Power data processing method, system and medium based on satellite time synchronization

Abstract

The present application provides a method, system and medium for processing electric data based on satellite time synchronization. The method obtains the equipment time signal through the equipment clock in the electric equipment to be monitored, receives the satellite time signal through the satellite signal receiver, and then determines the frequency difference signal between the equipment time signal and the satellite time signal, so as to determine the adjustment voltage by using the frequency difference signal, and then adjust the signal output frequency of the equipment clock by using the adjustment voltage, so as to provide accurate satellite time for the electric data obtained by the electric equipment to be monitored, effectively reduce the time error during the processing of electric data, and then ensure the accuracy of the subsequent data management platform for the processing of electric data, so as to realize that the data management platform automatically determines the operating state corresponding to the electric data according to the preset electric data processing model, improves the intelligence of the electric system, and greatly improves the management efficiency of the electric system.

Description

Technical Field

[0001] The present application relates to data processing technology, and in particular to a method, system and medium for processing electric power data based on satellite time synchronization. Background Art

[0002] Satellite time synchronization refers to the process of correcting and synchronizing the clocks of ground equipment using precise time signals sent by satellite navigation systems in Earth orbit.

[0003] In the power system, the power equipment to be monitored usually has a local clock built in, and the local clock can be used to record the acquisition time of the acquired power data. However, the local clock may cause time errors due to factors such as original equipment errors, equipment aging, and capacitance changes in the clock circuit. Therefore, a power data processing method that can be synchronized based on satellite time is urgently needed. Summary of the invention

[0004] The present application provides a method, system and medium for processing power data based on satellite time synchronization to reduce the time error during power data processing, thereby ensuring the accuracy of power data processing by a subsequent data management platform.

[0005] In a first aspect, the present application provides a method for processing electric power data based on satellite time synchronization, which is applied to an electric power operation data management system, wherein the electric power operation data management system includes an electric power operation data management platform and a cluster of electric power equipment to be monitored, wherein each electric power equipment to be monitored in the cluster of electric power equipment to be monitored is communicatively connected to the electric power operation data management platform, and the electric power equipment to be monitored is provided with a processor, an equipment clock and a satellite signal receiver, and the equipment clock and the satellite signal receiver are respectively communicatively connected to the processor; the method includes:

[0006] The first power device to be monitored acquires a first device time signal through a first device clock, and receives a first satellite time signal through a first satellite signal receiver, wherein the first power device to be monitored is a power device to be monitored in the power device cluster to be monitored;

[0007] The first monitored power device determines a first frequency difference signal between the first device time signal and the first satellite time signal using a first processor, so as to determine a first adjustment voltage using the first frequency difference signal, wherein the first adjustment voltage is used to adjust the signal output frequency of the first device clock;

[0008] The first monitored power device acquires first power data, where the first power data includes a first monitoring time and first monitoring data, where the first monitoring time is the device time output by the first device clock after being adjusted by the first regulating voltage;

[0009] The first monitored power equipment sends the first power data to the data management platform, so that the data management platform determines the first operating state corresponding to the first power data according to a preset power data processing model, wherein the preset power data processing model is a decision classification model generated after training a preset neural network model using a preset power data training set.

[0010] Optionally, the determining a first adjustment voltage by using the first frequency difference signal includes:

[0011] The first processor determines the first adjustment voltage according to a preset characteristic curve and the first frequency difference signal, wherein the preset characteristic curve is a voltage and frequency characteristic curve corresponding to the first device clock.

[0012] Optionally, after the first processor determines the first adjustment voltage according to a preset characteristic curve and the first frequency difference signal, the method further includes:

[0013] The first monitored power device obtains a second device time signal through the first device clock, and receives a second satellite time signal through the first satellite signal receiver, wherein the first satellite time signal and the second satellite time signal are heterogeneous satellite signals;

[0014] The first monitored power device determines a second frequency difference signal between the second device time signal and the second satellite time signal using the first processor, so as to determine a second regulated voltage using the second frequency difference signal;

[0015] If the second regulation voltage is less than a preset minimum regulation voltage of the first device clock, adjusting the signal output frequency of the first device clock using the first regulation voltage;

[0016] If the second regulation voltage is greater than the preset minimum regulation voltage of the first device clock, a third regulation voltage is used to adjust the signal output frequency of the first device clock, and the third regulation voltage is the sum of the first regulation voltage and the second regulation voltage.

[0017] Optionally, before the data management platform determines the first operating state corresponding to the first power data according to a preset power data processing model, the method further includes:

[0018] Determine, by using a preset vector machine, a state decision boundary between a first normal operating data set and a first abnormal operating data set, wherein the amount of data in the first normal operating data set is greater than the amount of data in the first abnormal operating data set;

[0019] Calculating a characteristic distance between the state decision boundary and each normal operation data in the first normal operation data set, and deleting normal operation data corresponding to a characteristic distance greater than a first preset distance threshold from the first normal operation data set to form a second normal operation data set;

[0020] Calculating a characteristic distance between the state decision boundary and each abnormal operation data in the first abnormal operation data set, and deleting abnormal operation data corresponding to a characteristic distance greater than a second preset distance threshold from the first abnormal operation data set to form a second abnormal operation data set, wherein the first preset distance threshold is greater than the second preset distance threshold;

[0021] The preset power data training set is determined according to the second normal operation data set and the second abnormal operation data set.

[0022] Optionally, the first monitored power device determines a first frequency difference signal between the first device time signal and the first satellite time signal using a first processor, including:

[0023] The first processor uses formula 1 and according to the first device time signal and the first satellite time signal Determine the first frequency difference signal ΔT k , the formula 1 is:

[0024]

[0025] Among them, T k is the standard time signal at time k, ω k is the random noise generated by the standard time signal in the communication link between the satellite and the first satellite signal receiver, ρ is the system frequency offset coefficient of the first device clock, is the system original deviation of the clock of the first device;

[0026] The first processor calculates the first frequency difference signal ΔT according to a preset deviation estimation elimination algorithm. k Deviation elimination is performed to determine a first frequency difference update signal, wherein the first frequency difference update signal is used to replace the first frequency difference signal to determine the first adjustment voltage.

