Method, device and equipment for predicting power failure range of low-voltage power grid, medium and product

Abstract

The invention provides a prediction method, device and equipment for a power failure range of a low-voltage power grid, a medium and a product. The method comprises the following steps: acquiring a low-voltage power grid operation variable after disturbance is added into a low-voltage power grid and a variable reference value during normal operation of the low-voltage power grid; determining a time sequence data set of each component in the low-voltage power grid operation variables according to the low-voltage power grid operation variables; according to the time sequence data set of each component and the prediction register, performing prediction processing on the low-voltage power grid operation variables to obtain a predicted operation variable sequence; determining a component power failure prediction index corresponding to each component according to the prediction operation variable sequence; according to the variable reference value during normal operation, the component power failure prediction index corresponding to each component is verified to obtain the low-voltage power failure prediction index, and the accuracy of predicting the power failure range of the low-voltage power grid is improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology for power systems, and in particular to a method, device, equipment, medium, and product for predicting the range of a low-voltage power grid outage. Background Art

[0002] Low-voltage outage scope prediction is a core component of intelligent grid operation and maintenance. By accurately locating faults, optimizing resource allocation, and improving collaborative efficiency, it provides solid support for protecting core grid equipment, improving emergency repair efficiency, and ensuring power supply reliability. With continuous technological advancements, outage scope prediction will become even more accurate and efficient, laying a solid foundation for building a robust smart grid.

[0003] Existing low-voltage power outage scope prediction methods usually use a machine learning method with a hybrid algorithm architecture to predict the scope of low-voltage power outages. This prediction method can integrate multimodal data and overcome the limitations of a single model in modeling complex correlations.

[0004] However, the hybrid algorithm architecture has complex parameter adjustment and inaccurate data sample division, which reduces the accuracy of low-voltage power outage range prediction. Summary of the Invention

[0005] The present application provides a method, device, equipment, medium and product for predicting the range of low-voltage power outages in a low-voltage power grid, so as to solve the technical problems in the prior art of complex testing process and low prediction accuracy when predicting the range of low-voltage power outages in a low-voltage power grid.

[0006] In a first aspect, the present application provides a method for predicting the range of a low-voltage power outage, comprising:

[0007] Collect the low-voltage grid operation variables after the disturbance is added and the variable baseline values ​​during normal operation;

[0008] Determining, according to the low-voltage power grid operation variable, a time series data set of each component in the low-voltage power grid operation variable;

[0009] Performing prediction processing on the low-voltage power grid operating variables according to the time series data sets and prediction registers of the various components to obtain a predicted operating variable sequence;

[0010] determining, based on the predicted operating variable sequence, a component power outage prediction index corresponding to each component;

[0011] According to the variable reference value during normal operation, the component power outage prediction index corresponding to each component is verified to obtain a low-voltage power outage prediction index, which is used to predict the power outage range of the low-voltage power grid.

[0012] Furthermore, the predicting process is performed on the low-voltage power grid operation variables according to the time series data sets and prediction registers of the various components to obtain a predicted operation variable sequence, including:

[0013] Determining, based on the time series data sets of the respective components, a data training set corresponding to the time series data sets of the respective components;

[0014] Updating each of the data training sets to a prediction register respectively to obtain an updated prediction register corresponding to each of the components;

[0015] According to the updated prediction register, the low-voltage power grid operation variables are predicted to obtain a predicted operation variable sequence.

[0016] Furthermore, the updating of each of the data training sets to a prediction register to obtain an updated prediction register corresponding to each of the components includes:

[0017] Updating each of the data training sets to the prediction registers respectively, so as to update the register parameters in each of the prediction registers and obtain updated register parameters corresponding to each of the prediction registers;

[0018] According to each of the update register parameters, update processing is performed on each of the prediction registers to obtain an updated prediction register corresponding to each of the components.

[0019] Furthermore, according to the updated prediction register, the low-voltage power grid operation variable is predicted to obtain a predicted operation variable sequence of the low-voltage power grid operation variable, including:

[0020] Determining, according to the update prediction registers, the prediction sequence time and each component prediction value corresponding to each of the update prediction registers;

[0021] According to the prediction sequence time, the predicted values ​​of the various components are combined and processed to obtain a predicted operating variable sequence of the low-voltage power grid operating variables.

[0022] Furthermore, determining the component outage prediction index corresponding to each component according to the predicted operating variable sequence includes:

[0023] Establishing an operating variable determination sequence of the low-voltage power grid operating variables according to the predicted operating variable sequence;

[0024] performing classification processing on the operating variable determination sequence to obtain a component determination sequence corresponding to each component;

[0025] The component power outage prediction index corresponding to each component is determined according to the component determination sequence and historical power outage data information of each component, wherein the historical power outage data information is power outage data information of the low-voltage power grid during historical operation.

[0026] Furthermore, the component power outage prediction index corresponding to each component is verified based on the variable reference value during normal operation to obtain a low-voltage power outage prediction index, including:

[0027] Calculating the total offset distance between each component and the variable reference value according to the variable reference value during normal operation;

[0028] performing an offset clustering process on each component according to the total offset distance of each component to obtain offset status information of each component;

[0029] performing verification processing on the component power outage prediction index corresponding to each component according to the offset state information of each component to obtain a verification result of the component power outage prediction index corresponding to each component;

[0030] If the verification result meets the preset prediction index requirement, weighting the component power outage prediction indexes to obtain a low-voltage power outage prediction index;

[0031] The low-voltage power outage prediction index is used to predict the power outage scope of the low-voltage power grid.

[0032] Furthermore, the verifying process of the component power outage prediction index corresponding to each component according to the offset state information of each component to obtain the verification result of the component power outage prediction index corresponding to each component includes:

[0033] Determining power outage probability information corresponding to each component according to each of the offset state information and historical power outage data information of each component;

[0034] Comparing the power outage probability information with the component power outage prediction index corresponding to each component to verify the component power outage prediction index corresponding to each component, and obtaining a verification result of the component power outage prediction index corresponding to each component;

[0035] Among them, the verification results include meeting the preset prediction index requirements and not meeting the preset prediction index requirements, and the preset prediction index requirements are that the comparison difference between the power outage probability information and the component power outage prediction index corresponding to each classification is within a preset threshold range.

