Power grid fault early warning system based on machine learning and related device

Abstract

The invention relates to the technical field of data processing, and discloses a power grid fault early warning system based on machine learning and a related device, and the system is characterized in that an initial monitoring module identifies and identifies abnormal data points through monitoring the voltage, current and load of a power grid, and obtains an abnormal mode identification result; the load fluctuation analysis module positions the abnormal node according to the abnormal mode identification result to obtain abnormal node positioning information; the trend analysis and prediction module evaluates and predicts fault nodes and time according to the abnormal node positioning information to obtain a risk prediction result; the strategy formulation and response module formulates an operation instruction based on the risk prediction result to obtain a fault prevention measure; the fault simulation and test module performs virtual fault simulation by adopting fault prevention measures, and performs iterative optimization on a strategy to obtain a test evaluation result; and the fault early warning and optimization module optimizes a fault prevention strategy according to a test evaluation result, and a more accurate target fault early warning and a faster and more effective target response strategy can be obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing and fault prediction, and in particular to a power grid fault early warning system based on machine learning and related devices. Background Art

[0002] The power grid fault early warning system is a key technology used in the power industry. It aims to predict and warn of impending failures by monitoring the operation status of the power grid and environmental variables. The system can use various sensors to collect real-time data from the power grid, such as voltage, current and frequency, as well as environmental factors such as temperature and humidity, and feed the collected data into an analysis model to identify abnormal patterns that lead to failures and issue early warnings in a timely manner so that operators can take necessary measures to avoid or mitigate the impact of failures, thereby ensuring the stability and safe operation of the power grid.

[0003] In the prediction and management of power grid faults, the existing technology mainly relies on traditional fault indicators and early warning systems. The system lacks the ability to deeply analyze and respond to complex data fluctuations in real time. Especially in the context of the increasing scale of power grids and the increasing complexity of data types, the limitations of traditional technologies in processing multi-node and multi-variable data lead to inaccurate fault predictions and the inability to effectively foresee and handle sudden and complex fault modes. In addition, the existing technology adopts a responsive maintenance strategy, that is, processing only after the fault occurs. This strategy not only increases maintenance costs, but also affects the stable operation of the power grid and the reliability of power supply. In an emergency, this low-accuracy fault prediction mode and delayed response mode may lead to a wider range of service interruptions and safety risks. Therefore, how to improve the fault prediction accuracy and fault response speed of the power grid fault early warning system has become a problem that needs to be solved urgently. Summary of the invention

[0004] The present invention provides a power grid fault early warning system and related devices based on machine learning, which can improve the accuracy of power grid fault prediction and the speed of fault response, so as to solve the technical problems of low accuracy of fault prediction mode and lagging response mode in the prior art.

[0005] In a first aspect, a machine learning-based power grid fault early warning system is provided, which includes an initial monitoring module, a load fluctuation analysis module, a trend analysis and prediction module, a strategy formulation and response module, a fault simulation and testing module, and a fault early warning and optimization module, wherein:

[0006] The initial monitoring module monitors the voltage, current and load of the power grid, identifies and marks abnormal data points, and obtains abnormal pattern identification results;

[0007] The load fluctuation analysis module locates the abnormal nodes according to the abnormal pattern identification results and obtains the abnormal node location information;

[0008] The trend analysis and prediction module evaluates and predicts the fault node and time based on the abnormal node location information to obtain the risk prediction results;

[0009] The strategy formulation and response module formulates operational instructions based on risk prediction results and obtains preventive measures for failures;

[0010] The fault simulation and testing module uses preventive measures to simulate virtual faults and iteratively optimizes the strategies to obtain test evaluation results.

[0011] The fault warning and optimization module optimizes the fault prevention strategy according to the test evaluation results to obtain the target fault warning and target response strategy.

[0012] In a second aspect, a power grid fault early warning device based on machine learning is provided, wherein:

[0013] The first processing module is used to identify and mark abnormal data points by monitoring the voltage, current and load of the power grid to obtain abnormal pattern identification results;

[0014] The second processing module is used to locate the abnormal node according to the abnormal pattern identification result to obtain abnormal node location information;

[0015] The third processing module is used to evaluate and predict the fault node and time according to the abnormal node location information to obtain the risk prediction result;

[0016] The fourth processing module is used to formulate operation instructions based on the risk prediction results and obtain preventive measures for failures;

[0017] The fifth processing module is used to perform virtual fault simulation using preventive fault measures and iteratively optimize the strategy to obtain a test evaluation result;

[0018] The sixth processing module is used to optimize the fault prevention strategy according to the test evaluation results to obtain the target fault warning and target response strategy.

[0019] In a third aspect, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the processing steps of the above-mentioned machine learning-based power grid fault warning system when executing the computer program.

[0020] In a fourth aspect, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the processing steps of the above-mentioned power grid fault warning system based on machine learning are implemented.

[0021] In the scheme implemented by the above-mentioned machine learning-based power grid fault warning system and related devices, the initial monitoring module, by monitoring the voltage, current and load of the power grid, identifies and marks abnormal data points, and obtains abnormal pattern identification results, so that the load fluctuation analysis module can locate the abnormal node according to the abnormal pattern identification results and obtain abnormal node location information, and further enables the trend analysis and prediction module to evaluate and predict the fault node and time according to the abnormal node location information to obtain risk prediction results, so that the strategy formulation and response module can formulate operation instructions based on the risk prediction results and obtain fault prevention measures, so that the fault simulation and testing module can use fault prevention measures to perform virtual fault simulation, and iteratively optimize the strategy to obtain test evaluation results, and then the fault warning and optimization module can optimize the fault prevention strategy according to the test evaluation results, and obtain accurate target fault warning and fast and effective target response strategy, which can effectively improve the accuracy of power grid fault prediction and the speed of fault response, and is conducive to achieving more accurate and efficient power grid fault warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0023] Figure 1 is a schematic diagram of an application environment of a power grid fault early warning system based on machine learning in one embodiment of the present invention;

[0024] Figure 2 It is a processing flow diagram of a power grid fault early warning system based on machine learning in one embodiment of the present invention;

[0025] Figure 3 is a structural schematic diagram of a power grid fault early warning device based on machine learning in one embodiment of the present invention;

[0026] Figure 4 is a schematic diagram of a structure of a computer device in one embodiment of the present invention;

[0027] Figure 5 It is another structural schematic diagram of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] The power grid fault early warning system based on machine learning provided by the embodiment of the present invention can be applied in the following aspects: Figure 1 In the application environment, the client communicates with the machine learning-based power grid fault early warning system through the network. Figure 1 As shown, the power grid fault early warning system based on machine learning can include an initial monitoring module, a load fluctuation analysis module, a trend analysis and prediction module, a strategy formulation and response module, a fault simulation and testing module, and a fault early warning and optimization module.