[0027] Optionally, after determining the preset power data training set according to the second normal operation data set and the second abnormal operation data set, the method further includes:

[0028] According to the preset power data training set D={d1,…,d i ,…,p n ,}Determine the power data input set P={p1,…,p i ,…,p n} and the running state output set Q = {q1,…,q i ,…,q n}, where p i is the ith power data input in the power data input set P, q i For p i The corresponding preset operating state output, n is the number of power data inputs in the preset power data training set D;

[0029] According to the power data input set P = {p1, ..., p i ,…,p n} and the power data processing model to be trained determines the operating state estimation output set Q' = {q'1, ..., q' i ,…,q' n};

[0030] The first estimated deviation δ1 and the second estimated deviation δ2 corresponding to the running state estimated output set Q' are determined using Formula 2, which is:

[0031]

[0032] If the first estimated deviation δ1 is smaller than a preset first estimated deviation threshold and the second estimated deviation δ2 is smaller than a preset second estimated deviation threshold, the preset power data processing model is generated according to the trained power data processing model to be trained.

[0033] Optionally, after determining the first estimation deviation δ1 and the second estimation deviation δ2 corresponding to the running state estimation output set Q', the method further includes:

[0034] If the first estimated deviation δ1 is greater than the preset first estimated deviation threshold and / or the second estimated deviation δ2 is greater than the preset second estimated deviation threshold, then the error adjustment weight K of each layer in the power data processing model to be trained is determined using Formula 3 according to the first estimated deviation δ1 and the second estimated deviation δ2. Formula 3 is:

[0035]

[0036] Among them, r is the number of neural network layers of the power data processing model to be trained, and μ is the current learning rate of the power data processing model to be trained.

[0037] In a second aspect, the present application provides an electric power operation data management system, including: an electric power operation data management platform and a cluster of electric power equipment to be monitored, wherein each electric power equipment to be monitored in the cluster of electric power equipment to be monitored is communicatively connected to the electric power operation data management platform, and the electric power equipment to be monitored is provided with a processor, an equipment clock and a satellite signal receiver, and the equipment clock and the satellite signal receiver are communicatively connected to the processor respectively;

[0038] The first power device to be monitored acquires a first device time signal through a first device clock, and receives a first satellite time signal through a first satellite signal receiver, wherein the first power device to be monitored is a power device to be monitored in the power device cluster to be monitored;

[0039] The first monitored power device determines a first frequency difference signal between the first device time signal and the first satellite time signal using a first processor, so as to determine a first adjustment voltage using the first frequency difference signal, wherein the first adjustment voltage is used to adjust the signal output frequency of the first device clock;

[0040] The first monitored power device acquires first power data, where the first power data includes a first monitoring time and first monitoring data, where the first monitoring time is the device time output by the first device clock after being adjusted by the first regulating voltage;

[0041] The first monitored power equipment sends the first power data to the data management platform, so that the data management platform determines the first operating state corresponding to the first power data according to a preset power data processing model, wherein the preset power data processing model is a decision classification model generated after training a preset neural network model using a preset power data training set.

[0042] Optionally, the first processor determines the first adjustment voltage according to a preset characteristic curve and the first frequency difference signal, wherein the preset characteristic curve is a voltage and frequency characteristic curve corresponding to the first device clock.

[0043] Optionally, the first monitored power device obtains a second device time signal through the first device clock, and receives a second satellite time signal through the first satellite signal receiver, and the first satellite time signal and the second satellite time signal are heterogeneous satellite signals;

[0044] The first monitored power device determines a second frequency difference signal between the second device time signal and the second satellite time signal using the first processor, so as to determine a second regulated voltage using the second frequency difference signal;

[0045] If the second regulation voltage is less than a preset minimum regulation voltage of the first device clock, adjusting the signal output frequency of the first device clock using the first regulation voltage;

[0046] If the second regulation voltage is greater than the preset minimum regulation voltage of the first device clock, a third regulation voltage is used to adjust the signal output frequency of the first device clock, and the third regulation voltage is the sum of the first regulation voltage and the second regulation voltage.

[0047] Optionally, a state decision boundary between a first normal operation data set and a first abnormal operation data set is determined using a preset vector machine, wherein the amount of data in the first normal operation data set is greater than the amount of data in the first abnormal operation data set;

[0048] Calculating a characteristic distance between the state decision boundary and each normal operation data in the first normal operation data set, and deleting normal operation data corresponding to a characteristic distance greater than a first preset distance threshold from the first normal operation data set to form a second normal operation data set;

[0049] Calculating a characteristic distance between the state decision boundary and each abnormal operation data in the first abnormal operation data set, and deleting abnormal operation data corresponding to a characteristic distance greater than a second preset distance threshold from the first abnormal operation data set to form a second abnormal operation data set, wherein the first preset distance threshold is greater than the second preset distance threshold;

[0050] The preset power data training set is determined according to the second normal operation data set and the second abnormal operation data set.

[0051] Optionally, the first processor uses formula 1 and calculates the time signal of the first device according to the first device time signal. and the first satellite time signal Determine the first frequency difference signal ΔT k , the formula 1 is:

[0052]

[0053] Among them, T k is the standard time signal at time k, ω kis the random noise generated by the standard time signal in the communication link between the satellite and the first satellite signal receiver, ρ is the system frequency offset coefficient of the first device clock, is the system original deviation of the clock of the first device;

[0054] The first processor calculates the first frequency difference signal ΔT according to a preset deviation estimation elimination algorithm. k Deviation elimination is performed to determine a first frequency difference update signal, wherein the first frequency difference update signal is used to replace the first frequency difference signal to determine the first adjustment voltage.