[0036] In a second aspect, the present application provides a device for predicting the range of a low-voltage power outage, comprising:

[0037] An acquisition module is used to acquire the operating variables of the low-voltage power grid after a disturbance is added and the reference values ​​of the variables during normal operation;

[0038] A time series data set determination module, configured to determine a time series data set of each component in the low-voltage power grid operation variable according to the low-voltage power grid operation variable;

[0039] An operating variable sequence prediction module is used to predict the operating variables of the low-voltage power grid according to the time series data sets of each component and the prediction register to obtain a predicted operating variable sequence;

[0040] a component power outage prediction index determination module, configured to determine a component power outage prediction index corresponding to each component according to the predicted operation variable sequence;

[0041] The low-voltage power outage prediction index obtaining module is used to verify the component power outage prediction index corresponding to each component according to the variable reference value during normal operation to obtain the low-voltage power outage prediction index.

[0042] In a third aspect, the present application provides an electronic device, comprising: a processor, and a memory communicatively connected to the processor;

[0043] The memory stores computer-executable instructions;

[0044] The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of the first aspects.

[0045] In a fourth aspect, the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the method as described in any one of the first aspects.

[0046] In a fifth aspect, the present application provides a computer program product, comprising a computer program, which implements the method as described in any one of the first aspects when executed by a processor.

[0047] The present application provides a method, device, equipment, medium and product for predicting the power outage range of a low-voltage power grid, including collecting the low-voltage power grid operating variables after the low-voltage power grid is disturbed and the variable baseline values ​​during normal operation; determining the time series data set of each component in the low-voltage power grid operating variables based on the low-voltage power grid operating variables; predicting the low-voltage power grid operating variables based on the time series data set and prediction registers of each component to obtain a predicted operating variable sequence; determining the component power outage prediction index corresponding to each component based on the predicted operating variable sequence; verifying the component power outage prediction index corresponding to each component based on the variable baseline value during normal operation to obtain a low-voltage power outage prediction index, wherein the low-voltage power outage prediction index is used to predict the power outage range of the low-voltage power grid, thereby improving the accuracy of predicting the power outage range of the low-voltage power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] 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.

[0049] Figure 1 This is a flow chart of Example 1 of the method for predicting the range of a low-voltage power outage in the present application;

[0050] Figure 2 This is a flow chart of Example 2 of the method for predicting the range of low-voltage power outages in the present application;

[0051] Figure 3 This is a flow chart of Example 3 of the method for predicting the range of low-voltage power outages in the present application;

[0052] Figure 4 This is a flow chart of a fourth embodiment of the method for predicting the range of a low-voltage power outage in the present application;

[0053] Figure 5 This is a flow chart of Example 5 of the method for predicting the range of low-voltage power outages in the present application;

[0054] Figure 6 A simulation diagram of the predicted running variable sequence provided by this application;

[0055] Figure 7 This is a schematic diagram of the structure of the device for predicting the range of low-voltage power outages proposed in this application;

[0056] Figure 8 This is a schematic diagram of the structure of the electronic device proposed in this application.

[0057] The above drawings illustrate specific embodiments of the present application, which will be described in more detail below. These drawings and the textual description are not intended to limit the scope of the present application in any way, but rather to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0058] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0059] Existing low-voltage power grid outage prediction technology mainly relies on fixed thresholds and single-variable analysis, which makes it difficult to accurately identify dynamic disturbances and multi-variable coupled faults. It also lacks in-depth correlation analysis of historical power outage events and the time series characteristics of operating variables, making it difficult to predict new complex faults, resulting in a high false alarm rate and insufficient ability to predict new faults.

[0060] To address the above technical issues, this solution matches dynamic multi-dimensional evaluation, cross-variable collaborative prediction and historical data, and extracts the time series features of operating variables based on prediction registers. At the same time, it combines multi-dimensional indicators such as amplitude, trend, and duration to generate a judgment sequence, significantly improving the accuracy and real-time performance of power outage predictions and adapting to the complex operating conditions of new power grids.

[0061] The following specific embodiments describe in detail the technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0062] Figure 1 This is a flow chart of the first embodiment of the method for predicting the range of low-voltage power outages proposed in this application. Figure 1 As shown, including:

[0063] S101. Collecting operating variables of the low-voltage power grid after a disturbance is added and reference values ​​of the variables during normal operation.

[0064] The low-voltage power grid operating variables collected after the disturbance in the embodiment of the present application include the bus voltage amplitude, bus voltage phase, active power flow on each branch line, and reactive power flow on each branch line of the low-voltage power grid. Among them, the bus voltage replication can reflect the voltage level of the power grid and can detect voltage sag or surge. The bus voltage phase can be used to analyze the synchronization state of the power grid and then determine whether the phase is out of step after the disturbance is added. The active power flow on each branch line can reflect the actual transmission power of the power grid and can be used to identify overload or reverse power flow. The reactive power flow on each branch line can be used to evaluate the stability of the power grid, such as the voltage support capacity.

[0065] In this step, by monitoring the dynamic response of the power grid after the disturbance (such as voltage fluctuations and power oscillations), the prediction method can be adapted to different fault scenarios (such as short circuits and sudden load changes). In addition, combining historical disturbance data can optimize the prediction algorithm and reduce the error rate (for example, distinguishing between transient disturbances and persistent faults).

[0066] The normal operating variable baseline value serves as a reference for comparing post-disturbance data changes, improving forecast reliability. For example, after predicting the scope of a power outage, the prediction results can be verified by comparing them to the baseline value (for example, checking whether the voltage has returned to normal levels) to ensure forecast accuracy.

[0067] S102: Determine a time series data set of each component of the low-voltage power grid operation variable according to the low-voltage power grid operation variable.

[0068] The components refer to individual variables in the above-mentioned low-voltage power grid operation variables, such as the bus voltage amplitude, bus voltage phase, active power flow on each branch line, and reactive power flow on each branch line of the low-voltage power grid.

[0069] A time series dataset samples each component in a time series, for example, once per second or dynamically adjusts the sampling rate based on the disturbance characteristics. Alternatively, continuous signals can be converted into discrete time series data using methods such as sliding windows or wavelet transforms for subsequent analysis.

[0070] S103 : Predicting the low-voltage power grid operation variables according to the time series data sets and prediction registers of each component to obtain a predicted operation variable sequence.

[0071] Among them, the prediction register is a storage unit for storing historical data of variables in the low-voltage power grid. It can be implemented in the form of a sliding window queue to ensure that the earliest data can be removed and the latest data can be inserted. It can retain the timing dependencies of the power grid operation (such as voltage relationships and power fluctuation trends) to provide the connection between the upper and lower data for the prediction of the scope of low-voltage power outages.

[0072] A predicted operating variable sequence is a collection of predicted values ​​of grid variables arranged in chronological order, such as the voltage, phase, or power predicted values ​​for the next 5 seconds. Each predicted value is accompanied by a timestamp, forming a traceable time series curve.

