Power system fault prediction and self-healing control method and system

Abstract

The invention relates to the field of power systems, in particular to a power system fault prediction and self-healing control method and system. The method comprises the steps of monitoring electrical quantity of a power system in real time to obtain a monitoring data set; predicting potential fault points and fault types based on the monitoring data set; carrying out key monitoring on the area where the potential fault point is located; once abnormity or fault signs are found in key monitoring, fault diagnosis is carried out, the fault position is determined, according to a fault diagnosis report, a fault area is isolated, an instruction is generated and controlled, power supply of a non-fault area is automatically recovered, and a recovery operation log is recorded; a fault recovery strategy is automatically generated, and a control instruction is updated; and adjusting the operation of the power system according to the updated control instruction, and monitoring the operation state of the power system. According to the invention, the accuracy of fault prediction and the self-healing response speed are improved.

Description

Technical Field

[0001] The present invention belongs to the field of power systems, and in particular relates to a method and system for predicting and self-healing faults in a power system. Background Art

[0002] As the infrastructure of modern society, the stability and reliability of the power system are crucial to economic operation and residents' lives. With the development of technology, the scale and complexity of the power system are increasing, which puts higher requirements on the intelligent management of the power system. Especially when a power system fails, how to quickly and accurately predict and locate the fault, and how to quickly restore power supply to non-fault areas have become key technical problems in power system management.

[0003] Prior art CN112819334A proposes an intelligent fault-tolerant self-healing control method and system for power generation and distribution of a ship's regional power distribution system, aiming to improve the fault-tolerant self-healing capability of the ship's regional power distribution system. However, the system is mainly focused on specific application scenarios of ship power systems. For a wider range of power systems, especially when faced with diverse grid fault types and complex grid structures, its universality and adaptability may be insufficient. In addition, the system's real-time and accuracy in fault prediction, as well as the response speed and efficiency of self-healing control, still need to be further improved.

[0004] Prior art CN116388382A discloses a distribution network self-healing control system and method, which collects data information of the distribution network, analyzes the current operating status of the distribution network and predicts the future operating status to select the corresponding control strategy. Although this method has made certain progress in the self-healing control of the distribution network, in practical applications, how to more accurately predict faults, locate fault areas faster, and how to achieve more efficient fault isolation and power supply restoration are still technical problems that need to be solved. Especially in the context of the continuous expansion of the scale of power grids and the increasingly complex operating environment, the efficiency and effect of existing self-healing control systems are still insufficient when dealing with large-scale power grid faults.

[0005] In summary, the main deficiencies of the existing technologies in power system fault prediction and self-healing control include: insufficient accuracy and real-time performance of fault prediction, slow fault location and isolation, and low efficiency of power supply restoration in non-fault areas. These problems limit the power system's response capability and recovery speed in the face of sudden faults, affecting the continuity and reliability of power supply. Therefore, it is urgent to develop a new power system fault prediction and self-healing control method. Summary of the invention

[0006] The present invention provides a method and system for predicting and self-healing faults in an electric power system, which solves the problems of low fault prediction accuracy and slow self-healing response speed in the prior art.

[0007] A method for predicting and self-healing control of power system faults, comprising:

[0008] Use an intelligent sensor network to monitor the electrical quantities of the power system in real time to obtain a monitoring data set; based on the monitoring data set, use a machine learning model to predict potential fault points and fault types; set up a long-term monitoring plan to focus on monitoring areas where potential fault points are located; once an abnormality or fault sign is found in the key monitoring, use an artificial intelligence algorithm to diagnose the fault, determine the fault location, and generate a fault diagnosis report; according to the fault diagnosis report, dynamically adjust the operating parameters of the power grid equipment, isolate the fault area, and generate control instructions; execute the control instructions, automatically restore the power supply to the non-fault area, and record the recovery operation log; build an intelligent decision-making engine to analyze the power grid status and the recovery operation log in real time, automatically generate a fault recovery strategy, and update the control instructions; according to the updated control instructions, adjust the operation of the power system and monitor the operating status of the power system.

[0009] Optionally, the following steps are also included:

[0010] Based on the monitoring data set and fault diagnosis report, the hybrid simulation mode and dynamic simulation technology are used to simulate various fault conditions in the power system and verify the implementation effect of the control instructions, including:

[0011] Create fault simulations that simulate the types and conditions of faults that may occur in the power system, using the initial data set as simulation input;

[0012] Apply fault simulation to the power system simulation environment, execute control instructions, and record system responses and self-healing operations during the simulation;

[0013] Analyze the simulation results and verify the response speed and recovery efficiency of control instructions to different fault types.