[0055] Optionally, according to the preset power data training set D={d1,…,d i ,…,p n ,}Determine the power data input set P={p1,…,p i ,…,p n} and the running state output set Q = {q1,…,q i ,…,q n}, where p i is the ith power data input in the power data input set P, q i For p i The corresponding preset operating state output, n is the number of power data inputs in the preset power data training set D;

[0056] According to the power data input set P = {p1, ..., p i ,…,p n} and the power data processing model to be trained determines the operating state estimation output set Q' = {q'1, ..., q' i ,…,q' n};

[0057] The first estimated deviation δ1 and the second estimated deviation δ2 corresponding to the running state estimated output set Q' are determined using Formula 2, which is:

[0058]

[0059] If the first estimated deviation δ1 is smaller than a preset first estimated deviation threshold and the second estimated deviation δ2 is smaller than a preset second estimated deviation threshold, the preset power data processing model is generated according to the trained power data processing model to be trained.

[0060] Optionally, if the first estimated deviation δ1 is greater than the preset first estimated deviation threshold and / or the second estimated deviation δ2 is greater than the preset second estimated deviation threshold, formula 3 is used to determine the error adjustment weight K of each layer in the power data processing model to be trained according to the first estimated deviation δ1 and the second estimated deviation δ2. Formula 3 is:

[0061]

[0062] Among them, r is the number of neural network layers of the power data processing model to be trained, and μ is the current learning rate of the power data processing model to be trained.

[0063] In a third aspect, the present application provides an electronic device, including:

[0064] processor; and,

[0065] A memory, configured to store executable instructions of the processor;

[0066] The processor is configured to perform any possible method described in the first aspect by executing the executable instructions.

[0067] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement any possible method described in the first aspect.

[0068] The power data processing method, system and medium based on satellite time synchronization provided by the present application obtain the device time signal through the device clock in the power equipment to be monitored, and receive the satellite time signal through the satellite signal receiver, and then determine the frequency difference signal between the device time signal and the satellite time signal, so as to determine the adjustment voltage using the frequency difference signal, and then adjust the signal output frequency of the device clock using the adjustment voltage, so as to provide accurate satellite time for the power data obtained by the power equipment to be monitored, effectively reduce the time error during power data processing, and then ensure the accuracy of the subsequent data management platform's processing of the power data, so as to realize the data management platform automatically determining the operating status corresponding to the power data according to the preset power data processing model, thereby improving the intelligence level of the power system and greatly improving the management efficiency of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0070] Figure 1is a flowchart of a method for processing power data based on satellite time synchronization according to an exemplary embodiment of the present application;

[0071] Figure 2 is a flowchart of a method for processing power data based on satellite time synchronization according to another exemplary embodiment of the present application;

[0072] Figure 3 is a structural schematic diagram of a power operation data management system according to an exemplary embodiment of the present application;

[0073] Figure 4 It is a schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present application.

[0074] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0075] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0076] Figure 1 FIG. 1 is a flow chart of a method for processing power data based on satellite time synchronization according to an exemplary embodiment of the present application. Figure 1 As shown, the method provided in this embodiment includes:

[0077] S101. A first power device to be monitored obtains a first device time signal through a first device clock, and receives a first satellite time signal through a first satellite signal receiver.

[0078] The method provided in this embodiment can be applied to an electric power operation data management system. Specifically, the electric power operation data management system includes an electric power operation data management platform and a cluster of electric power equipment to be monitored, wherein each electric power equipment to be monitored in the cluster of electric power equipment to be monitored is communicatively connected to the electric power operation data management platform, and a processor, an equipment clock, and a satellite signal receiver are provided in the electric power equipment to be monitored, and the equipment clock and the satellite signal receiver are communicatively connected to the processor respectively.

[0079] In this step, the first power device to be monitored can obtain the first device time signal through the first device clock, and receive the first satellite time signal through the first satellite signal receiver, wherein the first power device to be monitored is a power device to be monitored in the power device cluster to be monitored. Optionally, the power device to be monitored can be a power generation device, such as a turbine, a generator, a power station boiler, and a solar panel, etc., and can also be a power supply device, such as a transformer, a mutual inductor, a switch cabinet, and a circuit breaker, etc. In addition, the above-mentioned first device clock can be an electronic clock or a quartz crystal oscillator, and the above-mentioned satellite signal receiver can receive a Beidou satellite timing signal or a Global Positioning System (GPS) timing signal.

[0080] S102: The first power device to be monitored uses a first processor to determine a first frequency difference signal between a first device time signal and a first satellite time signal.

[0081] In this step, the first power device to be monitored can use the first processor to determine the first frequency difference signal between the first device time signal and the first satellite time signal, and use the first frequency difference signal to determine the first adjustment voltage, and the first adjustment voltage is used to adjust the signal output frequency of the first device clock.

[0082] Specifically, the first processor may use formula 1 and calculate the time signal of the first device according to and the first satellite time signal Determine the first frequency difference signal ΔT k , where the above formula 1 is:

[0083]

[0084] Among them, T k is the standard time signal at time k, ω k is the random noise generated by the standard time signal in the communication link between the satellite and the first satellite signal receiver, ρ is the system frequency offset coefficient of the first device clock, is the system raw deviation of the first device clock.

[0085] Then, the first processor performs an error estimation and elimination algorithm on the first frequency difference signal ΔT k Deviation elimination is performed to determine a first frequency difference update signal, and the first frequency difference update signal is used to replace the first frequency difference signal to determine the first adjustment voltage.

[0086] It is worth noting that the above-mentioned preset deviation estimation elimination algorithm may adopt particle filtering, Kalman filtering, Hotelling filtering, dynamic Bayesian network and other algorithms. k When it is the frequency difference signal between the Beidou second pulse signal and the crystal oscillator output signal, the above algorithm is used to filter the random deviation and accumulated error included therein to obtain the optimal estimate of the frequency difference signal.