[0073] Specifically, data of the first length (e.g., the most recent 100 sampling points) is selected from the time series dataset as the training set. It is important to note that the data within the first length must contain disturbance characteristics (e.g., short-circuit faults typically require millisecond-level data). Furthermore, different lengths can be set for different variables (e.g., a short window is used to capture sudden changes in voltage, and a long window is used to smooth noise in power).

[0074] The learning data is then stored by variable type (voltage, power, etc.) and updated with a window shift, removing the oldest data and inserting the latest measured values ​​to maintain register data timeliness. This is followed by prediction and weight adjustment, using the current weights to calculate the predicted value for the next moment. Finally, the predicted value is compared with the measured value, and the weights are updated using algorithms such as gradient descent.

[0075] By repeating the above steps, the voltage amplitude, phase, and active / reactive power flow are predicted separately and finally combined into a complete sequence.

[0076] The embodiment of the present application reduces the prediction error caused by the access of new energy (such as photovoltaic output fluctuation) through dynamic weight adjustment. In addition, the sliding window mechanism only requires local data, avoiding the backtracking of all historical data, and is suitable for real-time control.

[0077] S104. Determine a component power outage prediction index corresponding to each component according to the predicted operating variable sequence.

[0078] Among them, the component power outage prediction index refers to a quantitative indicator, which is used to evaluate the probability or risk level of a power outage failure of a certain component of the power grid (such as voltage, active power, etc.) in the future period.

[0079] Specifically, historical data on key operating parameters of the power grid must be collected. These variables are fundamental to its normal operation. Based on this historical data, machine learning algorithms are then used to build predictive models. Commonly used models include time series analysis and neural network models, which are used to predict the changing trends of various operating variables over future time periods.

[0080] For each operating variable, a quantitative indicator of power outage risk can be defined based on its predicted value and corresponding thresholds (such as low voltage and power imbalance). For example, low voltage can lead to a low-voltage power outage, so the power outage prediction index for the voltage component is determined based on the difference between the predicted voltage value and the set low-voltage threshold. This component power outage prediction index can be obtained by calculating the degree of abnormality of each operating parameter. If the predicted value of a component exceeds the normal operating range or threshold, it means that the probability of a power outage for that component increases, and the corresponding power outage prediction index is calculated.

[0081] S105 . Verify the component power outage prediction index corresponding to each component according to the variable reference value during normal operation to obtain a low-voltage power outage prediction index.

[0082] The low-voltage power outage prediction index is used to predict the extent of a low-voltage power outage. Specifically, the component power outage prediction index obtained in S104 is compared with the normal operating variable baseline values. If the predicted operating variables deviate from the normal range, it can be considered that there may be a power outage risk. For example, if the predicted voltage value is significantly lower than the normal baseline value, the corresponding low-voltage power outage risk increases.

[0083] Furthermore, during the comparison process, if the predicted values ​​of certain components (such as voltage and current) deviate from the baseline values, a set algorithm (such as a deviation threshold) can be used to determine whether there is a potential failure risk. If the predicted value exceeds the set deviation range, the component outage prediction index will reflect a higher probability of power outage.

[0084] Finally, based on the verification of each component, a comprehensive assessment of the risk of a low-voltage power outage is made. This step is usually achieved by weighting the power outage prediction indices of multiple components to obtain an overall low-voltage power outage prediction index.

[0085] The present embodiment accurately predicts changes in grid operating variables and compares them with normal operating baseline values, thereby identifying potential power outage risks in advance. By calculating component power outage prediction indices and low-voltage power outage prediction indices, grid operators can promptly identify abnormal conditions that may lead to power outages, take preventive measures in advance, reduce the occurrence of power outages, and improve the safety and reliability of the grid.

[0086] Figure 2 This is a flow chart of the second embodiment of the method for predicting the range of low-voltage power outages provided by this application. Figure 2 As shown, in Figure 1 Based on the embodiment, the low-voltage power grid operation variables are predicted according to the time series data sets and prediction registers of each component to obtain a predicted operation variable sequence, including:

[0087] S201. Determine, based on the time series data sets of each component, a data training set corresponding to the time series data sets of each component.

[0088] Among them, the data training set refers to a data subset of a specific time period extracted from time series data and used to train the prediction model. Its time length needs to be adjusted according to the dynamic characteristics of the power grid.

[0089] In this step, the validity of each time series data set must be verified to remove outliers caused by communication interruptions or sensor failures, such as zero or out-of-limit values. Digital filtering technology is also used to eliminate high-frequency noise, preserving the power frequency fundamental and major harmonic components.

[0090] Specifically, the time window can be dynamically selected based on the grid's operating status. For example, during steady-state operation, a fixed-length window is used for data collection, while during transient disturbances, the window is automatically shortened to a quarter of the power frequency cycle to capture rapid changes. For example, when a voltage drop is detected, data is immediately captured 10ms forward and 20ms backward from the start of the drop as the training set.

[0091] S202 , updating each data training set to a prediction register respectively, to obtain an updated prediction register corresponding to each component.

[0092] In this step, the prediction register must first be initialized. For example, when power is turned on, the most recent one-minute historical data is loaded from the power grid system to fill the prediction register, and the depth of the prediction register is set to 50-100 sampling points (to ensure coverage of 1-2 power frequency cycles). Then, each time new data is added, the oldest record in the prediction register is automatically eliminated, and the new data is threshold-checked. If the rate of change of a variable exceeds 10% per millisecond, it is marked as "urgent data" and processed first. For example, if the voltage of a node drops sharply from 0.95pu to 0.82pu, the data is marked as urgent data, immediately triggering a reload of the prediction register.

[0093] Furthermore, each data training set is updated to the prediction register respectively to obtain an updated prediction register corresponding to each component, including:

[0094] Each data training set is updated to the prediction register respectively, so as to update the register parameters in each prediction register and obtain the updated register parameters corresponding to each prediction register.

[0095] Prediction register parameters refer to dynamically adjusted parameters stored in the prediction register, including the data storage weight coefficient (the weight distribution of new data and historical data), the variable association matrix (describing the coupling relationship between variables such as voltage, phase, and power), and the time decay factor (controlling the decay rate of the influence of historical data). Register parameter updates are generated by statistically analyzing the data training set and adjusting the internal parameters of the register in real time. Typical parameters include the sliding window size (dynamically adjusted data storage duration), the variable sensitivity threshold (the critical value of the rate of change that triggers parameter updates), and the prediction confidence weight (the contribution of each variable to the comprehensive prediction).