[0030] Exemplarily, taking the scenario of power grid fault early warning as an example, the management personnel can receive analysis data or prediction results from the power grid fault early warning system based on machine learning through the client. Accordingly, the initial monitoring module in the power grid fault early warning system based on machine learning can identify and mark abnormal data points by monitoring the voltage, current and load of the power grid, and obtain abnormal pattern identification results, so that the load fluctuation analysis module can locate the abnormal node according to the abnormal pattern identification results, and obtain abnormal node location information, and further enable the trend analysis and prediction module to evaluate and predict the fault node and time according to the abnormal node location information, and obtain risk prediction results, so that the strategy formulation and response module can formulate operation instructions based on the risk prediction results, and obtain preventive measures, so that the fault simulation and testing module can use preventive measures to simulate virtual faults, and iteratively optimize the strategy to obtain test evaluation results, and then the fault warning and optimization module can optimize the fault prevention strategy according to the test evaluation results, obtain accurate target fault warning and fast and effective target response strategy, and feed back the target fault warning and target response strategy to the client. Correspondingly, the client can receive the target fault warning and target response strategy feedback from the server, and can display the target fault warning and target response strategy feedback on the client for the target user to query or browse. By adopting the machine learning-based power grid fault warning system provided by this application, the accuracy of power grid fault prediction and the speed of fault response can be effectively improved, which is conducive to achieving the purpose of more accurate and efficient power grid fault warning.

[0031] The client may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, and portable wearable devices. The server may be implemented by an independent server or a server cluster consisting of multiple servers. The present invention is described in detail below through specific embodiments.

[0032] See also Figure 2 As shown, Figure 2 A schematic diagram of a processing flow of a power grid fault early warning system based on machine learning provided by an embodiment of the present invention includes the following steps:

[0033] S10: The initial monitoring module monitors the voltage, current and load of the power grid, identifies and marks abnormal data points, and obtains abnormal pattern identification results.

[0034] The initial monitoring module monitors the voltage, current and load of the power grid to identify data points that do not conform to the normal behavior pattern, analyzes the power grid data by setting thresholds, and obtains abnormal pattern identification results by identifying the abnormalities of the data points. The abnormal pattern identification results include but are not limited to: abnormal identification time, abnormal type, and abnormal intensity.

[0035] It should be understood that the initial monitoring module, by monitoring the voltage, current and load of the power grid, identifies and marks abnormal data points, and obtains abnormal pattern identification results, which refers to the process of how to obtain abnormal pattern identification results. Among them, in step S10, that is, the initial monitoring module, by monitoring the voltage, current and load of the power grid, identifies and marks abnormal data points, and obtains abnormal pattern identification results, can include the following steps:

[0036] S11: The voltage and current monitoring submodule samples the voltage, current and load of the power grid to obtain sampling data;

[0037] S12: the voltage and current monitoring submodule performs frequency analysis and stability detection on the sampled data to generate a stability evaluation result;

[0038] S13: The load analysis submodule dynamically tracks the power grid load based on the stability evaluation result to obtain load data;

[0039] S14: the load analysis submodule performs periodic analysis on the load data to obtain a load consistency analysis result;

[0040] S15: The abnormal pattern recognition submodule uses the load consistency analysis result to analyze the abnormal pattern and obtain abnormal data;

[0041] S16: The abnormal pattern recognition submodule performs data pattern comparison on the abnormal data to obtain an abnormal pattern identification result.

[0042] Among them, the voltage and current monitoring submodule can monitor the voltage, current and load of the power grid to continuously sample and record signals, and perform frequency analysis and stability detection on the sampled data, such as optimizing the monitoring process by adjusting the sampling interval and analysis period to further generate stability assessment results.

[0043] To monitor the voltage, current and load of the power grid, the voltage and current values ​​can be accurately measured every second, and the power consumption can be calculated to continuously sample and record the signals. The sampling frequency can be set to 1000Hz to capture instantaneous changes, and the noise can be removed by filtering to ensure the clarity and accuracy of the data. For frequency analysis, Fourier transform can be used to convert time domain signals into frequency domain signals and analyze the main frequency components. Stability detection can evaluate the amplitude of power grid load fluctuations by calculating the standard deviation and peak factor of voltage and current, and optimize the monitoring process by adjusting the sampling interval and analysis period to generate stability evaluation results.

[0044] Optionally, the calculation formula for stability detection can be as follows:

[0045]

[0046] Where S represents the result of stability test; N represents the total number of samples; i represents the index of a single sample; x i represents the voltage or current value of the i-th sample, μ represents the average value; σ represents the standard deviation; k a It represents an adjustment parameter, which can be used to enhance the sensitivity to extreme values. By increasing the value of k, the ability to determine grid instability can be enhanced.

[0047] The load analysis submodule can dynamically track the grid load through the stability assessment results, perform periodic analysis of the load data, further adjust the threshold to identify load fluctuations, and perform anomaly detection by comparing the real-time data with the set threshold, thereby obtaining the load consistency analysis results.

[0048] Through the stability assessment results, the dynamic tracking of the power grid load can be carried out, the highest and lowest load values ​​every minute can be recorded, and the mean and variance of the sliding window data can be calculated to determine the normal fluctuation range of the load, and the load data can be analyzed periodically, and the autocorrelation function can be used to identify the periodic characteristics of load changes. The load fluctuations identified by adjusting the threshold are based on the comparison between the load data at the same time of the previous day and the current data, and the load consistency analysis results are obtained by comparing the real-time data with the set threshold for anomaly detection.

[0049] Optionally, the following formula may be used to calculate the load deviation:

[0050]

[0051] Where D is the calculated result of load deviation; L t Indicates the load value at the current time point t, L avg Represents the average load value at the same time point on the previous day; σ L Indicates the standard deviation of the load.

[0052] The abnormal pattern identification submodule can use the load consistency analysis results to perform abnormal pattern analysis, such as performing time series analysis on the identified abnormal data to adjust identification parameters and match differentiated power grid states, and further identify abnormal behaviors through data pattern comparison to obtain abnormal pattern identification results.

[0053] The results of load consistency analysis can be used to analyze abnormal patterns. First, define the abnormal pattern as an event in which the load changes suddenly beyond the set threshold, perform time series analysis on the identified abnormal data, and use moving average and exponential smoothing techniques to perform trend analysis to identify abnormal changes in the short term, adjust the identification parameters and match the differentiated power grid status. The parameters include window size and smoothing coefficient to adapt to different monitoring needs, identify abnormal behaviors through data pattern comparison, and then obtain abnormal pattern identification results.

[0054] S20: The load fluctuation analysis module locates the abnormal node according to the abnormal pattern identification result to obtain abnormal node location information.

[0055] The load fluctuation analysis module can monitor the current and voltage of multiple nodes of the power grid according to the abnormal pattern identification results, and record the fluctuations that exceed the safety standards, so as to locate the abnormal nodes and obtain the abnormal node location information. The abnormal node location information includes but is not limited to: node identification, fluctuation duration, and fluctuation frequency.

[0056] It should be understood that the load fluctuation analysis module locates the abnormal node according to the abnormal pattern identification result to obtain the abnormal node location information, which refers to the process of how to obtain the abnormal node location information. Among them, in step S20, that is, the load fluctuation analysis module locates the abnormal node according to the abnormal pattern identification result to obtain the abnormal node location information, which may include the following steps:

[0057] S21: The real-time monitoring submodule monitors the operation status of the power grid in real time based on the abnormal mode identification result to obtain a real-time data snapshot;

[0058] S22: the fluctuation detection submodule extracts current and voltage data from the real-time data snapshot, identifies abnormal fluctuations, and generates an excessive fluctuation analysis result;

[0059] S23: The node positioning submodule uses the excessive fluctuation analysis result to identify and locate nodes with abnormal fluctuations and establish abnormal node positioning information.

