A power data management method and system

Abstract

The application discloses a kind of electric power data management method and system, the method comprises: first, obtain historical data of power equipment, construct the mathematical model of historical data value and equipment life change;Then, according to the model, calculate the data value of different life stages, draw life change curve and mark key nodes and features;Curve is used as a reference, compared with real-time data, determine the warning level and data influence value;Combine multiple factors to generate electric power space situation estimation index;According to the index, determine the accuracy of the prediction node, analyze the curve fitting degree, and trace back the real cause of the equipment;The scheme quantifies the relationship between equipment data value and life by constructing a mathematical model, intuitively presents the trend, can timely warn life anomaly, comprehensively assesses the operation situation, traces back the failure cause, improves the management accuracy and efficiency, and ensures the stable operation of power system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system data management, and particularly relates to a power data management method and system. BACKGROUND

[0002] At present, the number of power equipment is increasing with the vigorous development of the power industry, and the amount of data generated during operation is growing explosively. These power data contain key information such as equipment state and operation efficiency, which is crucial to ensure the stable operation of the power system. However, the current power data management faces many challenges. On the one hand, the data sources are extensive and diverse in format, lacking unified standards, which makes data integration difficult and makes it difficult to achieve efficient use. On the other hand, traditional data management methods cannot accurately mine data value and cannot timely discover potential equipment failures and performance trends.

[0003] The current power equipment data processing and analysis has significant deficiencies. The existing technology adopts a static mode when processing power equipment data, without considering the impact of equipment aging on the reference value of historical data, which makes it difficult for the power space situation estimation index to accurately judge power abnormalities, cannot present the changes in equipment life in detail, and also cannot identify the time nodes of life changes and the corresponding data changes. At the same time, power data analysis is processed in chronological order, and the real cause of major failures (such as insulator flashover) occurs on the negative time axis before the fault occurs, which is difficult to trace back, and there are problems such as time inversion asymmetry, cause concealment, and multi-order correlation masking. SUMMARY

[0004] The present application provides a power data management method and system to more accurately assess equipment status, warn of risks, and accurately trace the real cause of equipment failure.

[0005] The present application provides a power data management method, which comprises:

[0006] S1, obtaining historical data of a single power equipment and establishing a mathematical model of historical data value and equipment life change;

[0007] S2, calculating the data value of the equipment at different life stages according to the model, drawing a life change curve and marking the decay time node and data typical feature;

[0008] S3, taking the life change curve as a reference curve, comparing it with the real-time collected data, and determining the current data feature, the equipment life warning level, and the corresponding data influence value;

[0009] S4, combining the data influence value and other influence indexes to generate a power space situation estimation index;

[0010] S5, according to the power space situation estimation index, determine the accuracy of the predicted time node, analyze the fitting degree of the reference curve and the current device life curve, and trace back the real cause of the device.

[0011] Preferably, S5, the real cause of the device includes: collecting the time data of the actual failure or significant performance decline of the device, calculating the accuracy index according to the predicted time node and the actual failure time; Align the time axis of the reference curve and the current device life curve, calculate the performance index difference of the reference curve and the current device life curve at the same time point, and evaluate the fitting degree of the two curves based on the size and distribution of the difference; According to the fitting degree of the predicted time node, use the back propagation algorithm to trace back, simulate the reverse process from the failure performance to the initial cause.

[0012] Preferably, the reference curve comprises: constructing a mathematical model describing the relationship between historical data value V and device life L ; Wherein V is the data value of the device at the life L; is the initial data value of the device; is the attenuation coefficient, is a random error term; With device life L as the horizontal axis and data value V as the vertical axis, draw the life change curve, and take the life change curve as the reference curve.

[0013] Preferably, in S5, the real cause of the device also includes:

[0014] S51, count the actual number of power equipment groups covered by the power system, and collect the data set of the area where each power equipment is located;

[0015] S52, establish a device wave function model describing the probability distribution of each device in the device group in different states;

[0016] S53, construct a density matrix based on the device wave function model, and quantify the quantum entanglement degree and uncertainty of the device group;

[0017] S54, design a quantum correlation decoder for the quantum entangled state of the device group, decode the quantum correlation information between the devices and construct a cooperative diagnosis model of the device group;

[0018] S55, using the cooperative diagnosis model of the device group, process and analyze the real-time collected data, and predict the cooperative failure risk of the device group.