[0096] Exemplarily, the data at the next moment is predicted based on the updated prediction register and the prediction weight is updated, specifically:

[0097]

[0098] Among them, PD(x) represents the predicted data at the next moment, W represents the prediction weight, and W T represents the transpose of the prediction weight, Wnew represents the updated prediction weight, U represents the prediction register data, Unew represents the updated prediction register data, UnewT represents the transpose of the updated prediction register data, K1 represents the proportional coefficient, and TR(x+1) represents the learning data at the next moment.

[0099] In this step, the data storage order is rearranged according to the parameters of each update register and the updated window size, and expired data is soft-deleted (marked as invalid but not physically cleared). A dual buffer mechanism is then adopted. For example, the active area is used to store the current valid data, and the update area is used to receive new data and apply the new parameters. The buffer is switched after every 10 updates.

[0100] S203: Perform prediction processing on the low-voltage power grid operation variables according to the updated prediction register to obtain a predicted operation variable sequence.

[0101] Among them, the predicted operating variable sequence is specifically the power grid variable value of the future period predicted based on historical data. It is a sequence formed in chronological order, and each predicted value is marked with the prediction time and confidence level.

[0102] In this step, the latest data is extracted from the update prediction register and fed into the prediction model. The prediction model can then output predictions for multiple future time points and arrange them in chronological order. Finally, the prediction results are verified for physical plausibility. For example, the predicted voltage sequence for a line over the next 100ms is [0.90, 0.88, 0.85, 0.82], with the confidence level of each prediction being [92%, 89%, 85%, 80%].

[0103] Furthermore, according to the updated prediction register, the low-voltage power grid operation variables are predicted to obtain a predicted operation variable sequence of the low-voltage power grid operation variables, including:

[0104] According to the update prediction registers, the prediction sequence time and each component prediction value corresponding to each update prediction register are determined.

[0105] Among them, the prediction sequence time refers to the time coordinate system corresponding to the prediction result, including the time resolution and prediction time domain. The time resolution refers to the time interval between prediction points, and the prediction time domain is the future prediction duration calculated from the current moment. The component prediction value refers to the independent prediction result of each grid variable (voltage, phase and power, etc.), including the prediction index, confidence index and prediction method identifier. The predicted operating variable sequence is the final output time-aligned multivariate prediction result. All variables share the same time axis and include physical verification flags (such as power balance verification results).

[0106] According to the prediction sequence time, the prediction values ​​of each component are combined and processed to obtain the prediction operation variable sequence of the low-voltage power grid operation variables.

[0107] The merging process can employ a time alignment algorithm. For example, a unified prediction time coordinate system is established, with the current moment as the origin and a preset delay time interval to the end of the prediction time domain. The predicted values ​​of all variables need to be mapped onto this time axis. For example, the voltage prediction point [t0+10ms, t0+20ms] and the phase prediction point [t0+12ms, t0+22ms] need to be aligned to [t0+10ms, t0+20ms].

[0108] There is another case where, for prediction points whose time deviation exceeds 50% of the step size, cubic spline interpolation is required to generate new data points.

[0109] In addition, when outputting predicted data, conflicting data must be handled hierarchically based on the importance of the variables. For example, when voltage and power predictions conflict, voltage data is prioritized to reconstruct the power value. Finally, a structured prediction sequence is generated, including a timestamp array, a matrix of predicted values ​​for each variable, and a checksum. For example, the input prediction values ​​are voltage: [t0+10ms: 0.93pu, t0+30ms: 0.87pu]; phase: [t0+15ms: -5.8°, t0+25ms: -7.2°]. Alignment is then performed, where the phase data is interpolated to generate t0+10ms: -5.3° and t0+30ms: -7.8°. Finally, a check is performed. If the voltage drop (0.06pu / 20ms) at t0+30ms matches the phase change (-2.5° / 20ms), the final predicted operating variable sequence is generated.

[0110] The present embodiment ensures both computational efficiency in steady-state conditions and prediction accuracy in transient states by adjusting the training set window and prediction model in real time. Furthermore, multiple validation mechanisms are implemented during the prediction sequence generation process, including data validity checks, physical constraint verification, and confidence assessment.

[0111] Figure 3 This is a flow chart of the third embodiment of the method for predicting the range of low-voltage power outages provided by this application. Figure 3 As shown, in Figure 1 Based on the embodiment, determining the component outage prediction index corresponding to each component according to the predicted operating variable sequence includes:

[0112] S301. Establish an operating variable determination sequence for low-voltage power grid operating variables based on a predicted operating variable sequence.

[0113] Among them, the operating variable determination sequence refers to the key feature data set extracted from the predicted operating variable sequence for power outage risk assessment, including the degree of deviation of the predicted value (such as the voltage drop amplitude), change trend indicators (such as first-order / second-order derivatives), and duration characteristics (such as continuous over-limit time). Through the operating variable determination sequence, the original predicted value can be converted into determination features that can be quantified and analyzed.

[0114] Exemplarily, the running variable determination sequence JS(t) is constructed as follows:

[0115] JS(t)=[PV(t-num+1), PV(t-num+2),..., PV(t)]

[0116] Where t represents the sampling time, num represents the number of samples, and PV(t) represents the predicted operating variable at time t; PV(t) = [V t1 , V t2 ,……,V tm ],V t1 Represents the first component of the predicted operating variable at time t, V t2 Represents the second component of the predicted operating variable at time t, V tm Represents the mth component of the predicted operating variable at time t.

[0117] S302: Classify the operating variable determination sequence to obtain a component determination sequence corresponding to each component.

[0118] Among them, the component judgment sequence is a judgment sequence generated for a single grid variable (such as voltage, phase, etc.), which contains the threshold violation record of the variable (such as the period when the voltage is <0.8pu) and the dynamic weight coefficient (reflecting the sensitivity of the variable to power outages). It can ensure that each component sequence maintains time synchronization, but the evaluation indicators are calculated independently.

[0119] In this step, the classification rules for the operating variable determination sequence can be processed based on three components: voltage component, phase component, and power component. When classifying the voltage component, the focus is on amplitude drop and recovery speed, and the key indicator can be set as 80% rated voltage duration. When classifying the phase component, the focus is on detecting and monitoring sudden changes and loss of step risk, and the key indicator can be set as the accumulated time when the phase difference between adjacent nodes is greater than 30°. For the power component, the focus can be on analyzing power flow reversal and overload, and the key indicator can be the reverse power duration or the overload multiple.