[0060] Among them, the real-time monitoring submodule can collect current and voltage of multiple nodes of the power grid based on the abnormal mode identification results, and monitor the operating status of the power grid in real time by setting the sampling frequency and data recording interval, so as to obtain real-time data snapshots.

[0061] Based on the abnormal pattern identification results, current and voltage can be collected from multiple nodes of the power grid. First, the sampling frequency of current and voltage is set to 100 times per second, and the data recording interval is 1 minute. By adding data cache processing, the integrity and accuracy of the data can be maintained under high-frequency collection conditions. The real-time data processing algorithm is used to perform preliminary filtering and denoising on the collected current and voltage data. The sliding window technology is used to calculate the average current and voltage in each time window, generate a time series data snapshot, and realize real-time monitoring of the operation status of the power grid, thereby obtaining real-time data snapshots.

[0062] Optionally, the formula for data filtering can be as follows:

[0063]

[0064] Where Filtered represents the filtered value; i represents the number of sampling times; k represents the index of the sampling point; n represents the number of sampling points in the sliding window; x represents the number of sampling points in the sliding window; k represents the data of the kth sampling; w k Represents the weight coefficient, which can be decreased by the decay factor to give higher weight to recent data.

[0065] The fluctuation detection submodule can further extract current and voltage data from the real-time data snapshot, and monitor data fluctuations by setting thresholds, and mark data that exceeds the thresholds, so as to analyze the data change trend and fluctuation frequency to identify abnormal fluctuations, and then generate excessive fluctuation analysis results.

[0066] Extract current and voltage data from real-time data snapshots. First, set the current fluctuation threshold to 596 standard deviations and the voltage fluctuation threshold to 3% standard deviation. The thresholds can be set based on historical data distribution. Calculate the degree of deviation for each data point using statistical analysis methods, and mark data points where the deviation exceeds the threshold. In this way, each measurement of current and voltage can be monitored in real time, and data that exceeds the threshold can be marked. Fast Fourier transform can be used to analyze the data change trend and fluctuation frequency, thereby identifying abnormal fluctuations and generating excessive fluctuation analysis results.

[0067] Alternatively, the formula for fluctuation monitoring can be as follows:

[0068]

[0069] Where Fluctuation represents the result of fluctuation monitoring; m represents the total number of measured values; i represents the index of the number of measurements; x represents the number of measurements; i represents the i-th measurement of current or voltage, and μ represents the average value of the measurement series. This formula can determine the degree of fluctuation by calculating the absolute deviation of all measurements from the average value.

[0070] The node positioning submodule can use the excessive fluctuation analysis results to analyze the distribution of abnormal data points and locate the nodes in the power grid, so as to identify and locate the nodes with abnormal fluctuations by comparing the data patterns and abnormal behaviors, thereby establishing abnormal node positioning information.

[0071] The results of the over-standard fluctuation analysis can be used to analyze the distribution of abnormal data points. First, all data points marked as abnormal are collected, and the geographical and grid topological locations of the data points are classified using cluster analysis methods. The nodes in the grid are located by analyzing the concentration trend of the data points in each category, and then the nodes with abnormal fluctuations can be effectively identified and located based on the historical data patterns and the currently observed abnormal behaviors, and the abnormal node location information can be established.

[0072] S30: The trend analysis and prediction module evaluates and predicts the fault node and time according to the abnormal node location information to obtain a risk prediction result.

[0073] The trend analysis and prediction module can evaluate the changes in node data through the abnormal node location information, and can predict the fault node and time by drawing a potential fault map, thereby obtaining risk prediction results. The risk prediction results include but are not limited to: risk level, potential impact area, and impact duration.

[0074] It should be understood that the trend analysis and prediction module evaluates and predicts the fault node and time according to the abnormal node location information to obtain the risk prediction result, which refers to the process of how to obtain the risk prediction result. Among them, step S30, that is, the trend analysis and prediction module evaluates and predicts the fault node and time according to the abnormal node location information to obtain the risk prediction result, can include the following steps:

[0075] S31: The conventional data comparison submodule obtains the current abnormal node and historical fluctuation data based on the abnormal node location information;

[0076] S32: the conventional data comparison submodule uses time series analysis to identify the consistency and key deviation of the pattern according to the current abnormal node and the historical fluctuation data, and obtains similarity data and abnormality data;

[0077] S33: The conventional data comparison submodule constructs and obtains data comparison and analysis results according to the similarity data and the abnormality data;

[0078] S34: the fault trend analysis submodule clusters the data points according to the data comparison and analysis results to generate a fault trend graph;

[0079] S35: The risk prediction submodule predicts the fault nodes and times in the future time period based on the fault trend graph and adopts a time series prediction algorithm to obtain a reference fault prediction result;

[0080] S36: The risk prediction submodule performs risk assessment and prediction model optimization based on the reference fault prediction result by applying trend analysis and behavior pattern recognition to obtain a risk prediction result.

[0081] Among them, the conventional data comparison submodule can compare the current abnormal node with the previous fluctuation data based on the abnormal node location information, and use time series analysis to identify pattern consistency and key deviations, which can include data synchronization and deviation measurement, and further evaluate similarities and abnormalities to construct data comparison analysis results.

[0082] Based on the abnormal node location information, the real-time data of the current abnormal node and the stored historical fluctuation data can be extracted, and the data can be preprocessed, which may include noise removal and normalization processing; secondly, the time series analysis method can be used to compare data points point by point through the sliding window technology. The mean square error (MSE) between the current data and the historical data is calculated for each window to evaluate the pattern consistency and key deviations at each time point. The deviation threshold method can be used to determine the degree of abnormality, and each deviation value is compared with the preset threshold to construct data comparison analysis results to evaluate similarity and abnormality, which may include data synchronization and deviation measurement.

[0083] Optionally, the calculation formula for the data comparison analysis result can be as follows:

[0084]

[0085] Among them, MSE represents the result of data comparison analysis; n represents the total number of data points in the window; i represents the data point index; Xa i Represents real-time data points, Indicates the historical data point corresponding to the real-time data point.

[0086] The fault trend analysis submodule can cluster the data points according to the data comparison and analysis results by drawing fluctuation maps and trend lines to reveal the development path and pattern of potential faults, including map updates and fault prediction model adjustments, thereby generating a fault trend map.

[0087] Based on the data comparison and analysis results, we can use the failure modes in historical data to compare with current data and apply kernel density estimation to perform cluster analysis of data points. By identifying high-density areas in the data set, we can reveal potential failure development paths. We can also compare historical trend data with current trend data, update the map to reflect the latest operating status, draw fluctuation maps and trend lines, and use visual tools to intuitively display data clustering results and potential failure modes, which can include map updates and fault prediction model adjustments to generate fault trend maps.

[0088] The risk prediction submodule can use a time series prediction algorithm based on the fault trend map to predict fault nodes and times in future time periods, and perform risk assessment and prediction model optimization by applying trend analysis and behavior pattern recognition to establish and obtain risk prediction results.