[0019] Preferably, the S52, establishing a device wave function model describing the probability distribution of each device in the device group in different states, comprises: determining all states that each device in the device group can be in; according to the collected data, counting the number of occurrences of each state of the device, calculating the probability amplitude, and constructing a wave function Ψ describing the state of the device; verifying the constructed wave function by comparing it with the actual data to check whether it satisfies the normalization condition and whether it can reasonably describe the state distribution of the device.

[0020] Preferably, the S54 comprises: analyzing the quantum entangled state in which the device group is located based on the constructed device group density matrix; explicitly determining the causal quantum correlation information between the devices that need to be decoded, constructing a measurement basis according to the target correlation information; performing quantum measurement on the device group, projecting the state of the device group onto the measurement basis; analyzing the probability distribution of the measurement result to extract the quantum correlation information between the devices; collecting the classical operating characteristics of the device in addition to the quantum correlation information; and fusing the quantum correlation characteristics and the classical characteristics to form a comprehensive feature vector.

[0021] Preferably, the S51, collecting the data set of the area where each power device is located, further comprises: monitoring the conversion process between the quantum state and the classical state of the device group in real time, collecting data at the conversion instant of the quantum state and the classical state of the device group; identifying the critical state at the conversion instant to extract fault features specific to the boundary state; formulating a quantum-classical boundary state conversion control strategy based on the results of boundary state monitoring and identification; introducing boundary state characteristic parameters into the collaborative diagnosis model of the device group to monitor and warn the operating state of the device group in real time.

[0022] Preferably, the determination of the fault features specific to the boundary state comprises: pre-processing the collected data to remove noise and abnormalities; analyzing the change trend of the quantum state and classical state parameters, and considering that the device group is in the critical state at the conversion instant when the quantum state parameters and the classical state parameters change significantly at the same time; and extracting fault features specific to the boundary state from the data of the critical state.

[0023] Preferably, the real-time monitoring and warning of the operating state of the device group comprises: introducing the extracted boundary state characteristic parameters into the collaborative diagnosis model of the device group; retraining the updated model using historical data to adjust the parameters of the model; collecting the operating data of the device group in real time, extracting the boundary state characteristic parameters and the classical characteristic parameters, and forming a comprehensive feature vector; inputting the real-time feature vector into the updated collaborative diagnosis model, and outputting the operating state and fault risk prediction result of the device group from the model; and formulating a warning strategy according to the prediction result and the set risk threshold.

[0024] The application also provides a power data management method, and the system comprises:

[0025] an acquisition module configured to acquire historical data of the power equipment and establish a mathematical model of historical data value and equipment life change;

[0026] a classification module configured to calculate data value of the equipment at different life stages according to the model, draw a life change curve and mark a decay time node and data typicalization features;

[0027] an analysis module configured to compare the life change curve as a reference curve with real-time collected data, determine a current data feature, an equipment life warning level and a corresponding data influence value;

[0028] a calculation module configured to generate a power space situation estimation index in combination with the data influence value and other influence indexes;

[0029] a tracing module configured to determine accuracy of a predicted time node according to the power space situation estimation index, analyze a fitting degree of the reference curve and a current equipment life curve, and trace a real cause of the equipment in reverse.

[0030] One or more technical solutions provided in the application have at least the following technical effects or advantages:

[0031] The mathematical model of historical data value and equipment life is constructed to effectively solve the key technical problems in power equipment management. The traditional power equipment management cannot accurately quantify the change of equipment data value with life, cannot timely warn about the equipment life problems, and cannot comprehensively evaluate the power space situation and trace the equipment failure causes. The present application constructs the mathematical model to quantify the relationship between equipment data value and life, directly presents the value change trend, can timely warn about the equipment life abnormality, generates the situation estimation index in combination with multi-dimensional data to comprehensively evaluate the operation situation, and can also trace the real cause of the equipment in reverse. Therefore, the accuracy and efficiency of power equipment management are significantly improved, and the stable and reliable operation of the power system is ensured.

[0032] By counting the number of power equipment groups and collecting multi-dimensional data, a device wave function model is constructed to accurately describe the device state distribution, a density matrix is constructed to quantify the quantum entanglement degree and uncertainty of the equipment group, a quantum correlation decoder is designed to extract quantum correlation information between devices and construct a cooperative diagnosis model of the equipment group, and finally the model is used to predict the cooperative fault risk. This series of technical solutions realizes more comprehensive and in-depth state monitoring and fault prediction of the power equipment group, effectively improves the safety and reliability of the power system operation, and provides strong support for intelligent management of the power system.