[0087] In addition, during the operation of the system, if the Beidou receiver is interfered with or there are too few satellites to track, the second pulse signal will have a large jump wild value, which will cause the system error to increase. Therefore, the Kalman filter can be further optimized to improve the robustness and adaptability of the algorithm. Optionally, a radial basis function neural network (RBF) can be used as a feedforward neural network to map any complex nonlinear relationship through its nonlinear fitting ability, thereby improving robustness and self-learning ability. The RBF neural network is used to correct the Kalman filter error, and the corrected filter error is compensated to the filter estimate to obtain a new filter value.

[0088] The above-mentioned determination of the first adjustment voltage using the first frequency difference signal may specifically include: the first processor determines the first adjustment voltage according to a preset characteristic curve and the first frequency difference signal, wherein the preset characteristic curve is a voltage and frequency characteristic curve corresponding to the first device clock.

[0089] In addition, after the first processor determines the first adjustment voltage according to the preset characteristic curve and the first frequency difference signal, the first power device to be monitored can also obtain the second device time signal through the first device clock, and receive the second satellite time signal through the first satellite signal receiver, and the first satellite time signal and the second satellite time signal are heterogeneous satellite signals. The first power device to be monitored uses the first processor to determine the second frequency difference signal between the second device time signal and the second satellite time signal, so as to determine the second adjustment voltage using the second frequency difference signal. If the second adjustment voltage is less than the preset minimum adjustment voltage of the first device clock, the signal output frequency of the first device clock is adjusted using the first adjustment voltage. If the second adjustment voltage is greater than the preset minimum adjustment voltage of the first device clock, the signal output frequency of the first device clock is adjusted using the third adjustment voltage, and the third adjustment voltage is the sum of the first adjustment voltage and the second adjustment voltage. It is worth noting that by verifying the above-mentioned multi-source satellite signals and combining the verification results with the adjustment characteristics of the device clock to meet the voltage adjustment under different verification results, the accuracy and reliability of the satellite time synchronization signal are further improved. It is worth noting that the selection of the first satellite time signal and the second satellite time signal can also be determined based on the strength of the two signals, and the satellite time signal received first is used as the first satellite time signal, and the satellite time signal received later is used as the second satellite time signal, thereby ensuring the timeliness of the signal output frequency signal for the device clock. In addition, the satellite time signal with a weaker signal can be used as the first satellite time signal, and the stronger satellite time signal can be used as the second satellite time signal, thereby ensuring the accuracy of the signal output frequency signal for the device clock.

[0090] S103: The first power equipment to be monitored obtains first power data.

[0091] In this step, the first power device to be monitored acquires first power data, the first power data includes a first monitoring time and first monitoring data, and the first monitoring time is the device time output by the first device clock adjusted by the first regulating voltage.

[0092] S104: The first power equipment to be monitored sends first power data to the data management platform.

[0093] Specifically, the first power equipment to be monitored sends the first power data to the data management platform, so that the data management platform determines the first operating state corresponding to the first power data according to the preset power data processing model, wherein the preset power data processing model is a decision classification model generated after training the preset neural network model using the preset power data training set. It is worth noting that the above-mentioned data management platform can also use the preset neural network model to make a comprehensive judgment on the power data uploaded by multiple power equipment to be monitored, so as to meet the needs of judging the operating state of the power system under complex working conditions. It can be understood that after the above-mentioned satellite time synchronization for each power equipment to be monitored, the power data uploaded by each power equipment to be monitored has a unified time reference, which is of great significance for timely detection and prediction of power system failures on the basis of ensuring data accuracy and consistency.

[0094] In this embodiment, the device time signal is obtained by the device clock in the power equipment to be monitored, and the satellite time signal is received by the satellite signal receiver, and then the frequency difference signal between the device time signal and the satellite time signal is determined, so as to determine the adjustment voltage using the frequency difference signal, and then adjust the signal output frequency of the device clock using the adjustment voltage, so as to provide accurate satellite time for the power data obtained by the power equipment to be monitored, effectively reducing the time error during power data processing, and thus ensuring the accuracy of the subsequent data management platform's processing of the power data, so that the data management platform automatically determines the operating status corresponding to the power data according to the preset power data processing model, thereby improving the intelligence level of the power system and greatly improving the management efficiency of the power system.

[0095] Figure 2 FIG. 1 is a flow chart of a method for processing power data based on satellite time synchronization according to another exemplary embodiment of the present application. Figure 2 As shown, the power data processing method based on satellite time synchronization provided in this embodiment includes:

[0096] S201. A first power device to be monitored obtains a first device time signal through a first device clock, and receives a first satellite time signal through a first satellite signal receiver.

[0097] The method provided in this embodiment can be applied to an electric power operation data management system. Specifically, the electric power operation data management system includes an electric power operation data management platform and a cluster of electric power equipment to be monitored, wherein each electric power equipment to be monitored in the cluster of electric power equipment to be monitored is communicatively connected to the electric power operation data management platform, and a processor, an equipment clock, and a satellite signal receiver are provided in the electric power equipment to be monitored, and the equipment clock and the satellite signal receiver are communicatively connected to the processor respectively.

[0098] In this step, the first power device to be monitored can obtain the first device time signal through the first device clock, and receive the first satellite time signal through the first satellite signal receiver, wherein the first power device to be monitored is a power device to be monitored in the power device cluster to be monitored. Optionally, the power device to be monitored can be a power generation device, such as a turbine, a generator, a power station boiler, and a solar panel, etc., and can also be a power supply device, such as a transformer, a mutual inductor, a switch cabinet, and a circuit breaker, etc. In addition, the above-mentioned first device clock can be an electronic clock or a quartz crystal oscillator, and the above-mentioned satellite signal receiver can receive a Beidou satellite timing signal or a Global Positioning System (GPS) timing signal.

[0099] S202: The first power device to be monitored uses a first processor to determine a first frequency difference signal between a first device time signal and a first satellite time signal.

[0100] In this step, the first power device to be monitored can use the first processor to determine the first frequency difference signal between the first device time signal and the first satellite time signal, and use the first frequency difference signal to determine the first adjustment voltage, and the first adjustment voltage is used to adjust the signal output frequency of the first device clock.