[0089] Based on the fault trend map, by integrating the fault trend analysis results, the weighted moving average method can be used to predict the fault nodes and time that will occur in the future time period, analyze the behavior patterns in the historical data of each node, and calculate the future risk level based on the development trend of the pattern. By applying trend analysis and behavior pattern recognition, various influencing factors are comprehensively considered for risk assessment, and the prediction model is optimized to establish risk prediction results.

[0090] Optionally, the risk prediction submodule may implement the step of predicting the fault nodes and times in a future time period based on the fault trend map and using a time series prediction algorithm to obtain a reference fault prediction result by the following formula:

[0091]

[0092] Among them, X t represents the value at time t in the time series, i.e., the reference fault prediction result; c represents the constant term; α represents the trend term T t-1 Adjustment factor; T t-1 represents the trend term of the previous period; p represents the order of the autoregressive term; i represents the order index of the autoregressive term; φ i represents the coefficient of the autoregressive term; X t-i Represents the value at time ti in the time series; β i represents the adjustment coefficient of the autoregressive term coefficient; q represents the order of the moving average term; j represents the moving average term order index; θ i represents the coefficient of the moving average term; ∈ t-j represents the random error term at time tj; γ i represents the adjustment coefficient of the moving average term coefficient; δ represents the seasonal adjustment parameter; D t Indicator variable representing seasonal factors; ∈ t represents the random error term at time t.

[0093] The trend term T of the previous period can be calculated t-1 , introduce it into the model as an independent variable to improve the sensitivity to time series trends, and introduce the autoregressive term φ i X t-i and the moving average term θ j ∈ t-j , and introduce an adjustment coefficient β i and γ j , in order to dynamically adjust the influence of historical data on the current forecast, add the seasonal adjustment parameter δ, multiply it by the seasonal indicator variable D t , to consider the impact of seasonal factors on time series, calculate the model prediction result X tThe specific method steps of the weight coefficient include using historical data to optimize the parameters, such as determining the best β by minimizing the sum of squares of the prediction errors. i , γ j and δ, integrating the above steps to form a comprehensive risk prediction model.

[0094] S40: The strategy formulation and response module formulates operation instructions based on the risk prediction results to obtain preventive measures.

[0095] The strategy formulation and response module can formulate operation instructions based on the risk prediction results, which may include operations such as adjusting loads, switching lines, and reconfiguring grid resources to avoid potential failures, thereby obtaining preventive measures. Among them, the preventive measures include but are not limited to: emergency response level, resource allocation details, and key operation points.

[0096] It should be understood that the strategy formulation and response module formulates operation instructions based on the risk prediction results to obtain preventive measures, which refers to the process of how to obtain preventive measures. Among them, in step S40, that is, the strategy formulation and response module formulates operation instructions based on the risk prediction results to obtain preventive measures, which may include the following steps:

[0097] S41: The operation instruction formulation submodule adjusts the grid load distribution and formulates the emergency switch operation process based on the risk prediction result, and constructs an operation adjustment plan;

[0098] S42: the resource adjustment submodule implements the optimal configuration of power grid resources based on the operation adjustment plan to obtain a reference configuration result;

[0099] S43: the resource adjustment submodule adjusts and matches the predicted fault scenario according to the reference configuration result to generate a resource configuration result;

[0100] S44: The fault prevention submodule uses the resource configuration result to optimize the power grid operation parameters and perform early warning upgrades to obtain fault prevention measures.

[0101] Among them, the operation instruction formulation submodule can refine the operation instructions based on the risk prediction results, analyze the load demand, adjust the power grid load distribution according to the predicted risks, and formulate the emergency switch operation process, so as to construct an operation adjustment plan.

[0102] Based on the risk prediction results, we can first obtain the predicted high-risk areas and time periods from the data center, analyze the historical data and predicted data of power load in specific areas, calculate the optimal load distribution strategy of the power grid in different time periods based on the data, and use the linear programming model to adjust the load distribution to balance the overall energy efficiency and safety. By focusing on load adjustments during periods with high predicted risks, we can design specific operating instructions, such as starting backup power supplies and adjusting the load of the main power grid. We also formulate emergency switch operation procedures, such as switching instructions between manual and automatic modes, and finally build an operation adjustment plan.

[0103] Optionally, the formula for load sharing adjustment can be as follows:

[0104] L d =k b ×(P req -P sup )

[0105] Among them, L d Represents the adjusted load distribution; P req represents the predicted total load demand; P sup Indicates the existing supply capacity; k b Represents a factor adjusted based on risk assessment to ensure adequate power support during high-risk periods.

[0106] The resource adjustment submodule can implement the optimal configuration of power grid resources based on the operation adjustment plan, such as line switching and reallocation of key resources, and analyze the resource utilization efficiency under the current operating environment to adjust and match the predicted fault scenarios, thereby generating resource configuration results.

[0107] Based on the operation adjustment plan, the current grid resource configuration can be evaluated and optimized according to the operation adjustment plan. By using the resource configuration model, the power lines in the power grid can be dynamically switched to optimize the allocation of key resources, such as the reallocation of transformers and distribution stations, to deal with predicted high-risk events, and calculate the optimal solution for resource configuration. According to the current operating environment and historical fault data, resources are adjusted to match the predicted fault scenario to generate resource configuration results.

[0108] Optionally, the calculation formula for resource optimization configuration can be as follows:

[0109] R c =f(P cur , P opt )

[0110] Among them, R c Indicates the result of resource allocation; P cur Indicates the current resource status; P optrepresents the resource configuration state after optimization based on fault prediction; f represents the resource configuration function, which is used to consider the difference between the current state and the optimization target.

[0111] The fault prevention submodule can utilize the resource configuration results to further optimize power grid operation parameters and strengthen supervision, such as regular inspections and early warning upgrades, to avoid expected faults, thereby obtaining preventive measures.

[0112] The resource allocation results can be used to integrate optimized resource allocation and real-time monitoring data to regularly adjust grid operating parameters, such as voltage and frequency regulation, to maintain network stability and efficiency, implement enhanced supervision measures, such as upgraded fault detection and early warning, and regularly conduct line and equipment inspections, through measures aimed at minimizing and avoiding expected failures, and obtaining preventive measures.

[0113] Alternatively, the formula for the fault prevention adjustment may be as follows:

[0114] F p =γ c ×O p

[0115] Among them, F p Indicates the effect of fault prevention; p represents the optimized operating parameters; γ c It represents the prevention effect coefficient adjusted according to historical failure data, and can be used to indicate the ability to prevent failures by optimizing operating parameters.

[0116] S50: The fault simulation and testing module uses the fault prevention measures to perform virtual fault simulation, and iteratively optimizes the strategy to obtain a test evaluation result.

[0117] The fault simulation and testing module can use the fault prevention measures to perform virtual fault simulation, and adjust the operating parameters through simulation feedback, and iteratively optimize the strategy to obtain test evaluation results. The test evaluation results include but are not limited to: simulation accuracy, prediction deviation, and operation response time.

