cause, and set a fault time period corresponding to the target fault cause as a target fault time period;

[0017] The data uploading module is used to extract data from the recorded data according to the fault time period to obtain target recorded data, generate data to be analyzed according to the target recorded data, the target fault cause and the target credibility, and encrypt the data to be analyzed and upload it to the cloud server.

[0018] In a third aspect, an embodiment of the present application provides an electronic device, the electronic device including a memory and a processor;

[0019] The memory is used to store computer programs;

[0020] The processor is configured to execute the computer program and implement the real-time data management method for a pole-mounted circuit breaker as described in any one of the embodiments of the present application when executing the computer program.

[0021] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements a real-time data management method for a pole-mounted circuit breaker as described in any one of the embodiments of the present application.

[0022] An embodiment of the present application provides a real-time data management method for a pole-mounted circuit breaker, which is applied to an edge device. The method includes: obtaining recorded waveform data of the pole-mounted circuit breaker, classifying and processing the recorded waveform data to obtain electrical data and switch data; performing feature extraction based on the electrical data to obtain an electrical feature set; performing vectorization processing on the electrical feature set to obtain an electrical feature vector, and performing vectorization processing on the switch data to obtain a switch feature vector; inputting the electrical feature vector and the switch feature vector into a pre-trained fault analysis model, and outputting a fault time period, a fault cause within the fault time period, and a credibility of the fault cause; setting a credibility lower than a preset credibility threshold as a target credibility, setting the fault cause corresponding to the target credibility as a target fault cause, and setting the fault time period corresponding to the target fault cause as a target fault time period; extracting data from the recorded waveform data based on the fault time period to obtain target recorded waveform data, generating data to be analyzed based on the target recorded waveform data, the target fault cause, and the target credibility, and encrypting the data to be analyzed and uploading it to a cloud server. Through the above method, electrical data is divided into electrical data reflecting the system's operating status and switch data reflecting the mechanism and time of fault occurrence. By extracting and vectorizing the electrical data, complex electrical signals are converted into a machine-readable form, constructing an electrical feature vector. Similarly, the switch data is vectorized to form a switch feature vector. This conversion not only simplifies the data structure but also facilitates subsequent model analysis. Using a pre-trained fault analysis model, the electrical and switch feature vectors are analyzed to determine the fault time period, fault cause, and its credibility. High-confidence fault causes can be processed by the edge device, while low-confidence fault causes are packaged and sent along with the corresponding fault data to the cloud server. Sending only a small amount of high-value fault data prevents data congestion, enabling the cloud server to quickly assist in fault cause analysis and expedite fault resolution, thereby minimizing the damage caused by power failures. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0024] Figure 1 A schematic flow chart of a real-time data management method for a pole-mounted circuit breaker provided in an embodiment of the present application;

[0025] Figure 2 A schematic block diagram of a real-time data management device for a pole-mounted circuit breaker provided in an embodiment of the present application. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0027] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined, or partially merged, so the actual execution order may vary depending on the actual situation.

[0028] It should also be understood that the terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit the present application. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0029] It should be further understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0030] See also Figure 1 , Figure 1 A schematic flow chart showing a real-time data management method for a pole-mounted circuit breaker provided in an embodiment of the present application is shown. Figure 1 The real-time data management method for a pole-mounted circuit breaker shown is applied to an edge device, and the specific steps of the method include: S101-S106.

[0031] S101. Acquire recorded wave data of a pole-mounted circuit breaker, classify and process the recorded wave data, and obtain electrical data and switch data.

[0032] For example, recorded data in a high-voltage transmission circuit system refers to a series of data automatically recorded by a fault recorder when an abnormality or fault occurs in the power system, reflecting the changes in the system's electrical quantities. Recorded data typically includes changes in electrical parameters (such as voltage, current, and power), as well as important switching data, such as the switch status (closed or open) of a pole-mounted circuit breaker. Recorded data is extremely important for analyzing the occurrence and development of faults in high-voltage transmission circuit systems, determining the operating behavior of protective devices, and verifying the correctness of system operation. A fault recorder is a specially designed device that can quickly activate and accurately record changes in electrical parameters for a period of time before and after a major disturbance such as a short circuit, oscillation, frequency collapse, or voltage collapse in a high-voltage transmission circuit system.

[0033] The edge device is connected to the fault recorder to obtain the recorded data collected by the fault recorder, analyze the faults in the high-voltage transmission circuit system based on the recorded data, and control the upload strategy of the filtered data.

[0034] S101. Extract features based on electrical data to obtain an electrical feature set.