[0033] By monitoring the quantum state and classical state conversion process of the equipment group in real time and collecting key data, the critical state at the conversion moment is accurately identified and the boundary state fault features are extracted, and based on this, an effective conversion control strategy is formulated, and the boundary state features are introduced into the collaborative diagnosis model. This scheme realizes the overall control of the quantum-electrostatic conversion process of the equipment group, can more timely discover potential fault risks, improves the stability and reliability of the equipment group operation, and provides strong and comprehensive support for the intelligent management and fault prevention of the equipment group. BRIEF DESCRIPTION OF DRAWINGS

[0034] Figure 1 A flowchart of a power data management method according to an embodiment of the present application is shown.

[0035] Figure 2 A block diagram of a power data management system according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0036] In order to facilitate the understanding of the present application, the present application will be described more fully below with reference to the accompanying drawings; the preferred embodiments of the present application are shown in the drawings, but the present application can be realized in many different forms and is not limited to the embodiments described herein; on the contrary, the purpose of providing these embodiments is to make the disclosure of the present application more thorough and comprehensive.

[0037] It should be noted that the terms "vertical", "horizontal", "up", "down", "left", "right" and similar expressions used herein are only for illustrative purposes and do not represent the only embodiment.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs; the terms used herein in the specification of the present application are only for the purpose of describing the specific embodiments and are not intended to limit the present application; the term "and / or" used herein includes any and all combinations of one or more related listed items.

[0039] Embodiment one: Figure 1 A flowchart of a power data management method according to an embodiment of the present application is shown.

[0040] As shown in Figure 1 A power data management method includes the following steps:

[0041] S1, the historical data of a single power equipment is acquired and a mathematical model of historical data value and equipment life change is established.

[0042] Specifically, the statistical power system covers the actual number of power equipment, and collects the data set of the area where each power equipment is located, including environmental monitoring data, basic operation state data, user power consumption data and historical power data. Analyze the influence of equipment aging on data value, consider the service life of equipment, maintenance records, environmental factors, etc., and build a mathematical model to describe the relationship between historical data value (V) and equipment life (L).

[0043] The mathematical model describing the relationship between historical data value (V) and equipment life (L) is as follows:

[0044]

[0045] V is the data value of the equipment at life L; is the initial data value of the equipment (i.e. the data value when the life is 0); is the decay coefficient, indicating the speed of data value decrease with the increase of life; is a random error term, which can usually be ignored or set to 0 in actual calculation (if the model is accurate enough and the data quality is high). Before calculation, the parameters and in the model need to be determined. and Use historical data to fit the model, that is, find a set of and

[0046] values that minimize the difference between the data value predicted by the model and the actual observed data value. Once the model parameters are determined, the data value of the equipment at different life stages can be calculated, and the life stage (such as 1 year, 5 years, 10 years, etc.) for which the data value needs to be calculated is determined. Replace each selected life stage L in the model formula to calculate the corresponding data value V.

[0047] Specifically, using the mathematical model established in S1, input different values of equipment life (L) to calculate the corresponding data value (V), and draw the life change curve with equipment life (L) as the horizontal axis and data value (V) as the vertical axis. Set the data value decay threshold, and mark the corresponding life time as the decay time node when the data value is lower than the threshold. Observe and record the typical characteristics of data in the decay process (such as increased data fluctuations and increased abnormal values).

[0048] Among them, the decay time node represents the time point when the data value of the equipment starts to decrease significantly, indicating the aging or performance decline of the equipment.

[0049] S3, compare the life change curve as a reference curve with the real-time collected data to determine the current data characteristics, the device life warning level and the corresponding data influence value.

[0050] Specifically, based on the constructed model, a device life change curve, i.e., an expected trend of device performance indicators changing over time, is generated. The real-time collected data is aligned with the corresponding time point on the device life change curve, the difference between the real-time data and the reference curve is calculated, the characteristics related to the device life are extracted from the real-time data, and compared with the characteristics on the reference curve. According to the size of the difference and the change trend of the characteristics, the device life warning level is divided, and based on the difference and the change of the characteristics, the data influence value is calculated to quantify the influence degree of the real-time data on the device life.

[0051] wherein the data influence value calculation formula is:

[0052]

[0053] , is a weight coefficient, is the difference between the real-time data and the reference curve, and ΔF is the change amount of the characteristics, is the remaining life of the device (estimated from the reference curve).

[0054] S4, generate a power space situation estimation index combining the data influence value and other influence indexes.

[0055] wherein the other influence indexes include: calculating a predicted environmental influence index according to environmental monitoring data, calculating a basic operation state index according to basic operation state data (which can be calculated according to the deviation of the device operation state from the ideal state), and calculating a predicted power load influence index according to user power consumption data and historical power data (predicted using a time series prediction model, and calculated based on the difference between the predicted value and the historical average value).