[0118] It should be understood that the fault simulation and test module uses the fault prevention measures to perform virtual fault simulation, and iteratively optimizes the strategy to obtain the test evaluation result, which refers to the process of how to obtain the test evaluation result. Among them, in step S50, that is, the fault simulation and test module uses the fault prevention measures to perform virtual fault simulation, and iteratively optimizes the strategy to obtain the test evaluation result, which may include the following steps:

[0119] S51: the parameter adjustment submodule performs a virtual fault simulation according to the fault prevention measures to obtain simulation feedback optimization parameters;

[0120] S52: the parameter adjustment submodule constructs an operation parameter adjustment table according to the simulation feedback optimization parameters;

[0121] S53: the strategy iteration submodule optimizes the fault response strategy using the operation parameter adjustment table to obtain a strategy optimization result;

[0122] S54: The strategy optimization and evaluation submodule uses the Bayesian optimization algorithm to adjust the test parameters based on the strategy optimization result to obtain the test evaluation result.

[0123] The parameter adjustment submodule can perform virtual fault simulation according to the fault prevention measures to adjust the operating parameters, such as the adjustment of the fault response speed and the fault detection threshold, so as to optimize the parameters according to the simulation feedback, and construct an operating parameter adjustment table with reference to the response time and accuracy.

[0124] Virtual fault simulation can be performed based on the preventive measures, virtual fault events can be generated through a high-fidelity simulation platform, and the fault response speed and fault detection threshold can be tested in a controlled environment. According to the simulation results, the relationship between response time and fault detection accuracy can be analyzed, and the operating parameters can be adjusted to reduce response delays and improve fault detection accuracy. After each simulation, the operating parameters can be gradually adjusted according to the difference between the actual and expected results, the fault response and detection threshold can be optimized, and a detailed operating parameter adjustment table can be generated. Among them, the operating parameter adjustment table can list the optimal parameter settings under different fault types and conditions.

[0125] Optionally, the formula for adjusting the operating parameters can be as follows:

[0126]

[0127] Among them, P adj Indicates the adjusted operating parameter value; T resp，i represents the response time of the ith simulation; T acc，i represents the detection accuracy corresponding to the i-th simulation; N c represents the number of simulations; i represents the simulation number index.

[0128] The strategy iteration submodule can use the data in the operation parameter adjustment table to optimize the fault response strategy, such as adjusting the fault isolation and recovery order, and adjust the strategy details according to the results of each iteration to optimize the decision accuracy and response time to obtain the strategy optimization result.

[0129] The data in the operating parameter adjustment table can be used to optimize the fault response strategy using the adjustment results of the operating parameters. By simulating different fault conditions, the fault isolation and recovery sequence can be adjusted to ensure that system downtime and resource waste are minimized. According to the results of each iteration, the module adjusts the strategy in detail to gradually improve the accuracy of decision-making and response time to achieve the most optimized fault handling process. Through repeated iterations, the strategy optimization results are obtained to ensure that faults can be responded to and handled quickly and effectively in actual applications.

[0130] Optionally, the calculation formula for strategy optimization can be as follows:

[0131]

[0132] Among them, S opt Represents the optimized strategy result; τ j represents the total time of fault handling in the jth iteration; J represents the number of iterations, and j represents the iteration number index.

[0133] The strategy optimization and evaluation submodule can perform multiple rounds of test evaluation through the strategy optimization results, such as using a Bayesian optimization algorithm, wherein each round of testing can adjust the test parameters based on the evaluation results of the previous round, thereby summarizing the test data and establishing and obtaining the test evaluation results.

[0134] The results of strategy optimization can be achieved through a series of test evaluation processes to verify and refine the fault handling strategy. By building a learning model based on the results of previous iterative tests, multiple rounds of testing are performed, and the test parameters are adjusted in each round based on the evaluation results of the previous round to refine the strategy and improve the effectiveness of fault handling. In this way, test evaluation results are established to help ultimately determine the most effective fault response and resource allocation strategy.

[0135] Optionally, the strategy optimization and evaluation submodule may implement the step of adjusting the test parameters using the Bayesian optimization algorithm according to the strategy optimization result to obtain the test evaluation result through the following formula:

[0136] μ(x)=k(x,θ) T (K(θ)+σ n 2 I+λL) -1 y

[0137] Where μ(x) represents the predicted mean at position x, i.e., the test evaluation result; k(x, θ) represents the covariance vector of x and the training data point adjusted according to the parameter θ; k(x, θ) T represents the transposition operation of k(x, θ); K(θ) represents the covariance matrix between the training data points adjusted according to the parameter θ; σ nRepresents the noise term; I represents the identity matrix; λ represents the regularization coefficient; L represents the Laplace matrix; y represents the observed value of the objective function at the training data point.

[0138] Optionally, the parameter θ is optimized by maximizing the marginal likelihood function to ensure that the kernel function simulates the behavior of the objective function as accurately as possible. The parameter λ is determined by cross-validation to balance the effects of fitting and regularization. The Laplace matrix L is constructed based on the spatial relationship of the training data points to reflect the geographic or functional similarities between the data points. These parameters act together on the kernel matrix K and the covariance vector k(x, θ). The predicted objective function value at location x is calculated through the updated prediction model. The above steps are combined to obtain the optimal parameter combination.

[0139] S60: The fault warning and optimization module optimizes the fault prevention strategy according to the test evaluation result to obtain the target fault warning and target response strategy.

[0140] The fault warning and optimization module can optimize the fault prevention strategy based on the test evaluation results, such as updating the training materials of the operation team, and updating the operation team in real time through the network, so as to obtain the adjusted fault warning and response strategy, that is, the target fault warning and target response strategy. The adjusted fault warning and response strategy includes but is not limited to: warning start time, key response team, and warning communication path.

[0141] It should be understood that the fault warning and optimization module optimizes the fault prevention strategy according to the test evaluation results to obtain the target fault warning and target response strategy, which refers to the process of how to obtain the target fault warning and target response strategy. Among them, in step S60, that is, the fault warning and optimization module optimizes the fault prevention strategy according to the test evaluation results to obtain the target fault warning and target response strategy, which may include the following steps:

[0142] S61: The fault detection submodule performs operation log analysis based on the test evaluation result to obtain abnormal frequency and abnormal operation;

[0143] S62: the fault detection submodule classifies and marks the abnormalities according to the abnormal frequency and the abnormal operation to obtain an abnormal classification and marking table;

[0144] S63: the education update submodule uses the abnormal classification mark table to conduct training on fault response to obtain an updated training material set;

[0145] S64: The strategy updating submodule adjusts and optimizes the fault warning and response based on the updated training material set to obtain a reference fault warning and reference response strategy;

[0146] S65: The strategy updating submodule performs simulation exercises and verifies the implementation effect of the reference response strategy according to the reference fault warning and the reference response strategy, and obtains a target fault warning and a target response strategy.

[0147] Among them, the fault detection submodule can perform operation log analysis based on the test evaluation results to identify the abnormal frequency and impact of operational anomalies, and classify and mark the anomalies in combination with performance monitoring data, such as recording the manifestation and frequency of each anomaly, thereby constructing an anomaly classification and marking table.