[0035] For example, electrical data includes multiple electrical quantities, such as voltage, current, and power. When processing electrical data, it is necessary to perform feature extraction on each electrical quantity to obtain a corresponding electrical feature set.

[0036] S101 , performing vector processing on an electrical feature set to obtain an electrical feature vector, and performing vector processing on switch data to obtain a switch feature vector.

[0037] For example, before vectorization, the relevant data needs to be normalized so that all features are on the same scale for easier model processing. In the electrical feature set, the features to be extracted for each electrical quantity include three dimensions: frequency variation, amplitude fault, and energy distortion. These three dimensions are arranged in a specific order to form a vector, which is the electrical feature vector.

[0038] Switching data typically involves information such as the operating status of a pole-mounted circuit breaker (e.g., closed or open) and the time of operation. This event information must first be encoded into a machine-readable format, such as defining "closed" as 1 and "opened" as 0, or using a discretized time series representation. Because switching operations are time series, the sequence of switch states within a specific time window can be directly converted into a vector.

[0039] Both electrical and switch feature vectors, after vectorization, can be directly fed into a pre-trained fault analysis model as input. The fault analysis model will perform fault identification, fault cause analysis, and credibility assessment based on these vectors, thereby guiding subsequent operation and maintenance decisions.

[0040] S101 , inputting the electrical characteristic vector and the switch characteristic vector into a pre-trained fault analysis model, and outputting the fault time period, the fault cause within the fault time period, and the credibility of the fault cause.

[0041] For example, the fault analysis model analyzes the input electrical and switching feature vectors to identify time intervals where abnormal patterns in the data are concentrated. These typically correspond to periods when high-voltage transmission circuit systems experience faults. The fault analysis model may determine the fault time period based on pattern changes, energy spikes, or the presence of specific fault signatures within the feature vectors. After determining the fault time period, the fault analysis model further analyzes the electrical and switching feature vectors, matching them with the characteristic patterns of different fault types learned during training to identify the root cause of the fault. In addition to providing the fault cause, the fault analysis model also outputs the credibility of each fault cause, a probability value that reflects the fault analysis model's confidence in its predicted fault cause. This credibility assessment is based on factors such as the degree to which the feature vector matches various fault patterns and the frequency of similar cases in historical data. It helps operations and maintenance personnel determine the reliability of the analysis results and guides subsequent decision-making.

[0042] S101 : Set a credibility lower than a preset credibility threshold as a target credibility, set a fault cause corresponding to the target credibility as a target fault cause, and set a fault time period corresponding to the target fault cause as a target fault time period.

[0043] For example, edge devices have limited computing resources. When faced with complex and difficult-to-analyze faults, the fault analysis model running on the edge device needs to report to the cloud server. Leveraging the cloud server's computing resources and data coordination capabilities, the cloud server can provide a more comprehensive and accurate fault diagnosis for complex faults. The edge device needs to upload the fault cause and its corresponding waveform data to the cloud server. If the edge device uploads all waveform data to the cloud server without identifying it, in the event of a regional power failure, multiple edge devices will upload waveform data simultaneously, inevitably leading to data congestion. Therefore, the edge device needs to directly process the fault cause with higher diagnostic confidence and upload the waveform data generated during the time period corresponding to the fault cause with lower confidence to the cloud server. This allows for the precise separation of high-value fault waveform data, greatly reducing the amount of data transmission, avoiding data congestion, and improving the processing efficiency of the cloud server. This reduces the time required for fault analysis and prevents the spread of faults from causing greater damage to the high-voltage transmission circuit system.

[0044] S101. Extract data from the recorded data according to the fault time period to obtain target recorded data, generate data to be analyzed according to the target recorded data, target fault cause and target credibility, and encrypt the data to be analyzed and upload it to the cloud server.

[0045] For example, after determining the fault time period corresponding to the complex fault, target recorded data is extracted from the recorded data based on the fault time period and packaged together with the target fault cause and target credibility to generate the data to be analyzed. The data generated by the high-voltage transmission circuit system is sensitive data and typically requires encryption before transmission to prevent data leakage. Therefore, in this embodiment of the application, the data to be analyzed needs to be encrypted before being uploaded to the cloud server.