[0120] Secondly, when performing dynamic weight allocation, the basic weight of voltage can be set to 0.5, the basic weight of phase can be set to 0.3, and the basic weight of power can be set to 0.2. For example, when performing operational variable classification processing on a line in a low-voltage power grid, the component determination sequence of the line can be: (1) voltage sequence: over-limit amplitude 15%, lasting 40ms, weight 0.68; (2) phase sequence: out of step 25°, lasting 20ms, weight 0.33; (3) power sequence: overload 1.2 times, lasting 60ms, weight 0.24.

[0121] S303: Determine a component power outage prediction index corresponding to each component based on the component determination sequence and historical power outage data information of each component.

[0122] Among them, the historical power outage data information is the power outage data information of the low-voltage power grid during its historical operation, including historical power outage event records (time, location, type), the change trajectory of the power grid variables before the power outage (such as the voltage waveform 10 seconds before the fault) and environmental related data (such as thunderstorm, typhoon and other meteorological information), which can be stored in the form of a tokenized event set in the time series database. In this step, a dynamic time warping algorithm can be used to calculate the similarity between the current sequence and the historical sequence before the power outage. For example, the index calculation formula can be:

[0123] COP i =α·S {sim} +βW i T {dur} +γ·R {env}

[0124] Among them, S {sim} refers to the similarity between the current sequence and the historical power outages, Wi represents the component weight, T {dur} represents the abnormal duration coefficient, R {env} Represents environmental risk factors.

[0125] By establishing and classifying a determination sequence for low-voltage power grid operating variables, the present embodiment enables more accurate analysis of the operating status of each grid component. Combined with historical outage data, a component outage prediction index is calculated, effectively assessing the failure risk of each component. This provides a scientific basis for grid scheduling and maintenance decisions, identifies potential outage risks in advance, reduces the occurrence of grid failures, and improves grid stability and reliability.

[0126] Figure 4 This is a flow chart of the fourth embodiment of the method for predicting the range of low-voltage power outages provided by this application. Figure 4 As shown, in Figure 1 On the basis of the embodiment, according to the variable reference value during normal operation, the component power outage prediction index corresponding to each component is verified and processed to obtain the low-voltage power outage prediction index, including:

[0127] S401. Calculate the total offset distance between each component and the variable reference value according to the variable reference value during normal operation.

[0128] The normal operating variable baseline refers to the ideal or standard values ​​of all relevant electrical variables (such as current, voltage, and frequency) during normal grid operation. These baseline values ​​reflect the grid's performance under normal conditions. The offset distance refers to the difference or degree of deviation between the currently observed variable and its baseline value. A larger offset distance indicates a further deviation from normal operation, potentially posing a risk of failure.

[0129] Specifically, the baseline value of each component during normal operation is first obtained. Then, the current electrical variable value of each component is acquired through a real-time monitoring or data acquisition system. Finally, the deviation between the current variable value and the baseline value is calculated to obtain the offset distance. This offset distance is then recorded for each component to provide data for subsequent analysis.

[0130] For example, first, the offset distance of each component in the predicted operating variable needs to be calculated. For example, the first offset distance of each component in the predicted operating variable needs to be calculated:

[0131]

[0132] Where DisD1 represents the first offset distance, i represents the component number of the predicted operating variable, n represents the total number of components, V ti represents the i-th component of the predicted operating variable at time t, V refi Indicates the baseline value of the i-th component of the predicted operating variable.

[0133] Calculate the second offset distance for each component of the forecast run variable:

[0134]

[0135] Wherein, DisD2 represents the second offset distance;

[0136] Calculate the third offset distance for each component of the forecast run variable:

[0137]

[0138] Wherein, DisD3 represents the third offset distance;

[0139] Calculate the fourth offset distance for each component of the forecast run variable:

[0140]

[0141] Wherein, DisD4 represents the fourth offset distance;

[0142] Calculate the fifth offset distance for each component of the forecast run variable:

[0143]

[0144] Wherein, DisD5 represents the fifth offset distance;

[0145] Calculate the sixth offset distance for each component of the forecast run variable:

[0146]

[0147] Wherein, DisD6 represents the sixth offset distance;

[0148] Calculate the seventh offset distance for each component of the forecast run variable:

[0149]

[0150] Wherein, DisD7 represents the seventh offset distance;

[0151] Calculate the eighth offset distance for each component of the forecast run variable:

[0152]

[0153] Wherein, DisD8 represents the eighth offset distance;

[0154] Calculate the ninth offset distance for each component of the forecast run variable:

[0155]

[0156] Wherein, DisD9 represents the ninth offset distance.

[0157] Finally, based on the above first to ninth offset distances, the total offset distance at any time is calculated, specifically:

[0158]

[0159] Where CD(t1,t2) represents the total offset distance between time t1 and t2, k represents the sub-offset distance number, j represents the running variable number, n represents the total number of running variables, DisD t1kj DisD represents the kth sub-offset distance of the jth running variable at time t1. t2kj represents the kth sub-offset distance of the jth running variable at time t2, D avgt1t2kj represents the mean of the kth sub-offset distance of the jth running variable at time t1 and t2.

[0160] S402 : Perform offset clustering processing on each component according to the total offset distance of each component to obtain offset status information of each component.

[0161] Among them, the offset clustering process refers to dividing each component data into K clusters so that the data points in cu are as similar as possible, wherein each cluster has a mean center, and the cluster center is updated iteratively until the clustering result converges.

[0162] In this step, an operating variable (such as voltage, current, or frequency) is first selected as the central variable for clustering. This variable serves as the basis for measuring offset distance. Offset distance is then calculated: for each component, the offset distance from the normal baseline value is calculated. The offset distance is calculated based on the selected operating variable. Finally, K-means clustering is performed, and K initial cluster centers are selected. These centers can be initially randomly selected offset distance values ​​or selected based on the data distribution. Each component is assigned to the cluster center closest to it based on its offset distance. The cluster centers are then updated so that the center of each cluster is the mean of the offset distances of all components within the cluster. Finally, this process is repeated until the cluster centers remain unchanged or change only minimally (reaching convergence). After clustering is complete, each component is assigned to a cluster (or class) representing similar offset states. The clustering results provide information about the offset state of each component. For example, some components may belong to the normal class, while others may belong to the class with large offsets or abnormalities.

[0163] S403 : Verify the component power outage prediction index corresponding to each component according to the offset state information of each component, and obtain a verification result of the component power outage prediction index corresponding to each component.