[0056] Specifically, data related to the operating environment of power equipment, such as temperature, humidity, air quality, wind speed, etc., are collected, and environmental monitoring data are processed to extract key features such as extreme weather event frequency, environmental parameter change rate, etc. Using machine learning algorithms, an environmental impact prediction model is constructed, and based on current and historical environmental data, a predicted environmental impact index is calculated. Collecting basic operating state data of power equipment, such as voltage, current, power factor, equipment temperature, etc., according to the operating specifications provided by the equipment manufacturer or industry standards, the current operating state of the equipment is evaluated. Based on the evaluation results, a basic operating state index is calculated to reflect the health status and operating efficiency of the equipment. Collecting user electricity consumption data and historical power load data, using time series analysis to predict future power load, based on the prediction results, a predicted power load impact index is calculated to reflect the potential impact of future power load on power space situation. According to the influence degree of each index on power space situation, the data impact value, predicted environmental impact index, basic operating state index and predicted power load impact index are assigned weights.

[0057] wherein the power space situation estimation index can be calculated by weighted average of each index:

[0058]

[0059] and is the weight of each index, is the predicted environmental impact index, is the basic operating state index, is the predicted power load impact index.

[0060] S5, according to the power space situation estimation index, determine the accuracy of the predicted time node, analyze the fit degree of the reference curve and the current equipment life curve, and trace back the real cause of the equipment.

[0061] Specifically, collect the time data of actual equipment failure or significant performance decline, calculate the accuracy index according to the predicted time node and the actual failure time. Align the time axis of the reference curve and the current equipment life curve, calculate the performance index difference of the reference curve and the current equipment life curve at the same time point, and evaluate the fit degree of the two curves based on the size and distribution of the difference. According to the fit degree of the predicted time node, the back propagation algorithm is used for back tracing, and the reverse process from failure performance to initial cause (time reversal algorithm needs to consider time reversal asymmetry, i.e. information entropy decay from failure performance (t=0) to cause (t=-Δt)).

[0062] ​​In the time travel process, the time travel algorithm should construct a multi-stage correlation model, reveal the internal relationship between stages through time series analysis and causal relationship inference, restore the complete fault chain, identify and handle the hidden cause problems (such as: initial micro-defects (such as 0.1 mm cracks) may appear as normal noise in conventional monitoring. To this end, the algorithm should use high-sensitivity detection technology, combine historical data and real-time data, and use pattern recognition and anomaly detection methods to mine potential causes) and multi-stage correlation hidden problems (such as: bird droppings pollution → partial discharge → insulation aging → chain reaction of flashover may be broken by discrete data points). According to the results of the time travel algorithm, determine the real cause of the current device, which may include environmental factors (such as bird droppings pollution), equipment aging, improper operation, etc. Generate an evaluation report including device name, location, cause type, cause occurrence time and potential risks.

[0063] The technical solutions in the embodiments of the application have at least the following technical effects or advantages:

[0064] By constructing a mathematical model of historical data value and equipment life, the key technical problems in power equipment management are effectively solved. Traditional power equipment management cannot accurately quantify the change of equipment data value with life, cannot timely warn of equipment life problems, and cannot comprehensively evaluate the power space situation and trace the equipment failure cause. The present scheme constructs a mathematical model to quantify the relationship between equipment data value and life, intuitively presents the value change trend, can timely warn of equipment life anomalies; generates a situation estimation index by comprehensively analyzing multi-dimensional data, and comprehensively evaluates the operation situation; and can also trace the real cause of the equipment. Therefore, the accuracy and efficiency of power equipment management are significantly improved, and the stable and reliable operation of the power system is ensured.

[0065] Embodiment two: In the existing power data analysis and fault prediction technology, the main focus is on the fault cause tracing of single equipment, but in the actual operation of the power grid, multiple devices may exhibit coordinated failure characteristics due to being in a quantum entangled state (such as sharing an electromagnetic field, mechanical resonance). At this time, the fault cause tracing method of single equipment fails to accurately predict and diagnose the coordinated failure of the equipment group. To solve this problem, the embodiment of the present application is optimized based on the above embodiment.

[0066] In some embodiments, in step S5, the real cause of the equipment is also traced back, which includes:

[0067] S51, the actual number of power equipment groups covered by the power system is counted, and a data set of the area where each power equipment is located is collected.

[0068] The data set includes: environmental monitoring data, basic operation state data, user power consumption data, and historical power data.