[0046] An embodiment of the present application provides a real-time data management method for a pole-mounted circuit breaker, which is applied to an edge device. The method includes: obtaining recorded waveform data of the pole-mounted circuit breaker, classifying and processing the recorded waveform data to obtain electrical data and switch data; performing feature extraction based on the electrical data to obtain an electrical feature set; performing vectorization processing on the electrical feature set to obtain an electrical feature vector, and performing vectorization processing on the switch data to obtain a switch feature vector; inputting the electrical feature vector and the switch feature vector into a pre-trained fault analysis model, and outputting a fault time period, a fault cause within the fault time period, and a credibility of the fault cause; setting a credibility lower than a preset credibility threshold as a target credibility, setting the fault cause corresponding to the target credibility as a target fault cause, and setting the fault time period corresponding to the target fault cause as a target fault time period; extracting data from the recorded waveform data based on the fault time period to obtain target recorded waveform data, generating data to be analyzed based on the target recorded waveform data, the target fault cause, and the target credibility, and encrypting the data to be analyzed and uploading it to a cloud server. Through the above method, electrical data is divided into electrical data reflecting the system's operating status and switch data reflecting the mechanism and time of fault occurrence. By extracting and vectorizing the electrical data, complex electrical signals are converted into a machine-readable form, constructing an electrical feature vector. Similarly, the switch data is vectorized to form a switch feature vector. This conversion not only simplifies the data structure but also facilitates subsequent model analysis. Using a pre-trained fault analysis model, the electrical and switch feature vectors are analyzed to determine the fault time period, fault cause, and its credibility. High-confidence fault causes can be processed by the edge device, while low-confidence fault causes are packaged and sent along with the corresponding fault data to the cloud server. Sending only a small amount of high-value fault data prevents data congestion, enabling the cloud server to quickly assist in fault cause analysis and expedite fault resolution, thereby minimizing the damage caused by power failures.

[0047] In order to more clearly introduce the technical solution of the present application, the technical solution of the present application will be introduced through specific embodiments below. It should be noted that the specific embodiments are used to expand the technical solution of the present application, but are not intended to limit the present application.

[0048] In some embodiments, when implementing feature extraction based on electrical data to obtain an electrical feature set, the following steps are specifically implemented: based on the local maximum and local minimum of the electrical data; determining a first extreme line based on the local maximum, and determining a second extreme line based on the local minimum; determining an average extreme line based on the first extreme line and the second extreme line; determining multiple eigenmode function terms and a residual term based on the electrical data and the average extreme line; performing feature extraction based on the eigenmode function terms to obtain frequency change characteristics, amplitude fault characteristics, and energy distortion characteristics.

[0049] For example, the Hilbert-Huang Transform (HHT) is used to extract features from electrical data. By identifying local maxima and local minima in the electrical data, a first extreme line and a second extreme line are determined, respectively. These extreme lines can effectively reflect the fluctuation trend of the signal. The determination of the first and second extreme lines depends on the characteristics of the signal itself, does not require preset parameters or models, and has strong adaptability. For non-stationary and nonlinear signals, this method can effectively capture important information in the signal.

[0050] In some embodiments, feature extraction is performed based on the intrinsic mode function terms to obtain frequency variation features, amplitude fault features, and energy distortion features, including: performing a Hilbert transform on each intrinsic mode function term to obtain an analytical signal sequence, the analytical signal sequence including an amplitude coefficient and a phase coefficient, and differentiating the phase coefficient to obtain an instantaneous frequency sequence. The mean and standard deviation of the instantaneous frequency sequence are calculated, and the quotient of the mean of the instantaneous frequency sequence and the standard deviation of the instantaneous frequency sequence are calculated to obtain the frequency variation features of each intrinsic mode function term. Based on the amplitude coefficient, the instantaneous amplitude sequence is determined, the mean and standard deviation of the instantaneous amplitude sequence are calculated, and the quotient of the mean of the instantaneous amplitude sequence and the standard deviation of the instantaneous amplitude sequence are calculated to obtain the amplitude fault features of each intrinsic mode function term. The intrinsic mode function terms are divided into normal terms and fault terms according to preset fault marking points; the amplitude coefficients corresponding to the normal terms are obtained to obtain a first instantaneous amplitude sequence, and the amplitude coefficients corresponding to the fault terms are obtained to obtain a second instantaneous amplitude sequence; the normal energy sum is determined according to the normal terms, the normal energy mean of the normal terms is determined according to the normal energy sum, and the fault energy value of the fault term is calculated; the energy distortion characteristics of the fault term are determined according to the fault energy value and the normal energy mean.

[0051] Exemplarily, the preset fault point can be determined based on the switch data, for example, based on the state of the gate of the pole-mounted circuit breaker. Frequency change characteristics: can reflect the high-frequency faults or harmonic interference that may exist in the electrical system, and help to identify frequency anomalies. Amplitude fault characteristics: can reflect sudden events or faults in the electrical system, and help to detect instantaneous faults in electrical equipment. Energy distortion characteristics: can reflect changes in signal energy distribution, and help to identify long-term or cumulative faults. When extracting features from electrical data, it is necessary to perform feature extraction on electrical quantities such as voltage and current separately. In the embodiment of the present application, the main electrical quantities analyzed are voltage and current.