[0148] Based on the test evaluation results, operation logs can be extracted and combined with performance monitoring data to perform frequency analysis on operation anomalies in the logs. Automated tools can be used to identify abnormal patterns and distinguish normal operation fluctuations from real operation anomalies. For each type of identified anomaly, the module records its manifestation, frequency and impact in detail, uses machine learning technology to classify anomalies, automatically marks the type of each anomaly, and generates an anomaly classification marking table, which can include a specific description of the anomaly and the frequency of occurrence, providing a reference for subsequent processing.

[0149] Optionally, the calculation formula for abnormal classification can be as follows:

[0150] C e =f(D log ,D perf )

[0151] Among them, C e Indicates the abnormal type after classification; D log Indicates operation log data; D perf represents performance monitoring data; f represents a classification function, which can be used to convert log and performance data into abnormal classification labels.

[0152] The education update submodule can use the anomaly classification tag table to recompile the training materials and operation guides of the operation team, and conduct case analysis on the identified anomalies to implement training on fault response, thereby obtaining an updated set of training materials.

[0153] The exception classification marking table can be used to recompile the operation team's training materials and operation guidelines according to the marked exception types, analyze each exception case, and develop targeted training plans to enhance the team's understanding and reaction speed of fault response. Through analysis of actual cases and simulated fault responses, the training materials are updated modularly to ensure that the team can quickly identify and effectively handle various exceptions in actual operations, and ultimately obtain an updated set of training materials.

[0154] The strategy update submodule can adjust and optimize fault warning and response based on the updated training material set, and organize simulation exercises through the internal network to verify the implementation effect of the strategy in real time, so as to obtain the adjusted fault warning and response strategy, that is, the target fault warning and target response strategy.

[0155] Based on the updated set of training materials, the fault warning and response strategies can be adjusted and optimized, and new fault response strategies can be tested by organizing simulation exercises, such as collecting exercise data in real time through the internal network to evaluate the effectiveness and implementation effect of the strategy. The collected data can be used to make minor adjustments to the strategy to ensure that various fault situations can be warned and responded to efficiently and accurately in actual operations, thereby establishing an adjusted fault warning and response strategy.

[0156] Optionally, the calculation formula for policy adjustment can be as follows:

[0157] S new =g(S old ,R sim )

[0158] Among them, S new represents the adjusted strategy; S old Represents the original strategy; R sim represents the result data of the simulation exercise; g represents the strategy adjustment function, which can be used to optimize the original strategy according to the exercise results.

[0159] This application monitors the voltage, current and load of the power grid and compares them with historical behavior data to achieve accurate identification of data point anomalies and threshold analysis, which can improve the accuracy and response speed of fault prediction. On the basis of real-time monitoring of multi-node current and voltage, it can effectively detect and record data fluctuations that exceed safety standards. The strategy significantly improves the efficiency of fault location. By comparing the abnormal node location information with historical data, it can evaluate future fault trends and predict fault nodes and times through statistical tools to enhance the comprehensiveness and foresight of the prediction.

[0160] Based on the risk prediction results, operational instructions such as load adjustment, line switching and grid resource configuration can effectively avoid the occurrence of predicted faults and greatly improve the stability and safety of grid operation. Through iterative optimization of fault simulation and testing, as well as adjustment of fault warning and response strategies, the maintenance team can enhance its adaptability to strategy execution and operational accuracy, thereby ensuring efficient and dynamic updating of grid fault management.

[0161] It can be seen that in the above scheme, the initial monitoring module, by monitoring the voltage, current and load of the power grid, identifies and marks abnormal data points, and obtains abnormal pattern identification results, so that the load fluctuation analysis module can locate the abnormal node according to the abnormal pattern identification results and obtain abnormal node location information, and further enable the trend analysis and prediction module to evaluate and predict the fault node and time according to the abnormal node location information to obtain risk prediction results, so that the strategy formulation and response module can formulate operation instructions based on the risk prediction results to obtain fault prevention measures, so that the fault simulation and testing module can use the fault prevention measures to perform virtual fault simulation, and iteratively optimize the strategy to obtain test evaluation results, and then enable the fault warning and optimization module to optimize the fault prevention strategy according to the test evaluation results, and obtain accurate target fault warning and fast and effective target response strategy, which can effectively improve the accuracy of power grid fault prediction and the speed of fault response, and is conducive to achieving more accurate and efficient power grid fault warning.

[0162] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0163] In one embodiment, a machine learning-based power grid fault warning device is provided, and the machine learning-based power grid fault warning device corresponds one-to-one to the machine learning-based power grid fault warning system in the above embodiment. Figure 3 As shown, the machine learning-based power grid fault early warning device includes a first processing module 101, a second processing module 102, a third processing module 103, a fourth processing module 104, a fifth processing module 105 and a sixth processing module 106. Each functional module is described in detail as follows:

[0164] The first processing module 101 is used to identify and mark abnormal data points by monitoring the voltage, current and load of the power grid to obtain abnormal pattern identification results;

[0165] The second processing module 102 is used to locate the abnormal node according to the abnormal pattern identification result to obtain abnormal node location information;

[0166] The third processing module 103 is used to evaluate and predict the fault node and time according to the abnormal node location information to obtain a risk prediction result;

[0167] The fourth processing module 104 is used to formulate an operation instruction based on the risk prediction result to obtain a preventive measure for a fault;

[0168] The fifth processing module 105 is used to perform virtual fault simulation using the fault prevention measures, and iteratively optimize the strategy to obtain a test evaluation result;

[0169] The sixth processing module 106 is used to optimize the fault prevention strategy according to the test evaluation result to obtain the target fault warning and target response strategy.

[0170] In one embodiment, the first processing module 101 is used to identify and mark abnormal data points by monitoring the voltage, current and load of the power grid to obtain abnormal pattern identification results, specifically for:

[0171] Sampling the voltage, current and load of the power grid to obtain sampling data;

[0172] Performing frequency analysis and stability testing on the sampled data to generate a stability evaluation result;

[0173] Dynamically track the power grid load based on the stability assessment results to obtain load data;

[0174] Performing periodic analysis on the load data to obtain load consistency analysis results;

[0175] Utilizing the load consistency analysis result, analyzing the abnormal pattern to obtain abnormal data;

[0176] A data pattern comparison is performed on the abnormal data to obtain an abnormal pattern identification result.

[0177] In one embodiment, the second processing module 102 is used to locate the abnormal node according to the abnormal pattern identification result to obtain abnormal node location information, specifically for:

[0178] Based on the abnormal mode identification result, the operation status of the power grid is monitored in real time to obtain a real-time data snapshot;

[0179] Extracting current and voltage data from the real-time data snapshot, identifying abnormal fluctuations, and generating excessive fluctuation analysis results;

[0180] The excessive fluctuation analysis results are used to identify and locate nodes with abnormal fluctuations, and to establish abnormal node location information.

[0181] In one embodiment, the third processing module 103 is used to evaluate and predict the fault node and time according to the abnormal node location information to obtain a risk prediction result, specifically for:

[0182] Based on the abnormal node location information, obtain the current abnormal node and historical fluctuation data;

[0183] According to the current abnormal node and the historical fluctuation data, time series analysis is used to identify the consistency and key deviation of the pattern to obtain similarity data and abnormality data;

[0184] Constructing and obtaining data comparison and analysis results based on the similarity data and the abnormality data;

[0185] According to the data comparison and analysis results, the data points are clustered to generate a fault trend map;

[0186] Based on the fault trend graph, a time series prediction algorithm is used to predict the fault nodes and times in the future time period to obtain a reference fault prediction result;

[0187] According to the reference fault prediction results, trend analysis and behavior pattern recognition are applied to perform risk assessment and prediction model optimization to obtain risk prediction results.