[0069] S52, a device wave function model describing the probability distribution of each device in the device group in different states is established.

[0070] Specifically, each device in the device group is analyzed in detail, and it is determined that it can be in all states. For example, for a transformer device, the possible states include normal operation state, overload state, local overheating state, insulation aging state, fault shutdown state, etc. The determined device states are encoded for subsequent digital processing. Assuming that the device has n possible states, it is denoted as , where (j=1, 2, , n) represents the jth state of the device.

[0071] According to the collected data, the number of times the device is in each state is counted. Assuming that in a period of time, the device is in state times, the total number of device state transitions is M ( ), and the probability amplitude of the device in state is The calculation formula is: The probability of the device being in a certain state is proportional to the frequency of the state appearing in the historical data based on the frequency.

[0072] According to the calculated probability amplitude , the wave function Ψ describing the device state is constructed. For each device in the device group, its wave function Ψ can be expressed as a linear combination of the device being in various states, that is: , where is the probability amplitude of the device being in state , and satisfies the normalization condition . The physical meaning of the normalization condition is that the device must be in one of the n possible states, and the sum of the probabilities of the device being in various states is 1.

[0073] The constructed wave function is verified by comparing with the actual data to check whether it satisfies the normalization condition and whether it can reasonably describe the state distribution of the device.

[0074] S53, a density matrix is constructed based on the device wave function model to quantify the quantum entanglement degree and uncertainty of the device group.

[0075] The quantum entanglement degree of the device group is quantified using entanglement entropy. The device group is divided into two subsystems A and B. For example, a part of the devices can be assigned to subsystem A, and the remaining devices can be assigned to subsystem B. The trace of subsystem B is taken to obtain the reduced density matrix of subsystem A. The trace operation is to calculate the entanglement entropy after summing all possible states of subsystem B. For details of the entanglement entropy calculation process, refer to the prior art, which will not be described herein. The uncertainty of the device group is quantified using von Neumann entropy. For details of the calculation process, refer to the prior art, which will not be described herein.

[0076] S54, a quantum correlation decoder is designed for the quantum entangled state of the device group to decode the quantum correlation information between the devices and construct a cooperative diagnosis model of the device group.

[0077] Specifically, the entangled state is determined. Based on the constructed device group density matrix, the quantum entangled state in which the device group is located is analyzed. For example, if the device group consists of two subsystems A and B, the entangled state can be represented as wherein and are the states of subsystems A and B, respectively, is a complex coefficient satisfying the normalization condition .

[0078] A decoding measurement basis is designed. The causal quantum correlation information between the devices that needs to be decoded is determined, and the measurement basis is constructed according to the target correlation information. For a two-device entangled state , the measurement basis can be represented as wherein and are the measurement basis vectors of subsystems A and B, respectively. The measurement basis should be selected to maximize the extraction of the correlation information in the entangled state.

[0079] Quantum measurement is performed. Quantum measurement is performed on the device group to project the state of the device group onto the measurement basis. The measurement result appears in the form of probability. The probability of measuring that subsystem A is in state and subsystem B is in state is . Multiple measurements are performed to statistically analyze the distribution of the measurement results to obtain more accurate quantum correlation information.

[0080] Key information extraction. The probability distribution of the measurement results is analyzed to extract the quantum correlation information between the devices. For example, the correlation degree between the devices is quantified by calculating the mutual information I(A;B)=H(A)+H(B)-H(A,B) between the measurement results, wherein H(A), H(B), and H(A,B) are the Shannon entropies of subsystem A, subsystem B, and device group AB, respectively. The formula for calculating the Shannon entropy is H(X)=- , and P(x) is the probability of system X being in state x.

[0081] A device group collaborative diagnosis model is constructed. In addition to quantum correlation information, classical operating characteristics of the device, such as voltage, current, temperature, etc., are collected. The quantum correlation characteristics and the classical characteristics are fused to form a comprehensive feature vector. Real-time operating data of the device are collected, quantum correlation characteristics and classical characteristics are extracted, and a comprehensive feature vector is formed. The real-time feature vector is input into the trained collaborative diagnosis model for real-time diagnosis, and a diagnosis result of the device is obtained.

[0082] S55, using the device group collaborative diagnosis model, processing and analyzing the data collected in real time, predicting the collaborative fault risk of the device group.