[0052] In some embodiments, when inputting the electrical feature vector and the switch feature vector into a pre-trained fault analysis model and outputting the fault time period, the fault cause within the fault time period, and the credibility of the fault cause, the following steps are specifically implemented: inputting the electrical feature vector and the switch feature vector into the pre-trained fault analysis model, wherein the fault analysis model includes a circuit fault detection network and a decision network, the circuit fault detection network includes a first long short-term memory network, a second long short-term memory network, and a threshold recurrent network, and the decision network includes a fully connected layer and a Sigmoid function; performing hidden state feature extraction on the electrical feature vector using the first long short-term memory network in the circuit fault detection network to obtain a first hidden state feature vector; performing hidden state feature extraction on the switch feature vector using the second long short-term memory network in each circuit fault detection network to obtain a second hidden state feature vector; inputting the first hidden state feature vector and the second hidden state feature vector into the threshold recurrent network for feature integration to generate a fault feature vector; inputting the fault feature vector into the decision network, and performing fault cause analysis on the fault feature vector using the fully connected layer and the Sigmoid function in the decision network to obtain the fault cause and the credibility of the fault cause.

[0053] Specifically, the electrical feature vector and the switch feature vector are input into a pre-set fault analysis model. The fault analysis model is a complex neural network system consisting of a circuit fault detection network and a decision network. These networks are responsible for processing and analyzing the power system's electrical characteristics and the switching characteristics of the pole-mounted circuit breaker to detect the cause of the fault. The circuit fault detection network comprises a first long short-term memory (LSTM) network, a second LSTM network, and a gated recurrent network. The goal of these networks is to extract hidden state features from the feature vector to capture the pattern of the circuit fault. Next, the first LSTM network performs hidden state feature extraction on the electrical feature vector to obtain a first hidden state feature vector of the electrical feature vector. This hidden state feature vector contains important information about electrical quantity changes. Similarly, the second LSTM network performs hidden state feature extraction on the switch feature vector to obtain a second hidden state feature vector of the switch feature vector. This step more comprehensively captures relevant information about the switch abnormality. The first and second hidden state feature vectors are then input into a gated recurrent network for feature integration. The gated recurrent network combines these two hidden state feature vectors to generate a fault feature vector for the circuit associated with the pole-mounted circuit breaker. The fault feature vector is then input into the decision network, where it is analyzed for the cause of the fault using the fully connected layer and sigmoid function. The sigmoid function outputs the credibility of the fault cause to determine whether an abnormality exists. Next, based on the fault cause, fault detection results are generated for the circuit associated with the pole-mounted circuit breaker. These results indicate whether a circuit fault exists in the circuit associated with the pole-mounted circuit breaker. Based on the fault detection results, a target fault handling strategy is matched from multiple pre-set fault handling strategies. These strategies can include closing and opening the pole-mounted circuit breaker gate to isolate the circuit fault and address the circuit fault. For example, if a current overload abnormality is detected in a branch of a pole-mounted circuit breaker, the system can take appropriate measures based on the pre-set strategy, such as opening the gate on the pole-mounted circuit breaker corresponding to that branch, to prevent the fault from spreading and protect the normal operation of the rest of the power system.

[0054] In some embodiments, after the electrical feature vector and the switch feature vector are input into a pre-trained fault analysis model and the fault time period, the fault cause within the fault time period and the credibility of the fault cause are output, the following steps are also included: according to the fault cause whose credibility is higher than the preset credibility threshold, the corresponding target fault handling strategy is matched from multiple preset fault handling strategies.

[0055] In some embodiments, the data to be analyzed is encrypted and uploaded to a cloud server, including: obtaining the data to be analyzed from the pole-mounted circuit breaker, framing the data to be analyzed to obtain multiple groups of sampling data; performing Fourier transform on the sampling data to generate spectrum data; determining first target data and second target data from the spectrum data according to a preset first spectrum interval and a second spectrum interval, performing a scrambling operation on the first target data and the second target data according to a preset first key to obtain first scrambled data; performing an inverse Fourier transform on the first scrambled data to generate time domain data; performing a grouping operation on the time domain data to obtain multiple groups of frame data, performing a scrambling operation on each group of frame data to obtain second scrambled data; cross-reorganizing the second scrambled data according to a preset second key to obtain encrypted data, and sending the encrypted data to the server.

[0056] The efficiency of data processing is improved through frame processing and grouping operations, which is suitable for real-time data analysis and processing and can quickly respond to and process circuit breaker fault data.