[0164] The component outage prediction index predicts the probability or severity of a power outage by analyzing the potential failure risk of each component (such as power equipment or lines). Verification involves verifying the accuracy of the power outage prediction index to confirm whether it meets predetermined standards or requirements.

[0165] In this step, based on the offset state information obtained in step S402, a predicted power outage index is generated in combination with a machine learning model. The index can indicate the possibility of a power outage in the low-voltage power grid. The predicted power outage index is then verified, for example, by comparing historical power outage events with the predicted power outage index to test the accuracy of the prediction. This step may be performed using methods such as cross-validation. The effect of the prediction can then be evaluated using statistical methods, such as calculating the prediction accuracy (accuracy, recall rate, etc.) and checking whether there is over-prediction or underestimation. For the difference between the predicted results and the actual results, error analysis can be performed to identify potential problems in the model or data. During the verification process, the model will generate verification results, including an accuracy report of the power outage prediction index. These results are used to evaluate whether the current model is accurate enough and whether it needs to be adjusted or optimized.

[0166] S404: If the verification result meets the preset prediction index requirement, weighted processing is performed on each component power outage prediction index to obtain a low-voltage power outage prediction index.

[0167] The low-voltage power outage prediction index is used to predict the power outage range of the low-voltage power grid. This step mainly performs weighted calculation on the power outage indices predicted by each component to finally obtain a comprehensive low-voltage power outage prediction index, which is used to predict the power outage range of the low-voltage power grid.

[0168] Specifically, in S403, if the component power outage prediction indices obtained after verification are accurate and meet preset prediction requirements (for example, the prediction error is within an allowable range), these prediction indices can continue to be processed.

[0169] The power outage prediction index of each component may have different importance. For example, the power outage prediction index of some key equipment or important lines may be more important than that of other equipment. Therefore, it is necessary to weight the power outage prediction index of each component. The weighting can be based on the importance of the equipment, its location in the power grid, the historical failure frequency of the equipment, etc. Weighting is to multiply the prediction index of each component by a weight coefficient and then sum the weighted values ​​to obtain the low-voltage power outage prediction index. For example, the power outage prediction index of certain key equipment in the low-voltage power grid will be assigned a higher weight, while other equipment will be assigned a lower weight.

[0170] The resulting low-voltage power outage prediction index is used to predict the scope of power outages in the low-voltage power grid. This allows for early identification of areas where failures are likely to occur, allowing for the adoption of appropriate preventive measures (e.g., dispatching equipment, arranging maintenance, etc.) to improve the stability and reliability of the power grid.

[0171] This embodiment of the application calculates and verifies the power outage prediction index of each component to ensure the reliability of the prediction results, and then derives a comprehensive low-voltage power outage prediction index through weighted processing. This process can identify areas with high power outage risk in advance, effectively preventing power grid failures, optimizing resource scheduling, and improving the stability and reliability of the power grid, ensuring power supply security.

[0172] Figure 5 This is a flow chart of the fifth embodiment of the method for predicting the range of low-voltage power outages provided by this application. Figure 5 As shown, in Figure 4 On the basis of the embodiment, according to the offset state information of each component, the component power outage prediction index corresponding to each component is verified, and the verification result of the component power outage prediction index corresponding to each component is obtained, including:

[0173] S501 : Determine power outage probability information corresponding to each component based on each offset state information and historical power outage data information of each component.

[0174] In this step, we first collect information on the offset status of each component in the power grid (e.g., transformers, lines, and equipment). This information typically includes equipment operating status (e.g., normal, overloaded, faulty), environmental factors (e.g., temperature, humidity), and grid load status. By analyzing past power outage events, including their occurrence time, duration, impact range, and cause, we can estimate the outage risk of each component under different offset states. For example, a transformer may be more likely to experience an outage when it is overloaded.

[0175] Then, a power outage probability model is constructed using statistical methods (such as regression analysis and Bayesian analysis) or machine learning methods (such as decision trees and neural networks). The outage probability of each component is calculated based on the offset state information and historical outage data. For example, invalid data is removed through data cleaning; relevant features are extracted based on historical data and offset states; and the model is trained using an algorithm to obtain the power outage probability of each component. Finally, each component is assigned a power outage probability value, indicating the likelihood of a power outage in the current state of the component.

[0176] S502: Compare the power outage probability information with the component power outage prediction index corresponding to each component to verify the component power outage prediction index corresponding to each component and obtain a verification result of the component power outage prediction index corresponding to each component;

[0177] Among them, the verification results include meeting the preset prediction index requirements and not meeting the preset prediction index requirements. The preset prediction index requirements are that the comparison difference between the power outage probability information and the component power outage prediction index corresponding to each classification is within the preset threshold range.

[0178] It's important to note that the outage probability information is calculated based on historical data and real-time status, while the outage prediction index is the prediction result provided by the model. When comparing the two, the difference or similarity between them can be calculated. For example, a simple difference calculation (outage probability minus prediction index) can be used to assess the consistency between the two. If the difference is too large, it indicates that the prediction index is biased. Alternatively, to more systematically evaluate the comparison results, an error metric can be used, such as absolute error (the absolute value of the difference between the prediction index and the actual outage probability) or relative error (the ratio of the error to the actual outage probability).

[0179] For example, during the verification process, a reasonable threshold range is set to determine whether the difference between the prediction index and the power outage probability is within an acceptable range. The threshold can be an absolute value (such as the difference cannot exceed 0.1) or a relative value (such as the difference cannot exceed 10% of the power outage probability). If the difference between the power outage probability and the power outage prediction index is within the preset threshold range, it means that the prediction index is accurate and the verification result is "satisfies the requirements". If the difference exceeds the threshold range, it means that the prediction index fails to accurately reflect the power outage risk and the verification result is "does not meet the requirements".

[0180] In another implementation of the present embodiment, for components that fail verification, methods for improving the prediction model can be used, such as optimizing feature selection, adjusting model parameters, or using other algorithms for prediction. Model training can also be performed by adding more historical data to improve prediction accuracy. For some components, manual intervention may be required, such as adjusting risk assessment parameters or performing equipment maintenance.

[0181] By accurately calculating the probability of a power outage and verifying it with a power outage prediction index, the embodiments of this application can improve the accuracy of power grid fault warnings, thereby identifying potential risks in advance, taking preventive measures, and reducing the occurrence of power outages. This not only ensures the stability of power supply, but also optimizes the operating efficiency of the low-voltage power grid, improves the targeted maintenance of equipment, and ensures the safety and reliability of the power system.

[0182] Figure 6 This is a simulation diagram of the predicted running variable sequence provided by this application. Figure 6 As shown, in Figure 2 Based on the embodiment, a simulation diagram of the predicted running variable sequence is obtained using a machine learning method.