[0083] Specifically, if the device group has quantum entanglement characteristics, the quantum state of the device group is estimated using quantum state tomography and other technologies based on the data collected in real time. For example, for a two-device entangled state, the density matrix ρ of the quantum state can be reconstructed by measuring the results in different basis vectors multiple times and using maximum likelihood estimation and other methods. Based on the estimated quantum state, the quantum correlation characteristics between the devices are calculated. The preprocessed classical characteristics and the extracted quantum correlation characteristics are fused to form a comprehensive feature vector wherein is a classical feature vector, is a quantum correlation feature vector. The real-time extracted comprehensive feature vector is input into the collaborative diagnosis model, and the model calculates based on the input feature vector to output the collaborative fault risk prediction result of the device group. According to actual requirements and historical data, a threshold of the collaborative fault risk is set. For example, when the probability of the device group being in a serious fault state is greater than 0.7, it is considered that there is a high collaborative fault risk. According to the prediction result and the risk threshold, a corresponding decision is made.

[0084] The technical solutions in the embodiments of the present application have at least the following technical effects or advantages:

[0085] By counting the number of power device groups and collecting multi-dimensional data, a device wave function model is constructed to accurately describe the device state distribution. Based on the model, a density matrix is constructed to quantify the quantum entanglement degree and uncertainty of the device group. A quantum correlation decoder is designed to extract quantum correlation information between devices and construct a device group collaborative diagnosis model. Finally, the model is used to predict the collaborative fault risk. This series of technical solutions realizes more comprehensive and in-depth state monitoring and fault prediction of the power device group, effectively improves the safety and reliability of the power system operation, and provides strong support for intelligent management of the power system.

[0086] Embodiment three: In a hybrid quantum-classical system, the power equipment group will generate boundary state information loss when converting between quantum states and classical states. Existing technologies cannot effectively capture the critical state at the moment of conversion, resulting in quantum correlation collapse during classical conversion, high information loss rate, boundary state failure characteristics being masked, and the failure of the collaborative prediction model in the conversion zone. To solve these problems, the embodiments of the present application make certain optimizations based on the above embodiments.

[0087] In some embodiments, the data set of the area where each power equipment is located collected in step S51 further includes:

[0088] 1A, real-time monitoring of the conversion process of the equipment group between quantum states and classical states, data acquisition at the moment of quantum state and classical state conversion of the equipment group.

[0089] Specifically, sensors capable of monitoring quantum states and classical states simultaneously are installed in the equipment group. For quantum state monitoring, special equipment such as quantum bit detectors is used; for classical state monitoring, conventional voltage, current, temperature, etc. sensors are used. The signals collected by the sensors are digitized and transmitted to the data processing center. Continuously monitor the state of the equipment group, and when it is detected that the equipment group may be in the conversion region of quantum state and classical state, increase the data acquisition frequency. For example, increase the data acquisition from once every second to once every millisecond. Record the timestamp, quantum state parameters (such as the spin state of the quantum bit, entanglement degree, etc. if measurable) and classical state parameters (such as voltage, current, temperature, etc.) at the moment of conversion.

[0090] 1B, identifying the critical state at the moment of conversion to extract the fault characteristics unique to the boundary state.

[0091] Specifically, the collected data is preprocessed to remove noise and outliers. For example, moving average filtering is used to smooth the classical state data. Analyze the change trend of quantum state and classical state parameters, when quantum state parameters and classical state parameters change significantly at the same time, consider that the equipment group is in the critical state at the moment of conversion. Extract the fault characteristics unique to the boundary state from the data of the critical state. For example, extract the change rate of quantum state parameters and the mutation amplitude of classical state parameters (such as the maximum mutation amplitude of current).

[0092] 1C, based on the results of boundary state monitoring and identification, develop a quantum-classical boundary state conversion control strategy.

[0093] Specifically, according to the extracted boundary state characteristics, the stability and reliability of the device group in the quantum state and classical state conversion process are analyzed. For example, if the change rate of quantum state entanglement entropy is too large, it may cause the device group to be unstable. According to the analysis result, a control strategy is formulated. For example, when the change rate of quantum state entanglement entropy exceeds a certain threshold α, the operating parameters of the device group are adjusted, such as reducing the input power, changing the working frequency, etc.

[0094] 1D, the boundary state characteristic parameters are introduced into the device group collaborative diagnosis model, and the running state of the device group is monitored and warned in real time.

[0095] Specifically, the extracted boundary state characteristic parameters are introduced into the device group collaborative diagnosis model. For example, in the neural network model, these characteristic parameters are used as new input nodes. The updated model is retrained using historical data to adjust the parameters of the model. The running data of the device group is collected in real time, the boundary state characteristic parameters and the classical characteristic parameters are extracted, and a comprehensive feature vector is formed. The real-time feature vector is input into the updated collaborative diagnosis model, and the model outputs the running state of the device group and the fault risk prediction result. According to the prediction result and the set risk threshold, a warning strategy is formulated. For example, when the probability of the device group being in a serious fault state is greater than a certain threshold β, an early warning signal is sent to inform relevant personnel to take measures.