[0188] In one embodiment, the fourth processing module 104 is used to formulate an operation instruction based on the risk prediction result to obtain a fault prevention measure, specifically for:

[0189] Based on the risk prediction results, adjust the grid load distribution and formulate emergency switch operation procedures to build an operation adjustment plan;

[0190] Based on the operation adjustment plan, the optimal configuration of power grid resources is implemented to obtain a reference configuration result;

[0191] According to the reference configuration result, adjusting and matching the predicted fault scenario to generate a resource configuration result;

[0192] The fault prevention submodule utilizes the resource configuration result to optimize the power grid operation parameters and perform early warning upgrades to obtain fault prevention measures.

[0193] In one embodiment, the fifth processing module 105 is used to perform virtual fault simulation using the fault prevention measures and iteratively optimize the strategy to obtain a test evaluation result, specifically for:

[0194] Perform virtual fault simulation according to the fault prevention measures to obtain simulation feedback optimization parameters;

[0195] Constructing an operating parameter adjustment table according to the simulation feedback optimization parameters;

[0196] The operation parameter adjustment table is used to optimize the fault response strategy to obtain a strategy optimization result;

[0197] Based on the strategy optimization results, the Bayesian optimization algorithm is used to adjust the test parameters to obtain the test evaluation results.

[0198] In one embodiment, the fifth processing module 105 is specifically configured to:

[0199] The following formula is used to optimize the results according to the strategy, adjust the test parameters using the Bayesian optimization algorithm, and obtain the test evaluation results:

[0200] μ(x)=k(x,θ) T (K(θ)+σ n 2 I+λL) -1 y

[0201] Where μ(x) represents the predicted mean at position x, i.e., the test evaluation result; k(x, θ) represents the covariance vector of x and the training data point adjusted according to the parameter θ; k(x, θ) T represents the transposition operation of k(x, θ); K(θ) represents the covariance matrix between the training data points adjusted according to the parameter θ; σ n Represents the noise term; I represents the identity matrix; λ represents the regularization coefficient; L represents the Laplace matrix; y represents the observed value of the objective function at the training data point.

[0202] In one embodiment, the sixth processing module 106 is used to optimize the fault prevention strategy according to the test evaluation result to obtain the target fault warning and target response strategy, specifically for:

[0203] Based on the test evaluation results, analyze the operation logs to obtain abnormal frequencies and abnormal operations;

[0204] Classify and mark the abnormalities according to the abnormal frequency and the abnormal operation to obtain an abnormal classification and marking table;

[0205] Using the abnormal classification marking table, training on fault response is performed to obtain an updated training material set;

[0206] Based on the updated training material set, fault warning and response are adjusted and optimized to obtain reference fault warning and reference response strategies;

[0207] According to the reference fault warning and the reference response strategy, a simulation exercise is carried out and the implementation effect of the reference response strategy is tested to obtain a target fault warning and a target response strategy.

[0208] The present invention provides a power grid fault early warning device based on machine learning. By monitoring the voltage, current and load of the power grid, abnormal data points are identified and marked to obtain abnormal pattern identification results. Abnormal nodes can be located according to the abnormal pattern identification results to obtain abnormal node positioning information. Fault nodes and times are further evaluated and predicted according to the abnormal node positioning information to obtain risk prediction results. Therefore, operation instructions can be formulated based on the risk prediction results to obtain fault prevention measures. Virtual fault simulation can be performed using the fault prevention measures, and the strategy is iteratively optimized to obtain test evaluation results. Further, the fault prevention strategy can be optimized according to the test evaluation results to obtain accurate target fault early warning and fast and effective target response strategy. The accuracy of power grid fault prediction and the speed of fault response can be effectively improved, which is conducive to achieving the purpose of more accurate and efficient power grid fault early warning.

[0209] For the specific definition of the power grid fault warning device based on machine learning, please refer to the definition of the power grid fault warning system based on machine learning above, which will not be repeated here. Each module in the above-mentioned power grid fault warning device based on machine learning can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0210] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client through a network connection. When the computer program is executed by the processor, a function or step on the service side of a power grid fault warning system based on machine learning is implemented.

[0211] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 5As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When the computer program is executed by the processor, a function or step on the client side of a power grid fault early warning system based on machine learning is implemented.

[0212] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0213] By monitoring the voltage, current and load of the power grid, abnormal data points are identified and marked to obtain abnormal pattern identification results;

[0214] Locating the abnormal node according to the abnormal pattern identification result to obtain abnormal node location information;

[0215] Evaluate and predict the fault node and time according to the abnormal node location information to obtain a risk prediction result;

[0216] Formulate operation instructions based on the risk prediction results to obtain preventive measures for failures;

[0217] The fault prevention measures are used to perform virtual fault simulation, and the strategies are iteratively optimized to obtain test evaluation results;

[0218] According to the test evaluation results, the fault prevention strategy is optimized to obtain the target fault warning and target response strategy.

[0219] The present invention provides a computer device, which monitors the voltage, current and load of the power grid, identifies and marks abnormal data points, obtains abnormal pattern identification results, can locate abnormal nodes according to the abnormal pattern identification results, obtain abnormal node location information, further evaluate and predict fault nodes and times according to the abnormal node location information, and obtain risk prediction results, so that operation instructions can be formulated based on the risk prediction results to obtain fault prevention measures, and virtual fault simulation can be performed using the fault prevention measures, and the strategy can be iteratively optimized to obtain test evaluation results, and then the fault prevention strategy can be optimized according to the test evaluation results to obtain accurate target fault warnings and fast and effective target response strategies, which can effectively improve the accuracy of power grid fault predictions and the speed of fault response, and is conducive to achieving more accurate and efficient power grid fault warnings.

[0220] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0221] By monitoring the voltage, current and load of the power grid, abnormal data points are identified and marked to obtain abnormal pattern identification results;

[0222] Locating the abnormal node according to the abnormal pattern identification result to obtain abnormal node location information;

[0223] Evaluate and predict the fault node and time according to the abnormal node location information to obtain a risk prediction result;

[0224] Formulate operation instructions based on the risk prediction results to obtain preventive measures for failures;

[0225] The fault prevention measures are used to perform virtual fault simulation, and the strategies are iteratively optimized to obtain test evaluation results;

[0226] According to the test evaluation results, the fault prevention strategy is optimized to obtain the target fault warning and target response strategy.