[0101] Specifically, the first processor may use formula 1 and calculate the time signal of the first device according to and the first satellite time signal Determine the first frequency difference signal ΔT k , where the above formula 1 is:

[0102]

[0103] Among them, T k is the standard time signal at time k, ω k is the random noise generated by the standard time signal in the communication link between the satellite and the first satellite signal receiver, ρ is the system frequency offset coefficient of the first device clock, is the system raw deviation of the first device clock.

[0104] Then, the first processor performs an error estimation and elimination algorithm on the first frequency difference signal ΔT k Deviation elimination is performed to determine a first frequency difference update signal, and the first frequency difference update signal is used to replace the first frequency difference signal to determine the first adjustment voltage.

[0105] It is worth noting that the above-mentioned preset deviation estimation elimination algorithm may adopt particle filtering, Kalman filtering, Hotelling filtering, dynamic Bayesian network and other algorithms. k When it is the frequency difference signal between the Beidou second pulse signal and the crystal oscillator output signal, the above algorithm is used to filter the random deviation and accumulated error included therein to obtain the optimal estimate of the frequency difference signal.

[0106] In addition, during the operation of the system, if the Beidou receiver is interfered with or there are too few satellites to track, the second pulse signal will have a large jump wild value, which will cause the system error to increase. Therefore, the Kalman filter can be further optimized to improve the robustness and adaptability of the algorithm. Optionally, a radial basis function neural network (RBF) can be used as a feedforward neural network to map any complex nonlinear relationship through its nonlinear fitting ability, thereby improving robustness and self-learning ability. The RBF neural network is used to correct the Kalman filter error, and the corrected filter error is compensated to the filter estimate to obtain a new filter value.

[0107] The above-mentioned determination of the first adjustment voltage using the first frequency difference signal may specifically include: the first processor determines the first adjustment voltage according to a preset characteristic curve and the first frequency difference signal, wherein the preset characteristic curve is a voltage and frequency characteristic curve corresponding to the first device clock.

[0108] In addition, after the first processor determines the first adjustment voltage according to the preset characteristic curve and the first frequency difference signal, the first power device to be monitored can also obtain the second device time signal through the first device clock, and receive the second satellite time signal through the first satellite signal receiver, and the first satellite time signal and the second satellite time signal are heterogeneous satellite signals. The first power device to be monitored uses the first processor to determine the second frequency difference signal between the second device time signal and the second satellite time signal, so as to determine the second adjustment voltage using the second frequency difference signal. If the second adjustment voltage is less than the preset minimum adjustment voltage of the first device clock, the signal output frequency of the first device clock is adjusted using the first adjustment voltage. If the second adjustment voltage is greater than the preset minimum adjustment voltage of the first device clock, the signal output frequency of the first device clock is adjusted using the third adjustment voltage, and the third adjustment voltage is the sum of the first adjustment voltage and the second adjustment voltage. It is worth noting that by verifying the above-mentioned multi-source satellite signals and combining the verification results with the adjustment characteristics of the device clock to meet the voltage adjustment under different verification results, the accuracy and reliability of the satellite time synchronization signal are further improved. It is worth noting that the selection of the first satellite time signal and the second satellite time signal can also be determined based on the strength of the two signals, and the satellite time signal received first is used as the first satellite time signal, and the satellite time signal received later is used as the second satellite time signal, thereby ensuring the timeliness of the signal output frequency signal for the device clock. In addition, the satellite time signal with a weaker signal can be used as the first satellite time signal, and the stronger satellite time signal can be used as the second satellite time signal, thereby ensuring the accuracy of the signal output frequency signal for the device clock.

[0109] S203: The first power equipment to be monitored obtains first power data.

[0110] In this step, the first power device to be monitored acquires first power data, the first power data includes a first monitoring time and first monitoring data, and the first monitoring time is the device time output by the first device clock adjusted by the first regulating voltage.

[0111] S204: Determine a preset power data training set.

[0112] Specifically, a preset vector machine may be used to determine the state decision boundary between the first normal operation data set and the first abnormal operation data set, and the amount of data in the first normal operation data set is greater than the amount of data in the first abnormal operation data set. The preset vector machine may be a support vector machine (SVM), which may be used to determine the distance classification hyperplane of the sample, thereby determining the position of the hyperplane, and in this embodiment, the sample decision boundary may be the classification hyperplane determined by the support vector machine. In addition, the preset clustering algorithm may be a clustering algorithm based on one or more combinations of K-Means, hierarchical clustering, spectral clustering, etc.

[0113] Then, the characteristic distance between the state decision boundary and each normal operation data in the first normal operation data set is calculated, and the normal operation data corresponding to the characteristic distance greater than the first preset distance threshold is deleted from the first normal operation data set to form a second normal operation data set. And the characteristic distance between the state decision boundary and each abnormal operation data in the first abnormal operation data set is calculated, and the abnormal operation data corresponding to the characteristic distance greater than the second preset distance threshold is deleted from the first abnormal operation data set to form a second abnormal operation data set, wherein the first preset distance threshold is greater than the second preset distance threshold. Finally, the preset power data training set is determined according to the second normal operation data set and the second abnormal operation data set. It is worth noting that in the early stage of model training, in order to effectively train the model, the original operation data set is usually collected from different power systems, which may have large data differences. If training is performed directly, it is easy to cause the model to fail to converge for a long time, thereby affecting the efficiency of model training. By deleting the data with a large distance from the state decision boundary respectively, the noise data is effectively excluded, which is beneficial to the training efficiency of the model in the early stage. In the subsequent actual operation process, the operation data of the specific power system used by the model can be further added to the preset power data training set, so as to train a special model suitable for different power systems. In addition, since the amount of abnormal operation state data in the monitoring data of the power system is usually much smaller than the normal operation state data, if the method of directly using the collected data to train the model is used, it is easy to cause the trained model to shift to the majority class samples in the classification results. By limiting the first preset distance threshold to be greater than the second preset distance threshold, the amount of data in the second normal operation data set is close to that in the second abnormal operation data set. For example, the ratio of the two can be controlled to be less than the preset ratio, so that the reliability of the training data can be trained while further improving the accuracy of the trained model.