[0096] The technical solutions in the embodiments of the application have at least the following technical effects or advantages:

[0097] By monitoring the quantum state and classical state conversion process of the device group in real time and collecting key data, the critical state at the conversion moment is accurately identified and the boundary state fault characteristics are extracted, and based on this, an effective conversion control strategy is formulated, and the boundary state characteristics are introduced into the collaborative diagnosis model. This scheme realizes the comprehensive control of the quantum-electrostatic conversion process of the device group, can more timely discover potential fault risks, improves the stability and reliability of the device group operation, and provides strong and comprehensive support for the intelligent management and fault prevention of the device group.

[0098] Further, the embodiment of the application also provides an electric power data management system.

[0099] Figure 2 is a structural schematic diagram of an electric power data management system according to the embodiment of the application.

[0100] As shown in Figure 2 , an electric power data management system comprises an acquisition module, a classification module, an analysis module, a calculation module and a tracing module.

[0101] The acquisition module is used to acquire the historical data of the electric power equipment and establish a mathematical model of the historical data value and the change of the equipment life;

[0102] a classification module for calculating the data value of the model computing device at different life stages, drawing a life change curve and marking the decay time node and data typicalization features;

[0103] an analysis module for comparing the life change curve as a reference curve with the real-time collected data to determine the current data features, the device life warning level and the corresponding data influence value;

[0104] a calculation module for combining the data influence value and other influence indexes to generate a power space situation estimation index;

[0105] a tracing module for determining the accuracy of the predicted time node according to the power space situation estimation index, analyzing the fitting degree of the reference curve and the current device life curve, and reversely tracing the real cause of the device.

[0106] It should be noted that other specific implementation contents of the power data management system of the embodiment of the present application can refer to the power data management method described above.

[0107] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A power data management method, characterized in that, The method includes: S1, acquire historical data of a single power device and establish a mathematical model of the value of historical data and changes in the lifespan of the device; among which, the data for establishing the value of historical data in the digital model includes environmental monitoring data, basic operating status data, user electricity consumption data, and historical power data of the area where each power device is located; S2, calculate the data value of the device at different life stages based on the model, plot the life change curve and mark the decay time nodes and typical data characteristics; wherein, the decay time nodes are obtained by setting a data value decay threshold, and marking the corresponding life time when the data value is lower than the threshold. S3, based on the constructed model, generates an equipment lifespan curve; aligns the real-time collected data with the corresponding time points on the equipment lifespan curve, calculates the difference between the real-time data and the reference curve, extracts features related to equipment lifespan from the real-time data, and compares them with the features on the reference curve; classifies the equipment lifespan warning level according to the magnitude of the difference and the trend of feature changes; calculates the data impact value based on the difference and feature changes, quantifying the degree of impact of real-time data on equipment lifespan; the formula for calculating the data impact value is: , , These are weighting coefficients. It represents the difference between real-time data and a reference curve; ΔF is the change in the feature. It is the estimated remaining lifespan of the equipment based on the reference curve; S4, combining data impact values ​​and other impact indices to generate a power spatial situation prediction index; among which, other impact indices include the predicted environmental impact index calculated from environmental monitoring data, the basic operating status index calculated from basic operating status data, and the predicted power load impact index calculated from user electricity consumption data and historical power data, through calculation. Generate a power spatial situation prediction index, among which, , , and It is the weight of each index. It is a predictive environmental impact index. It is a basic operating status index. It is a predictive index of the impact of electricity load; S5 determines the accuracy of the predicted time nodes based on the power spatial situation prediction index, analyzes the fit between the reference curve and the current equipment life curve, and traces back to the real cause of the equipment failure; among them, by collecting the time data of actual equipment failure or significant performance degradation, the accuracy index is calculated based on the predicted time nodes and the actual failure time.

2. The power data management method as described in claim 1, characterized in that, S5, the reverse tracing of the true cause of the device includes: collecting time data of the actual occurrence of device failure or significant performance degradation, calculating accuracy indicators based on the predicted time nodes and the actual failure time; aligning the time axes of the reference curve and the current device life curve, calculating the performance index difference between the reference curve and the current device life curve at the same time point, evaluating the fit between the two curves based on the magnitude and distribution of the difference; and using the backpropagation algorithm to perform reverse tracing based on the fit of the predicted time nodes, simulating the reverse process from the fault manifestation to the initial cause.