[0057] Through a series of complex scrambling and transformation operations, this solution effectively encrypts the data to be analyzed from pole-mounted circuit breakers, ensuring data security and integrity, enhancing data resistance to attacks, and preventing unauthorized access and cracking. Fourier transform and inverse transform operations ensure the integrity and accuracy of data when converting between the frequency and time domains, preserving the key characteristics of the original data. Multiple scrambling and key control operations enhance the randomness and attack resistance of the data, making it suitable for the high-security requirements of real-time data in power grids. This solution excels in improving data processing efficiency and ensuring data transmission security, effectively ensuring the secure transmission and analysis of pole-mounted circuit breaker fault data.

[0058] In some embodiments, a scrambling operation is performed on first target data and second target data according to a preset first key to obtain first scrambled data, including: inputting the first key into a preset quantum number generator to generate a first random number and a second random number, the number of characters of the first random number is the same as the number of characters of the first target data, and the number of characters of the second random number is the same as the number of characters of the second target data; performing an offset operation on the first target data according to the first random number to obtain first offset data, and performing an offset operation on the second target data according to the second random number to obtain second offset data; performing an exchange operation on the first offset data and the second offset data according to the first key and a preset exchange function to obtain first exchange data and second exchange data; generating a set of complex weight coefficients according to the first key, performing weighted processing on the first exchange data according to the complex weight coefficients to obtain first scrambled data, and performing weighted processing on the second exchange data according to the complex weight coefficients to obtain second scrambled data.

[0059] Specifically, the random numbers generated by the quantum number generator are extremely random and unpredictable, ensuring that the first and second random numbers generated based on the first key are unpredictable. Only the decryption party with the same quantum number generator can obtain the same first and second random numbers. When offsetting according to the first and second random numbers, each data point of the first target data can be offset by the numerical value of the character corresponding to the first random number, and the same applies to the second target data. The introduction of complex weight coefficients provides data protection in two dimensions: amplitude and phase, enhancing the data's resistance to attacks. By generating highly random random numbers through a quantum number generator and combining them with keys for offset, exchange, and weighting, the security and resistance to attacks of the data are significantly improved. The multi-level scrambling operation and the introduction of complex weight coefficients provide multi-dimensional protection for the data, ensuring the complexity and unpredictability of the data. The scheme has high flexibility and adaptability, is suitable for a variety of data encryption needs, and can effectively protect the data security of key equipment such as pole-mounted circuit breakers.

[0060] See also Figure 2 , Figure 2 The present application also provides a schematic block diagram of a real-time data management device for a pole-mounted circuit breaker. The real-time data management device 300 for a pole-mounted circuit breaker is configured to execute the aforementioned real-time data management method for a pole-mounted circuit breaker. The real-time data management device 300 for a pole-mounted circuit breaker can be configured in a server.

[0061] Among them, the server can be an independent server, a server cluster, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0062] like Figure 2 As shown, the real-time data management device 300 for a pole-mounted circuit breaker includes: a data processing module 301 , a feature extraction module 302 , a vector conversion module 303 , a fault analysis module 304 , a fault classification module 305 and a data upload module 306 .

[0063] The data processing module 301 is used to obtain the recorded wave data of the pole-mounted circuit breaker, classify and process the recorded wave data, and obtain electrical data and switch data.

[0064] The feature extraction module 302 is used to extract features based on the electrical data to obtain an electrical feature set.

[0065] In some embodiments, when the feature extraction module 302 is used to implement feature extraction based on electrical data to obtain an electrical feature set, it is specifically used to implement: based on the local maximum and local minimum of the electrical data; determining the first extreme line based on the local maximum, and determining the second extreme line based on the local minimum; determining the average extreme line based on the first extreme line and the second extreme line; determining multiple eigenmode function terms and a residual term based on the electrical data and the average extreme line; performing feature extraction based on the eigenmode function terms to obtain frequency change characteristics, amplitude fault characteristics and energy distortion characteristics.