[0227] The present invention provides a computer-readable storage medium, which monitors the voltage, current and load of a power grid, identifies and marks abnormal data points, obtains abnormal pattern identification results, can locate abnormal nodes according to the abnormal pattern identification results, obtain abnormal node location information, further evaluate and predict fault nodes and times according to the abnormal node location information, obtain risk prediction results, so that operation instructions can be formulated based on the risk prediction results, fault prevention measures can be obtained, virtual fault simulation can be performed using the fault prevention measures, and the strategy can be iteratively optimized to obtain test evaluation results, and then the fault prevention strategy can be optimized according to the test evaluation results to obtain accurate target fault warnings and fast and effective target response strategies, which can effectively improve the accuracy of power grid fault predictions and the speed of fault response, and is conducive to achieving more accurate and efficient power grid fault warnings.

[0228] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can refer to the relevant descriptions on the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0229] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0230] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0231] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A power grid fault early warning system based on machine learning, characterized in that: The machine learning-based power grid fault early warning system includes an initial monitoring module, a load fluctuation analysis module, a trend analysis and prediction module, a strategy formulation and response module, a fault simulation and testing module, and a fault early warning and optimization module, wherein: The initial monitoring module monitors the voltage, current and load of the power grid, identifies and marks abnormal data points, and obtains abnormal pattern identification results; The load fluctuation analysis module locates the abnormal node according to the abnormal pattern identification result to obtain abnormal node location information; The trend analysis and prediction module evaluates and predicts the fault node and time according to the abnormal node location information to obtain a risk prediction result; The strategy formulation and response module formulates operation instructions based on the risk prediction results to obtain preventive measures for failures; The fault simulation and testing module uses the fault prevention measures to perform virtual fault simulation and iteratively optimizes the strategy to obtain a test evaluation result; The fault warning and optimization module optimizes the fault prevention strategy according to the test evaluation result to obtain the target fault warning and target response strategy.

2. The power grid fault early warning system based on machine learning according to claim 1 is characterized in that: The initial monitoring module includes a voltage and current monitoring submodule, a load analysis submodule and an abnormal pattern recognition submodule, wherein: The voltage and current monitoring submodule samples the voltage, current and load of the power grid to obtain sampled data; The voltage and current monitoring submodule performs frequency analysis and stability detection on the sampled data to generate a stability evaluation result; The load analysis submodule dynamically tracks the power grid load through the stability evaluation result to obtain load data; The load analysis submodule performs periodic analysis on the load data to obtain a load consistency analysis result; The abnormal pattern recognition submodule uses the load consistency analysis result to analyze the abnormal pattern and obtain abnormal data; The abnormal pattern recognition submodule performs data pattern comparison on the abnormal data to obtain an abnormal pattern identification result.

3. The machine learning-based power grid fault early warning system according to claim 2, characterized in that: The load fluctuation analysis module includes a real-time monitoring submodule, a fluctuation detection submodule and a node positioning submodule, wherein: The real-time monitoring submodule monitors the operation status of the power grid in real time based on the abnormal mode identification result to obtain a real-time data snapshot; The fluctuation detection submodule extracts current and voltage data from the real-time data snapshot, identifies abnormal fluctuations, and generates an excessive fluctuation analysis result; The node positioning submodule uses the excessive fluctuation analysis result to identify and locate nodes with abnormal fluctuations and establish abnormal node positioning information.

4. The machine learning-based power grid fault early warning system according to claim 3, characterized in that: The trend analysis and prediction module includes a conventional data comparison submodule, a fault trend analysis submodule and a risk prediction submodule, wherein: The conventional data comparison submodule obtains the current abnormal node and historical fluctuation data based on the abnormal node location information; The conventional data comparison submodule uses time series analysis to identify the consistency and key deviation of the pattern according to the current abnormal node and the historical fluctuation data, and obtains similarity data and abnormality data; The conventional data comparison submodule constructs and obtains data comparison and analysis results based on the similarity data and the abnormality data; The fault trend analysis submodule clusters the data points according to the data comparison and analysis results to generate a fault trend graph; The risk prediction submodule predicts the fault nodes and times in the future time period based on the fault trend map and adopts a time series prediction algorithm to obtain a reference fault prediction result; The risk prediction submodule performs risk assessment and prediction model optimization based on the reference fault prediction result by applying trend analysis and behavior pattern recognition to obtain a risk prediction result.

5. The machine learning-based power grid fault early warning system according to claim 4, characterized in that: The strategy formulation and response module includes an operation instruction formulation submodule, a resource adjustment submodule, and a fault prevention submodule: The operation instruction formulation submodule adjusts the grid load distribution and formulates the emergency switch operation process based on the risk prediction result, and constructs an operation adjustment plan; The resource adjustment submodule implements the optimal configuration of power grid resources based on the operation adjustment scheme to obtain a reference configuration result; The resource adjustment submodule adjusts and matches the predicted fault scenario according to the reference configuration result to generate a resource configuration result; The fault prevention submodule utilizes the resource configuration result to optimize the power grid operation parameters and perform early warning upgrades to obtain fault prevention measures.

6. The machine learning-based power grid fault early warning system according to claim 5, characterized in that: The fault simulation and testing module includes a parameter adjustment submodule, a strategy iteration submodule and a strategy optimization and evaluation submodule, wherein: The parameter adjustment submodule performs virtual fault simulation according to the fault prevention measures to obtain simulation feedback optimization parameters; The parameter adjustment submodule constructs an operation parameter adjustment table according to the simulation feedback optimization parameters; The strategy iteration submodule optimizes the fault response strategy using the operation parameter adjustment table to obtain a strategy optimization result; The strategy optimization and evaluation submodule uses the Bayesian optimization algorithm to adjust the test parameters based on the strategy optimization results to obtain the test evaluation results.

7. The machine learning-based power grid fault early warning system according to claim 6, characterized in that: in, The strategy optimization and evaluation submodule uses the following formula to adjust the test parameters according to the strategy optimization results and obtain the test evaluation results using the Bayesian optimization algorithm: μ(x)=k(x,θ) T (K(θ)+σ n 2 I+λL) -1 y Among them, μ(x) represents the predicted mean at position x, that is, the test evaluation result; k(x,θ) represents the covariance vector of x and the training data point adjusted according to the parameter θ; k(x,θ) T represents the transposition operation of k(x,θ); K(θ) represents the covariance matrix between the training data points adjusted according to the parameter θ; σ n Represents the noise term; I represents the identity matrix; λ represents the regularization coefficient; L represents the Laplace matrix; y represents the observed value of the objective function at the training data point.

8. The machine learning-based power grid fault early warning system according to any one of claims 1 to 7, characterized in that: The fault warning and optimization module includes a fault detection submodule, an education update submodule and a strategy update submodule, wherein: The fault detection submodule performs operation log analysis based on the test evaluation result to obtain abnormal frequency and abnormal operation; The fault detection submodule classifies and marks the abnormalities according to the abnormal frequency and the abnormal operation to obtain an abnormal classification and marking table; The education update submodule uses the abnormal classification mark table to conduct training on fault response to obtain an updated training material set; The strategy updating submodule adjusts and optimizes the fault warning and response based on the updated training material set to obtain a reference fault warning and reference response strategy; The strategy updating submodule performs simulation exercises and verifies the implementation effect of the reference response strategy according to the reference fault warning and the reference response strategy, and obtains a target fault warning and a target response strategy.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the processing method of the power grid fault early warning system based on machine learning is implemented as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processing method of the power grid fault early warning system based on machine learning as described in any one of claims 1 to 8 is implemented.

Images