3. The power data management method as described in claim 2, characterized in that, The reference curve includes: constructing a mathematical model describing the relationship between historical data value V and equipment lifespan L. Where V is the data value of the device during its lifespan L; It is the initial data value of the equipment; It is the attenuation coefficient. It is a random error term; plot the lifespan curve with equipment lifespan L as the horizontal axis and data value V as the vertical axis, and use the lifespan curve as a reference curve.

4. The power data management method as described in claim 1, characterized in that, In S5, the real motivations for reverse tracing devices also include: S51, count the actual number of power equipment groups covered by the power system and collect data sets of the area where each power equipment is located; S52, Establish a device wave function model that describes the probability distribution of each device in the device group under different states; S53, based on the device wavefunction model, constructs a density matrix to quantify the degree of quantum entanglement and uncertainty of device groups; S54, Design a quantum correlation decoder for quantum entangled states in a device group, decode the quantum correlation information between devices and build a collaborative diagnostic model for the device group; S55 utilizes a collaborative diagnostic model for equipment groups to process and analyze real-time collected data, predicting collaborative failure risks within the equipment group.

5. The power data management method as described in claim 4, characterized in that, S52 establishes a device wave function model describing the probability distribution of each device in the device group under different states, including: determining all possible states that each device in the device group can be in; based on the collected data, counting the number of times the device is in each state, calculating the probability amplitude, and constructing a wave function Ψ describing the device state; verifying the constructed wave function by comparing it with actual data to check whether it meets the normalization condition and whether it can reasonably describe the state distribution of the device.

6. The power data management method as described in claim 4, characterized in that, S54 includes: analyzing the quantum entangled state of the device group based on the constructed device group density matrix; identifying the quantum correlation information of causality between devices that needs to be decoded, and constructing a measurement basis according to the target correlation information; performing quantum measurement on the device group and projecting the state of the device group onto the measurement basis; analyzing the probability distribution of the measurement results and extracting the quantum correlation information between devices; collecting classical operating characteristics of the devices in addition to the quantum correlation information; and fusing the quantum correlation characteristics and classical characteristics to form a comprehensive feature vector.

7. The power data management method as described in claim 4, characterized in that, S51, the data set collected for the area where each power device is located, also includes: real-time monitoring of the transition process between quantum and classical states of the device group, data collection at the moment of transition between quantum and classical states; identifying the critical state at the moment of transition, extracting features to determine the fault characteristics unique to the boundary state; formulating a quantum-classical boundary state transition control strategy based on the results of boundary state monitoring and identification; introducing boundary state feature parameters into the device group collaborative diagnosis model, and real-time monitoring and early warning of the operating status of the device group.

8. The power data management method as described in claim 7, characterized in that, The process of determining the fault characteristics specific to the boundary state includes: preprocessing the collected data to remove noise and anomalies; analyzing the changing trends of quantum state and classical state parameters, and considering the equipment group to be in a critical state at the moment of transition when both quantum state and classical state parameters change significantly; and extracting the fault characteristics specific to the boundary state from the data of the critical state.

9. A power data management method as described in claim 7, characterized in that, The real-time monitoring and early warning of the operating status of the equipment group specifically includes: introducing the extracted boundary state feature parameters into the equipment group collaborative diagnosis model; retraining the updated model using historical data and adjusting the model parameters; collecting the operating data of the equipment group in real time, extracting boundary state feature parameters and classical feature parameters to form a comprehensive feature vector; inputting the real-time feature vector into the updated collaborative diagnosis model, and the model outputs the operating status and fault risk prediction results of the equipment group; and formulating early warning strategies based on the prediction results and the set risk thresholds.

10. A power data management system, applied to a power data management method as described in any one of claims 1 to 9, characterized in that, The system includes: The acquisition module is used to acquire historical data of power equipment and establish a mathematical model of the relationship between the value of historical data and changes in equipment lifespan. The classification module is used to calculate the data value of the device at different life stages based on the model, draw the life change curve, and mark the decay time nodes and typical data characteristics. The analysis module uses the life change curve as a reference curve and compares it with the real-time collected data to determine the impact of the current data characteristics on the equipment life warning level and the corresponding data impact value. Calculation module: Combines data impact values ​​and other impact indices to generate a power spatial situation prediction index; The tracing module determines the accuracy of the predicted time points based on the power spatial situation prediction index, analyzes the fit between the reference curve and the current equipment life curve, and traces back to the real cause of the equipment's failure.

Images