[0114] S205 , training the power data processing model to be trained using a preset power data training set to generate a preset power data processing model.

[0115] Optionally, the power data training set D = {d1, ..., d i ,…,p n ,}Determine the power data input set P={p1,…,p i ,…,p n} and the running state output set Q = {q1,…,q i ,…,q n}, where p i is the ith power data input in the power data input set P, q i For p i The corresponding preset operating state output, n is the number of power data inputs in the preset power data training set D.

[0116] According to the power data input set P = {p1, ..., p i ,…,p n} and the power data processing model to be trained determines the operating state estimation output set Q' = {q'1, ..., q' i ,…,q' n}.

[0117] Using Formula 2, the first estimated deviation δ1 and the second estimated deviation δ2 corresponding to the running state estimated output set Q' are determined. Formula 2 is:

[0118]

[0119] If the first estimated deviation δ1 is less than the preset first estimated deviation threshold and the second estimated deviation δ2 is less than the preset second estimated deviation threshold, a preset power data processing model is generated according to the trained power data processing model to be trained. Through the above control of the multi-dimensional estimated deviation threshold, the accuracy and stability of the recognition result of the power data processing model to be trained can be guaranteed.

[0120] In addition, if the first estimated deviation δ1 is greater than the preset first estimated deviation threshold and / or the second estimated deviation δ2 is greater than the preset second estimated deviation threshold, then the error adjustment weight K of each layer in the power data processing model to be trained is determined using Formula 3 according to the first estimated deviation δ1 and the second estimated deviation δ2. Formula 3 is:

[0121]

[0122] Among them, r is the number of neural network layers of the power data processing model to be trained, and μ is the current learning rate of the power data processing model to be trained.

[0123] S206: The first power equipment to be monitored sends the first power data to the data management platform.

[0124] Specifically, the first power equipment to be monitored sends the first power data to the data management platform, so that the data management platform determines the first operating state corresponding to the first power data according to the preset power data processing model, wherein the preset power data processing model is a decision classification model generated after training the preset neural network model using the preset power data training set. It is worth noting that the above-mentioned data management platform can also use the preset neural network model to make a comprehensive judgment on the power data uploaded by multiple power equipment to be monitored, so as to meet the needs of judging the operating state of the power system under complex working conditions. It can be understood that after the above-mentioned satellite time synchronization for each power equipment to be monitored, the power data uploaded by each power equipment to be monitored has a unified time reference, which is of great significance for timely detection and prediction of power system failures on the basis of ensuring data accuracy and consistency.

[0125] Figure 3 FIG. 1 is a schematic diagram of a power operation data management system according to an exemplary embodiment of the present application. Figure 3 As shown, the power operation data management system 300 provided in this embodiment includes: a power operation data management platform 310 and a power equipment cluster 320 to be monitored, wherein each power equipment to be monitored in the power equipment cluster 320 to be monitored is communicatively connected to the power operation data management platform 310, and the power equipment to be monitored is provided with a processor, a device clock and a satellite signal receiver, and the device clock and the satellite signal receiver are respectively communicatively connected to the processor;

[0126] The first power device to be monitored acquires a first device time signal through a first device clock, and receives a first satellite time signal through a first satellite signal receiver, wherein the first power device to be monitored is one of the power devices to be monitored in the power device cluster 320 to be monitored;

[0127] The first monitored power device determines a first frequency difference signal between the first device time signal and the first satellite time signal using a first processor, so as to determine a first adjustment voltage using the first frequency difference signal, wherein the first adjustment voltage is used to adjust the signal output frequency of the first device clock;

[0128] The first monitored power device acquires first power data, where the first power data includes a first monitoring time and first monitoring data, where the first monitoring time is the device time output by the first device clock after being adjusted by the first regulating voltage;

[0129] The first monitored power equipment sends the first power data to the data management platform, so that the data management platform determines the first operating state corresponding to the first power data according to a preset power data processing model, wherein the preset power data processing model is a decision classification model generated after training a preset neural network model using a preset power data training set.

[0130] Optionally, the first processor determines the first adjustment voltage according to a preset characteristic curve and the first frequency difference signal, wherein the preset characteristic curve is a voltage and frequency characteristic curve corresponding to the first device clock.

[0131] Optionally, the first monitored power device obtains a second device time signal through the first device clock, and receives a second satellite time signal through the first satellite signal receiver, and the first satellite time signal and the second satellite time signal are heterogeneous satellite signals;

[0132] The first monitored power device determines a second frequency difference signal between the second device time signal and the second satellite time signal using the first processor, so as to determine a second regulated voltage using the second frequency difference signal;

[0133] If the second regulation voltage is less than a preset minimum regulation voltage of the first device clock, adjusting the signal output frequency of the first device clock using the first regulation voltage;

[0134] If the second regulation voltage is greater than the preset minimum regulation voltage of the first device clock, a third regulation voltage is used to adjust the signal output frequency of the first device clock, and the third regulation voltage is the sum of the first regulation voltage and the second regulation voltage.

[0135] Optionally, a state decision boundary between a first normal operation data set and a first abnormal operation data set is determined using a preset vector machine, wherein the amount of data in the first normal operation data set is greater than the amount of data in the first abnormal operation data set;

[0136] Calculating a characteristic distance between the state decision boundary and each normal operation data in the first normal operation data set, and deleting normal operation data corresponding to a characteristic distance greater than a first preset distance threshold from the first normal operation data set to form a second normal operation data set;

[0137] Calculating a characteristic distance between the state decision boundary and each abnormal operation data in the first abnormal operation data set, and deleting abnormal operation data corresponding to a characteristic distance greater than a second preset distance threshold from the first abnormal operation data set to form a second abnormal operation data set, wherein the first preset distance threshold is greater than the second preset distance threshold;

[0138] The preset power data training set is determined according to the second normal operation data set and the second abnormal operation data set.