[0066] In some embodiments, the feature extraction module 302 is used to implement feature extraction based on the intrinsic mode function term to obtain frequency change characteristics, amplitude fault characteristics and energy distortion characteristics, and is specifically used to implement: performing Hilbert transform on each intrinsic mode function term to obtain an analytical signal sequence, the analytical signal sequence includes: amplitude coefficient and phase coefficient; taking the derivative of the phase coefficient to obtain an instantaneous frequency sequence; calculating the mean and standard deviation of the instantaneous frequency sequence, calculating the quotient of the mean of the instantaneous frequency sequence and the standard deviation of the instantaneous frequency sequence to obtain the frequency change characteristics of each intrinsic mode function term; determining the instantaneous amplitude sequence according to the amplitude coefficient. column; calculate the mean and standard deviation of the instantaneous amplitude sequence, calculate the quotient of the mean of the instantaneous amplitude sequence and the standard deviation of the instantaneous amplitude sequence, and obtain the amplitude fault characteristics of each eigenmode function item; divide the eigenmode function items into normal items and fault items according to the preset fault marking points; obtain the amplitude coefficients corresponding to the normal items to obtain the first instantaneous amplitude sequence, obtain the amplitude coefficients corresponding to the fault items to obtain the second instantaneous amplitude sequence; determine the normal energy sum according to the normal items, determine the normal energy mean of the normal items according to the normal energy sum, and calculate the fault energy value of the fault item; determine the energy distortion characteristics of the fault item according to the fault energy value and the normal energy mean.

[0067] The vector conversion module 303 is used to perform vector processing on the electrical feature set to obtain an electrical feature vector, and to perform vector processing on the switch data to obtain a switch feature vector.

[0068] The fault analysis module 304 is used to input the electrical feature vector and the switch feature vector into a pre-trained fault analysis model, and output the fault time period, the fault cause within the fault time period, and the credibility of the fault cause.

[0069] In some embodiments, when the fault analysis module 304 is used to input the electrical feature vector and the switch feature vector into a pre-trained fault analysis model and output the fault time period, the fault cause within the fault time period, and the credibility of the fault cause, it is specifically used to implement: inputting the electrical feature vector and the switch feature vector into the pre-trained fault analysis model, wherein the fault analysis model includes a circuit fault detection network and a decision network, the circuit fault detection network includes a first long short-term memory network, a second long short-term memory network, and a threshold recurrent network, and the decision network includes a fully connected layer and a sigmoid function; performing hidden state feature extraction on the electrical feature vector using the first long short-term memory network in the circuit fault detection network to obtain a first hidden state feature vector; performing hidden state feature extraction on the switch feature vector using the second long short-term memory network in each circuit fault detection network to obtain a second hidden state feature vector; inputting the first hidden state feature vector and the second hidden state feature vector into the threshold recurrent network for feature integration to generate a fault feature vector; inputting the fault feature vector into the decision network, and performing fault cause analysis on the fault feature vector using the fully connected layer and the sigmoid function in the decision network to obtain the fault cause and the credibility of the fault cause.

[0070] In some embodiments, after the fault analysis module 304 is used to input the electrical feature vector and the switch feature vector into a pre-trained fault analysis model and output the fault time period, the fault cause within the fault time period and the credibility of the fault cause, it is also specifically used to implement: matching the corresponding target fault handling strategy from multiple preset fault handling strategies based on the fault cause whose credibility is higher than the preset credibility threshold.

[0071] The fault classification module 305 is configured to set the credibility lower than a preset credibility threshold as a target credibility, set the fault cause corresponding to the target credibility as a target fault cause, and set the fault time period corresponding to the target fault cause as a target fault time period.

[0072] The data upload module 306 is used to extract data from the recorded waveform data according to the fault time period to obtain target recorded waveform data, generate data to be analyzed based on the target recorded waveform data, target fault cause, and target credibility, and encrypt the data to be analyzed before uploading it to the cloud server.

[0073] In some embodiments, when the data upload module 306 is used to encrypt the data to be analyzed and upload it to the cloud server, it is specifically used to achieve: obtaining the data to be analyzed from the pole-mounted circuit breaker, framing the data to be analyzed to obtain multiple groups of sampling data; performing Fourier transform operation on the sampling data to generate spectrum data; determining first target data and second target data from the spectrum data according to a preset first spectrum interval and a second spectrum interval, and performing a scrambling operation on the first target data and the second target data according to a preset first key to obtain first scrambled data; performing an inverse Fourier transform operation on the first scrambled data to generate time domain data; performing a grouping operation on the time domain data to obtain multiple groups of frame data, and performing a scrambling operation on each group of frame data to obtain second scrambled data; cross-reorganizing the second scrambled data according to a preset second key to obtain encrypted data, and sending the encrypted data to the server.

[0074] In some embodiments, when the data upload module 306 is used to implement a scrambling operation on the first target data and the second target data according to a preset first key to obtain first scrambled data, it is specifically used to implement: inputting the first key into a preset quantum number generator to generate a first random number and a second random number, the number of characters of the first random number is the same as the number of characters of the first target data, and the number of characters of the second random number is the same as the number of characters of the second target data; performing an offset operation on the first target data according to the first random number to obtain first offset data, and performing an offset operation on the second target data according to the second random number to obtain second offset data; performing an exchange operation on the first offset data and the second offset data according to the first key and a preset exchange function to obtain first exchange data and second exchange data; generating a set of complex weight coefficients according to the first key, performing weighted processing on the first exchange data according to the complex weight coefficients to obtain first scrambled data, and performing weighted processing on the second exchange data according to the complex weight coefficients to obtain second scrambled data.