[0183] Among them, the training data in the figure uses data with a sequence of 500-2000. The dotted line in the data with sequence numbers 2000-3000 is the predicted operating variable data, and the solid line is the actual operating variable data. It can be seen that although the predicted operating variable data deviates from the actual data in amplitude, it generally simulates the changing trend of the data well and can play a role in alleviating data noise and disturbance.

[0184] Figure 7 This is a schematic diagram of the structure of the device for predicting the range of low-voltage power outages provided by this application. Figure 7 As shown, the prediction device 70 for the range of low-voltage power outage includes an acquisition module 701, a time series data set determination module 702, an operation variable sequence prediction module 703, a component power outage prediction index determination module 704 and a low-voltage power outage prediction index acquisition module 705.

[0185] The acquisition module 701 is used to acquire the low-voltage grid operation variables after the disturbance is added and the variable reference values ​​during normal operation;

[0186] A time series data set determining module 702 is configured to determine a time series data set of each component of the low-voltage power grid operation variable according to the low-voltage power grid operation variable;

[0187] The operating variable sequence prediction module 703 is used to predict the low-voltage power grid operating variables based on the time series data sets and prediction registers of each component to obtain a predicted operating variable sequence;

[0188] A component power outage prediction index determination module 704 is configured to determine a component power outage prediction index corresponding to each component based on a predicted operating variable sequence;

[0189] The low voltage power outage prediction index obtaining module 705 is used to verify the component power outage prediction index corresponding to each component according to the variable reference value during normal operation.

[0190] Furthermore, the variable sequence prediction module 703 is further specifically configured to:

[0191] Based on the time series data sets and prediction registers of each component, the low-voltage power grid operation variables are predicted to obtain a predicted operation variable sequence, including:

[0192] According to the time series data sets of each component, determining the data training set corresponding to the time series data sets of each component;

[0193] Update each data training set to the prediction register respectively to obtain the updated prediction register corresponding to each component;

[0194] According to the updated prediction register, the low voltage power grid operation variables are predicted and processed to obtain a predicted operation variable sequence.

[0195] Furthermore, the variable sequence prediction module 703 is further specifically configured to:

[0196] Updating each data training set to the prediction register respectively to update the register parameters in each prediction register to obtain updated register parameters corresponding to each prediction register;

[0197] According to each update register parameter, each prediction register is updated to obtain an updated prediction register corresponding to each component.

[0198] Furthermore, the variable sequence prediction module 703 is further specifically configured to:

[0199] Determining prediction sequence times and component prediction values ​​corresponding to respective update prediction registers according to the update prediction registers;

[0200] According to the prediction sequence time, the prediction values ​​of each component are combined and processed to obtain the prediction operation variable sequence of the low-voltage power grid operation variables.

[0201] Furthermore, the component power outage prediction index determination module 704 is further specifically configured to:

[0202] Establishing an operating variable determination sequence for low-voltage power grid operating variables based on the predicted operating variable sequence;

[0203] Classify the operating variable determination sequence to obtain a component determination sequence corresponding to each component;

[0204] The component outage prediction index corresponding to each component is determined according to the component determination sequence and the historical outage data information of each component, where the historical outage data information is the outage data information of the low-voltage power grid during its historical operation.

[0205] Furthermore, the low-voltage power outage prediction index obtaining module 705 is further specifically configured to:

[0206] According to the variable reference value during normal operation, calculate the total offset distance between each component and the variable reference value;

[0207] Based on the total offset distance of each component, each component is subjected to offset clustering processing to obtain the offset status information of each component;

[0208] Verifying the component power outage prediction index corresponding to each component according to the offset state information of each component to obtain a verification result of the component power outage prediction index corresponding to each component;

[0209] If the verification result meets the preset prediction index requirements, the power outage prediction index of each component is weighted to obtain the low-voltage power outage prediction index;

[0210] Among them, the low-voltage power outage prediction index is used to predict the power outage scope of the low-voltage power grid.

[0211] Furthermore, the low-voltage power outage prediction index obtaining module 705 is further specifically configured to:

[0212] Determine the power outage probability information corresponding to each component based on each offset state information and historical power outage data information of each component;

[0213] Comparing the power outage probability information with the component power outage prediction index corresponding to each component to verify the component power outage prediction index corresponding to each component, and obtaining a verification result of the component power outage prediction index corresponding to each component;

[0214] Among them, the verification results include meeting the preset prediction index requirements and not meeting the preset prediction index requirements. The preset prediction index requirements are that the comparison difference between the power outage probability information and the component power outage prediction index corresponding to each classification is within the preset threshold range.

[0215] Figure 8 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present application. Figure 8 As shown, the electronic device 80 includes:

[0216] The electronic device 80 may include one or more processors 801 , one or more computer-readable storage media memories 802 , and a communication component 803 . The processor 801 , the memory 802 , and the communication component 803 are connected via a bus 804 .

[0217] In a specific implementation process, at least one processor 801 executes the computer-executable instructions stored in the memory 802 , so that the at least one processor 801 performs the above-mentioned power grid operation task processing method.

[0218] The specific implementation process of the processor 801 can be found in the above method embodiment. Its implementation principle and technical effects are similar and will not be repeated here in this embodiment.

[0219] In the above Figure 8In the illustrated embodiment, it should be understood that the processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present invention may be directly executed by a hardware processor or by a combination of hardware and software modules within the processor.

[0220] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (NVM), such as at least one disk memory.

[0221] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be divided into address buses, data buses, control buses, etc. For ease of illustration, the buses in the drawings of this application are not limited to just one bus or just one type of bus.

[0222] The specific implementation of the above operations can be found in the previous embodiments and will not be repeated here.

[0223] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be accomplished by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0224] To this end, an embodiment of the present application provides a computer-readable storage medium storing a plurality of instructions, which can be loaded by a processor to execute the steps of any low-voltage power grid outage range prediction method provided in the embodiment of the present application.

[0225] The storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0226] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of the present application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0227] It should be noted that for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all optional embodiments, and the actions and modules involved are not necessarily required by this application.

[0228] It should be further noted that, although the various steps in the flowchart are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps may be performed in other orders. Moreover, at least a portion of the steps in the flowchart may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily performed at the same time, but may be performed at different times. The execution order of these sub-steps or stages is not necessarily to be performed in sequence, but may be performed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0229] It should be understood that the above-described device embodiments are merely illustrative, and the device of the present application may also be implemented in other ways. For example, the division of units / modules in the above-described embodiments is merely a logical functional division, and actual implementations may employ other division methods. For example, multiple units, modules, or components may be combined or integrated into another system, or some features may be omitted or not implemented.