[0139] Optionally, the first processor uses formula 1 and calculates the time signal of the first device according to the first device time signal. and the first satellite time signal Determine the first frequency difference signal ΔT k , the formula 1 is:

[0140]

[0141] Among them, T k is the standard time signal at time k, ω k is the random noise generated by the standard time signal in the communication link between the satellite and the first satellite signal receiver, ρ is the system frequency offset coefficient of the first device clock, is the system original deviation of the clock of the first device;

[0142] The first processor calculates the first frequency difference signal ΔT according to a preset deviation estimation elimination algorithm. k Deviation elimination is performed to determine a first frequency difference update signal, wherein the first frequency difference update signal is used to replace the first frequency difference signal to determine the first adjustment voltage.

[0143] Optionally, according to the preset power data training set D={d1,…,d i ,…,p n ,}Determine the power data input set P={p1,…,p i ,…,p n} and the running state output set Q = {q1,…,q i ,…,q n}, where p i is the ith power data input in the power data input set P, q i For p i The corresponding preset operating state output, n is the number of power data inputs in the preset power data training set D;

[0144] According to the power data input set P = {p1, ..., p i,…,p n} and the power data processing model to be trained determines the operating state estimation output set Q' = {q'1, ..., q' i ,…,q' n};

[0145] The first estimated deviation δ1 and the second estimated deviation δ2 corresponding to the running state estimated output set Q' are determined using Formula 2, which is:

[0146]

[0147] If the first estimated deviation δ1 is smaller than a preset first estimated deviation threshold and the second estimated deviation δ2 is smaller than a preset second estimated deviation threshold, the preset power data processing model is generated according to the trained power data processing model to be trained.

[0148] Optionally, if the first estimated deviation δ1 is greater than the preset first estimated deviation threshold and / or the second estimated deviation δ2 is greater than the preset second estimated deviation threshold, formula 3 is used to determine the error adjustment weight K of each layer in the power data processing model to be trained according to the first estimated deviation δ1 and the second estimated deviation δ2. Formula 3 is:

[0149]

[0150] Among them, r is the number of neural network layers of the power data processing model to be trained, and μ is the current learning rate of the power data processing model to be trained.

[0151] Figure 4 is a schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present application. Figure 4 As shown, an electronic device 400 provided in this embodiment includes: a processor 401 and a memory 402; wherein:

[0152] The memory 402 is used to store computer programs, and the memory may also be a flash memory.

[0153] The processor 401 is used to execute the execution instructions stored in the memory to implement each step in the above method. For details, please refer to the relevant description in the above method embodiment.

[0154] Optionally, the memory 402 may be independent or integrated with the processor 401 .

[0155] When the memory 402 is a device independent of the processor 401, the electronic device 400 may further include:

[0156] The bus 403 is used to connect the memory 402 and the processor 401 .

[0157] This embodiment further provides a readable storage medium, in which a computer program is stored. When at least one processor of an electronic device executes the computer program, the electronic device executes the methods provided in the above-mentioned various implementation modes.

[0158] This embodiment also provides a program product, which includes a computer program stored in a readable storage medium. At least one processor of the electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device implements the methods provided in the above various embodiments.

[0159] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the claims.

[0160] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A method for processing electric power data based on satellite time synchronization, characterized in that: Applied to an electric power operation data management system, the electric power operation data management system includes an electric power operation data management platform and a cluster of electric power equipment to be monitored, wherein each electric power equipment to be monitored in the cluster of electric power equipment to be monitored is communicatively connected to the electric power operation data management platform, and the electric power equipment to be monitored is provided with a processor, an equipment clock and a satellite signal receiver, and the equipment clock and the satellite signal receiver are communicatively connected to the processor respectively; the method includes: The first power device to be monitored acquires a first device time signal through a first device clock, and receives a first satellite time signal through a first satellite signal receiver, wherein the first power device to be monitored is a power device to be monitored in the power device cluster to be monitored; The first monitored power device determines a first frequency difference signal between the first device time signal and the first satellite time signal using a first processor, so as to determine a first adjustment voltage using the first frequency difference signal, wherein the first adjustment voltage is used to adjust the signal output frequency of the first device clock; The first monitored power device acquires first power data, where the first power data includes a first monitoring time and first monitoring data, where the first monitoring time is the device time output by the first device clock after being adjusted by the first regulating voltage; The first monitored power equipment sends the first power data to the data management platform, so that the data management platform determines the first operating state corresponding to the first power data according to a preset power data processing model, wherein the preset power data processing model is a decision classification model generated after training a preset neural network model using a preset power data training set; The determining a first adjustment voltage by using the first frequency difference signal includes: The first processor determines the first adjustment voltage according to a preset characteristic curve and the first frequency difference signal, wherein the preset characteristic curve is a voltage and frequency characteristic curve corresponding to the first device clock; After the first processor determines the first adjustment voltage according to the preset characteristic curve and the first frequency difference signal, the method further includes: The first monitored power device obtains a second device time signal through the first device clock, and receives a second satellite time signal through the first satellite signal receiver, wherein the first satellite time signal and the second satellite time signal are heterogeneous satellite signals; The first monitored power device determines a second frequency difference signal between the second device time signal and the second satellite time signal using the first processor, so as to determine a second regulated voltage using the second frequency difference signal; If the second regulation voltage is less than a preset minimum regulation voltage of the first device clock, adjusting the signal output frequency of the first device clock using the first regulation voltage; If the second regulation voltage is greater than the preset minimum regulation voltage of the first device clock, a third regulation voltage is used to adjust the signal output frequency of the first device clock, and the third regulation voltage is the sum of the first regulation voltage and the second regulation voltage.