[0075] An embodiment of the present application provides an electronic device, which includes a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement a real-time data management method for a pole-mounted circuit breaker as described in any one of the embodiments of the present application when executing the computer program.

[0076] An embodiment of the present application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor implements a real-time data management method for a pole-mounted circuit breaker as described in any one of the embodiments of the present application.

[0077] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present application, and such modifications or substitutions should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A real-time data management method for a pole-mounted circuit breaker, characterized in that: Applied to an edge device, the method includes: Acquire recorded wave data of the pole-mounted circuit breaker, classify and process the recorded wave data to obtain electrical data and switch data; Perform feature extraction based on the electrical data to obtain an electrical feature set; Performing vector processing on the electrical feature set to obtain an electrical feature vector, and performing vector processing on the switch data to obtain a switch feature vector; Input the electrical feature vector and the switch feature vector into a pre-trained fault analysis model, wherein the fault analysis model includes a circuit fault detection network and a decision network, the circuit fault detection network includes a first long short-term memory network, a second long short-term memory network and a threshold recurrent network, and the decision network includes a fully connected layer and a Sigmoid function; perform hidden state feature extraction on the electrical feature vector through the first long short-term memory network in the circuit fault detection network to obtain a first hidden state feature vector; perform hidden state feature extraction on the switch feature vector through the second long short-term memory network in each circuit fault detection network to obtain a second hidden state feature vector; input the first hidden state feature vector and the second hidden state feature vector into the threshold recurrent network for feature integration to generate a fault feature vector; input the fault feature vector into the decision network, and perform fault cause analysis on the fault feature vector through the fully connected layer and Sigmoid function in the decision network to obtain the fault cause and the credibility of the fault cause; The credibility lower than the preset credibility threshold is set as the target credibility, the fault cause corresponding to the target credibility is set as the target fault cause, and the fault time period corresponding to the target fault cause is set as the target fault time period; Extract data from the recorded data according to the target fault time period to obtain target recorded data, generate data to be analyzed according to the target recorded data, the target fault cause and the target credibility, and encrypt the data to be analyzed and upload it to the cloud server.

2. The real-time data management method for a pole-mounted circuit breaker according to claim 1, wherein: The extracting features based on the electrical data to obtain an electrical feature set includes: Based on local maxima and local minima of the electrical data; Determine a first extreme value line according to the local maximum value, and determine a second extreme value line according to the local minimum value; determining an average extreme value line based on the first extreme value line and the second extreme value line; determining a plurality of eigenmode function terms and a residual term based on the electrical data and the average extreme value line; Feature extraction is performed based on the eigenmode function term to obtain frequency change features, amplitude fault features and energy distortion features.

3. The real-time data management method for a pole mounted circuit breaker according to claim 2, wherein: Feature extraction is performed based on the eigenmode function term to obtain frequency change features, amplitude fault features, and energy distortion features, including: Performing Hilbert transform on each of the eigenmode function terms to obtain an analytical signal sequence, wherein the analytical signal sequence includes: an amplitude coefficient and a phase coefficient; Derivative the phase coefficient to obtain an instantaneous frequency sequence; Calculating the mean and standard deviation of the instantaneous frequency sequence, and calculating the quotient of the mean of the instantaneous frequency sequence and the standard deviation of the instantaneous frequency sequence to obtain the frequency variation characteristics of each eigenmode function item; determining an instantaneous amplitude sequence according to the amplitude coefficient; Calculating the mean and standard deviation of the instantaneous amplitude sequence, and calculating the quotient of the mean of the instantaneous amplitude sequence and the standard deviation of the instantaneous amplitude sequence to obtain the amplitude fault feature of each of the eigenmode function items; Dividing the intrinsic mode function term into a normal term and a fault term according to a preset fault marking point; The amplitude coefficient corresponding to the normal item is obtained to obtain a first instantaneous amplitude sequence, and the amplitude coefficient corresponding to the fault item is obtained to obtain a second instantaneous amplitude sequence; Determine a normal energy sum according to the normal items, determine a normal energy mean of the normal items according to the normal energy sum, and calculate a fault energy value of the fault item; An energy distortion characteristic of the fault item is determined according to the fault energy value and the normal energy mean.