[0230] This embodiment further provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the method for predicting the power outage range of the low-voltage power grid provided by any of the above embodiments.

[0231] In addition, unless otherwise specified, the functional units / modules in the various embodiments of the present application may be integrated into a single unit / module, each unit / module may exist physically separately, or two or more units / modules may be integrated together. The aforementioned integrated units / modules may be implemented in the form of hardware or software program modules.

[0232] If the integrated unit / module is implemented in hardware, the hardware may be digital circuits, analog circuits, etc. The physical implementation of the hardware structure includes, but is not limited to, transistors, memristors, etc. Unless otherwise specified, the processor may be any appropriate hardware processor, such as a CPU, GPU, FPGA, DSP, and ASIC. Unless otherwise specified, the storage unit may be any appropriate magnetic storage medium or magneto-optical storage medium, such as resistive random access memory (RRAM), dynamic random access memory (DRAM), static random access memory (SRAM), enhanced dynamic random access memory (EDRAM), high-bandwidth memory (HBM), hybrid memory cube (HMC), etc.

[0233] If the integrated unit / module is implemented in the form of a software program module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a memory, including a number of instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0234] In the above embodiments, the description of each embodiment has its own emphasis. For parts not described in detail in a particular embodiment, please refer to the relevant description of other embodiments. The technical features of the above embodiments can be combined in any way. To keep the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0235] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of the present application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0236] It should be understood that the present application is not limited to the exact structure 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 predicting the scope of a low-voltage power outage, characterized in that: include: Collect the low-voltage grid operation variables after the disturbance is added and the variable baseline values ​​during normal operation; Determining, according to the low-voltage power grid operation variable, a time series data set of each component in the low-voltage power grid operation variable; Performing prediction processing on the low-voltage power grid operating variables according to the time series data sets and prediction registers of the various components to obtain a predicted operating variable sequence; determining, based on the predicted operating variable sequence, a component power outage prediction index corresponding to each component; According to the variable reference value during normal operation, the component power outage prediction index corresponding to each component is verified to obtain a low-voltage power outage prediction index, which is used to predict the power outage range of the low-voltage power grid.

2. The prediction method according to claim 1, characterized in that The step of performing prediction processing on the low-voltage power grid operation variables according to the time series data sets and prediction registers of the respective components to obtain a predicted operation variable sequence includes: Determining, based on the time series data sets of the respective components, a data training set corresponding to the time series data sets of the respective components; Updating each of the data training sets to a prediction register respectively to obtain an updated prediction register corresponding to each of the components; According to the updated prediction register, the low-voltage power grid operation variables are predicted to obtain a predicted operation variable sequence.

3. The prediction method according to claim 2, characterized in that The updating of each of the data training sets to a prediction register to obtain an updated prediction register corresponding to each of the components includes: Updating each of the data training sets to the prediction registers respectively, so as to update the register parameters in each of the prediction registers and obtain updated register parameters corresponding to each of the prediction registers; According to each of the update register parameters, update processing is performed on each of the prediction registers to obtain an updated prediction register corresponding to each of the components.

4. The prediction method according to claim 2, characterized in that According to the updated prediction register, the low-voltage power grid operation variable is predicted to obtain a predicted operation variable sequence of the low-voltage power grid operation variable, including: Determining, according to the update prediction registers, the prediction sequence time and each component prediction value corresponding to each of the update prediction registers; According to the prediction sequence time, the predicted values ​​of the various components are combined and processed to obtain a predicted operating variable sequence of the low-voltage power grid operating variables.

5. The prediction method according to claim 1, wherein: Determining, according to the predicted operating variable sequence, a component power outage prediction index corresponding to each component, including: Establishing an operating variable determination sequence of the low-voltage power grid operating variables according to the predicted operating variable sequence; performing classification processing on the operating variable determination sequence to obtain a component determination sequence corresponding to each component; The component power outage prediction index corresponding to each component is determined according to the component determination sequence and historical power outage data information of each component, wherein the historical power outage data information is power outage data information of the low-voltage power grid during historical operation.

6. The prediction method according to claim 1, characterized in that The verifying process of the component power outage prediction index corresponding to each component according to the variable reference value during normal operation to obtain the low-voltage power outage prediction index includes: Calculating the total offset distance between each component and the variable reference value according to the variable reference value during normal operation; performing an offset clustering process on each component according to the total offset distance of each component to obtain offset status information of each component; performing verification processing on the component power outage prediction index corresponding to each component according to the offset state information of each component to obtain a verification result of the component power outage prediction index corresponding to each component; If the verification result meets the preset prediction index requirement, weighting the component power outage prediction indexes to obtain a low-voltage power outage prediction index; The low-voltage power outage prediction index is used to predict the power outage scope of the low-voltage power grid.

7. The prediction method according to claim 6, characterized in that The verifying process of the component power outage prediction index corresponding to each component according to the offset state information of each component to obtain the verification result of the component power outage prediction index corresponding to each component includes: Determining power outage probability information corresponding to each component according to each of the offset state information and historical power outage data information of each component; Comparing the power outage probability information with the component power outage prediction index corresponding to each component to verify the component power outage prediction index corresponding to each component, and obtaining a verification result of the component power outage prediction index corresponding to each component; Among them, the verification results include meeting the preset prediction index requirements and not meeting the preset prediction index requirements, and the preset prediction index requirements are that the comparison difference between the power outage probability information and the component power outage prediction index corresponding to each classification is within a preset threshold range.

8. A device for predicting the range of a low-voltage power outage, characterized in that: include: An acquisition module is used to acquire the operating variables of the low-voltage power grid after a disturbance is added and the reference values ​​of the variables during normal operation; A time series data set determination module, configured to determine a time series data set of each component of the low-voltage power grid operation variable according to the low-voltage power grid operation variable; An operating variable sequence prediction module is used to predict the operating variables of the low-voltage power grid according to the time series data sets and prediction registers of each component to obtain a predicted operating variable sequence; a component power outage prediction index determination module, configured to determine a component power outage prediction index corresponding to each component according to the predicted operation variable sequence; The low-voltage power outage prediction index obtaining module is used to verify the component power outage prediction index corresponding to each component according to the variable reference value during normal operation to obtain the low-voltage power outage prediction index.

9. An electronic device, characterized in that: include: a processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 7.

10. 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 7 when executed by a processor.

11. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 7 when the computer program is executed by a processor.

Images