2. The power data processing method based on satellite time synchronization according to claim 1 is characterized in that: Before the data management platform determines the first operating state corresponding to the first power data according to the preset power data processing model, the method further includes: Determine, by using a preset vector machine, a state decision boundary between a first normal operating data set and a first abnormal operating data set, wherein the amount of data in the first normal operating data set is greater than the amount of data in the first abnormal operating data set; Calculating a characteristic distance between the state decision boundary and each normal operation data in the first normal operation data set, and deleting normal operation data corresponding to a characteristic distance greater than a first preset distance threshold from the first normal operation data set to form a second normal operation data set; Calculating a characteristic distance between the state decision boundary and each abnormal operation data in the first abnormal operation data set, and deleting abnormal operation data corresponding to a characteristic distance greater than a second preset distance threshold from the first abnormal operation data set to form a second abnormal operation data set, wherein the first preset distance threshold is greater than the second preset distance threshold; The preset power data training set is determined according to the second normal operation data set and the second abnormal operation data set.

3. The power data processing method based on satellite time synchronization according to claim 1 is characterized in that: The first monitored power device determines a first frequency difference signal between the first device time signal and the first satellite time signal using a first processor, including: The first processor uses formula 1 and according to the first device time signal and the first satellite time signal Determine the first frequency difference signal ΔT k , the formula 1 is: Among them, T k is the standard time signal at time k, ω k is the random noise generated by the standard time signal in the communication link between the satellite and the first satellite signal receiver, ρ is the system frequency offset coefficient of the first device clock, is the system original deviation of the clock of the first device; The first processor calculates the first frequency difference signal ΔT according to a preset deviation estimation elimination algorithm. k Deviation elimination is performed to determine a first frequency difference update signal, wherein the first frequency difference update signal is used to replace the first frequency difference signal to determine the first adjustment voltage.

4. The power data processing method based on satellite time synchronization according to claim 2 is characterized in that: After determining the preset power data training set according to the second normal operation data set and the second abnormal operation data set, the method further includes: According to the preset power data training set D={d1,…,d i ,…,p n ,}Determine the power data input set P={p1,…,p i ,…,p n } and the running state output set Q = {q1,…,q i ,…,q n }, where p i is the ith power data input in the power data input set P, q i For p i The corresponding preset operating state output, n is the number of power data inputs in the preset power data training set D; According to the power data input set P = {p1, ..., p i ,…,p n } and the power data processing model to be trained determines the operating state estimation output set Q ′ = {q ′ 1,…,q ′ ′ ,…,q ′ n }; Using formula 2, determine the running state estimation output set Q ′ The corresponding first estimated deviation δ1 and second estimated deviation δ2, the formula 2 is: If the first estimated deviation δ1 is smaller than a preset first estimated deviation threshold and the second estimated deviation δ2 is smaller than a preset second estimated deviation threshold, the preset power data processing model is generated according to the trained power data processing model to be trained.

5. The power data processing method based on satellite time synchronization according to claim 4 is characterized in that: In determining the operating state estimation output set Q ′ The corresponding first estimated deviation δ1 and second estimated deviation δ2 also include: If the first estimated deviation δ1 is greater than the preset first estimated deviation threshold and / or the second estimated deviation δ2 is greater than the preset second estimated deviation threshold, then the error adjustment weight K of each layer in the power data processing model to be trained is determined using Formula 3 according to the first estimated deviation δ1 and the second estimated deviation δ2. Formula 3 is: Among them, r is the number of neural network layers of the power data processing model to be trained, and μ is the current learning rate of the power data processing model to be trained.

6. A power operation data management system, characterized in that: include: An electric power operation data management platform and a cluster of electric power equipment to be monitored, wherein each electric power equipment to be monitored in the cluster of electric power equipment to be monitored is communicatively connected to the electric power operation data management platform, and a processor, an equipment clock and a satellite signal receiver are provided in the electric power equipment to be monitored, and the equipment clock and the satellite signal receiver are communicatively connected to the processor respectively; The first power device to be monitored acquires a first device time signal through a first device clock, and receives a first satellite time signal through a first satellite signal receiver, wherein the first power device to be monitored is a power device to be monitored in the power device cluster to be monitored; The first monitored power device determines a first frequency difference signal between the first device time signal and the first satellite time signal using a first processor, so as to determine a first adjustment voltage using the first frequency difference signal, wherein the first adjustment voltage is used to adjust the signal output frequency of the first device clock; The first monitored power device acquires first power data, where the first power data includes a first monitoring time and first monitoring data, where the first monitoring time is the device time output by the first device clock after being adjusted by the first regulating voltage; The first monitored power equipment sends the first power data to the data management platform, so that the data management platform determines the first operating state corresponding to the first power data according to a preset power data processing model, wherein the preset power data processing model is a decision classification model generated after training a preset neural network model using a preset power data training set; The determining a first adjustment voltage by using the first frequency difference signal includes: The first processor determines the first adjustment voltage according to a preset characteristic curve and the first frequency difference signal, wherein the preset characteristic curve is a voltage and frequency characteristic curve corresponding to the first device clock; After the first processor determines the first adjustment voltage according to the preset characteristic curve and the first frequency difference signal, the method further includes: The first monitored power device obtains a second device time signal through the first device clock, and receives a second satellite time signal through the first satellite signal receiver, wherein the first satellite time signal and the second satellite time signal are heterogeneous satellite signals; The first monitored power device determines a second frequency difference signal between the second device time signal and the second satellite time signal using the first processor, so as to determine a second regulated voltage using the second frequency difference signal; If the second regulation voltage is less than a preset minimum regulation voltage of the first device clock, adjusting the signal output frequency of the first device clock using the first regulation voltage; If the second regulation voltage is greater than the preset minimum regulation voltage of the first device clock, a third regulation voltage is used to adjust the signal output frequency of the first device clock, and the third regulation voltage is the sum of the first regulation voltage and the second regulation voltage.

7. An electronic device, characterized in that: include: processor; as well as, A memory, configured to store executable instructions of the processor; The processor is configured to perform the method of any one of claims 1 to 5 by executing the executable instructions.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 5 when executed by a processor.

Images