4. The real-time data management method for a pole mounted circuit breaker according to claim 1, wherein: After inputting the fault feature vector into the decision network and performing fault cause analysis on the fault feature vector using a fully connected layer and a Sigmoid function in the decision network to obtain the fault cause and the credibility of the fault cause, the method further includes: According to the fault cause whose credibility is higher than a preset credibility threshold, a corresponding target fault handling strategy is matched from a plurality of preset fault handling strategies.

5. The real-time data management method for a pole mounted circuit breaker according to claim 1, wherein: The step of encrypting the data to be analyzed and uploading it to the cloud server includes: Acquire data to be analyzed from a pole-mounted circuit breaker, perform frame processing on the data to be analyzed, and obtain multiple groups of sampling data; Performing a Fourier transform operation on the sampled data to generate spectrum data; Determining first target data and second target data from the spectrum data according to a preset first spectrum interval and a second spectrum interval, and performing a scrambling operation on the first target data and the second target data according to a preset first key to obtain first scrambled data; Performing an inverse Fourier transform operation on the first scrambled data to generate time domain data; performing a grouping operation on the time domain data to obtain multiple groups of framed data, and performing a scrambling operation on each group of the framed data to obtain second scrambled data; The second scrambled data is cross-reorganized according to a preset second key to obtain encrypted data, and the encrypted data is sent to the server.

6. The real-time data management method for a pole mounted circuit breaker according to claim 5, wherein: The performing a scrambling operation on the first target data and the second target data according to a preset first key to obtain first scrambled data includes: Inputting the first key into a preset quantum number generator to generate a first random number and a second random number, wherein the number of characters in the first random number is the same as the number of characters in the first target data, and the number of characters in the second random number is the same as the number of characters in the second target data; performing an offset operation on the first target data according to the first random number to obtain first offset data, and performing an offset operation on the second target data according to the second random number to obtain second offset data; performing an exchange operation on the first offset data and the second offset data respectively according to the first key and a preset exchange function to obtain first exchange data and second exchange data; A set of complex weight coefficients is generated according to the first key, the first exchange data is weighted according to the complex weight coefficients to obtain the first scrambled data, and the second exchange data is weighted according to the complex weight coefficients to obtain the second scrambled data.

7. A real-time data management device for a pole-mounted circuit breaker, characterized in that: The real-time data management device for a pole-mounted circuit breaker is used to execute the real-time data management method for a pole-mounted circuit breaker according to any one of claims 1 to 6, and the real-time data management device includes: A data processing module is used to obtain the recorded wave data of the pole-mounted circuit breaker, classify and process the recorded wave data, and obtain electrical data and switch data; A feature extraction module, configured to extract features based on the electrical data to obtain an electrical feature set; a vector conversion module, configured to perform vector processing on the electrical feature set to obtain an electrical feature vector, and to perform vector processing on the switch data to obtain a switch feature vector; A fault analysis module is configured to input the electrical feature vector and the switch feature vector into a pre-trained fault analysis model, wherein the fault analysis model includes a circuit fault detection network and a decision network, the circuit fault detection network includes a first long short-term memory network, a second long short-term memory network, and a threshold recurrent network, and the decision network includes a fully connected layer and a Sigmoid function; performing hidden state feature extraction on the electrical feature vector through the first long short-term memory network in the circuit fault detection network to obtain a first hidden state feature vector; performing hidden state feature extraction on the switch feature vector through the second long short-term memory network in each circuit fault detection network to obtain a second hidden state feature vector; inputting the first hidden state feature vector and the second hidden state feature vector into the threshold recurrent network for feature integration to generate a fault feature vector; inputting the fault feature vector into the decision network, and performing fault cause analysis on the fault feature vector through the fully connected layer and Sigmoid function in the decision network to obtain a fault cause and the credibility of the fault cause; a fault classification module, configured to set a credibility lower than a preset credibility threshold as a target credibility, set a fault cause corresponding to the target credibility as a target fault cause, and set a fault time period corresponding to the target fault cause as a target fault time period; The data uploading module is used to extract data from the recorded data according to the target fault time period to obtain target recorded data, generate data to be analyzed according to the target recorded data, the target fault cause and the target credibility, and encrypt the data to be analyzed and upload it to the cloud server.

8. An electronic device, characterized in that: The electronic device includes a memory and a processor; The memory is used to store computer programs; The processor is configured to execute the computer program and implement the real-time data management method for a pole-mounted circuit breaker according to any one of claims 1 to 6 when executing the computer program.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, enables the processor to implement the real-time data management method for a pole-mounted circuit breaker according to any one of claims 1 to 6.

Images