[0014] Optionally, the steps of using a smart sensor network to monitor the electrical quantity of the power system in real time and obtaining a monitoring data set include:

[0015] Deploy smart sensors at key nodes of power transmission lines, substations and distribution networks in the power system, including voltage sensors, current sensors, power sensors, frequency sensors and environmental parameter sensors, to monitor the electrical quantities and equipment status of the power system in real time;

[0016] The following data are collected in real time through smart sensors:

[0017] collecting voltage data of the power system from voltage sensors;

[0018] collecting current data of the power system from current sensors;

[0019] Collecting power data of the power system from power sensors;

[0020] collecting frequency data of the power system from frequency sensors;

[0021] Collect equipment operation status and environmental parameter data from environmental parameter sensors;

[0022] Integrate the collected data into an initial dataset;

[0023] Perform preliminary processing on the initial data set. The specific steps include:

[0024] Use edge computing devices to clean the received raw data to remove noise and outliers;

[0025] Perform feature extraction on the cleaned data to identify key parameters and patterns;

[0026] The extracted features are compressed to obtain a preprocessed data set as a monitoring data set.

[0027] Optionally, the steps of using an artificial intelligence algorithm to perform fault diagnosis, determine the fault location, and generate a fault diagnosis report include:

[0028] Based on the monitoring data set, key features are extracted through adaptive threshold and pattern recognition technology to obtain key feature sets;

[0029] Input the obtained key feature set into the pre-trained deep learning model;

[0030] The deep learning model outputs fault diagnosis results, including fault type and possible causes.

[0031] According to the fault diagnosis results, the graph theory algorithm is combined with the power grid topology structure to locate the fault location;

[0032] The fault diagnosis results and the located fault location are integrated as a fault diagnosis report to guide the adjustment of operating parameters of power grid equipment and the isolation of fault areas.

[0033] Optionally, according to the fault diagnosis report, the operation parameters of the power grid equipment are dynamically adjusted, the fault area is isolated, and the control instructions are generated, specifically including:

[0034] According to the fault diagnosis report, evaluate the impact of the fault on the grid operation, including the fault type, location and possible chain reaction, and obtain a fault assessment report;

[0035] Based on fault diagnosis reports, fault assessment reports, dynamic characteristics of the power grid and historical fault data, an adaptive algorithm is applied to calculate the optimal operating parameters of power grid equipment in real time;

[0036] Based on the fault location information in the fault diagnosis report, combined with the real-time topology and power flow distribution of the power grid, the shortest path algorithm and minimum cut algorithm in graph theory are used to intelligently identify the fault area that needs to be isolated and the isolation plan. The isolation plan includes the specific equipment that needs to be isolated and the operation sequence;

[0037] Automatically configure protection devices in the power grid, including relay protection and automatic reclosing devices, based on intelligently identified fault areas and isolation schemes;

[0038] Combined with the optimized grid equipment operating parameters, fault areas, isolation schemes and protection device configuration schemes, a multi-objective optimization algorithm including a genetic algorithm or a particle swarm optimization algorithm is used to generate control instructions for automatic control of grid equipment to achieve rapid isolation of fault areas and power supply restoration in non-fault areas.

[0039] Optionally, the steps of executing the control instruction, automatically restoring the power supply to the non-faulty area, and recording the restoration operation log specifically include:

[0040] Receive and execute generated control instructions, adjust grid equipment to new operating parameters, and isolate identified fault areas.

[0041] According to the results of the isolation operation performed, the power supply status of the non-fault area is determined, and the power supply path and equipment parameters of the non-fault area are automatically adjusted to bypass the fault area and restore the power supply to the non-fault area;

[0042] During execution, a detailed log is recorded for each operation, including the time of operation, type of operation, equipment involved, and results of the operation;

[0043] Using smart sensor networks to monitor the status of the power system after restoration operations and verify whether power supply has been successfully restored to non-faulty areas;

[0044] If the monitoring results show that the power supply has not been fully restored, the power supply to the non-fault area will be automatically restored. If the power supply is restored, the recovery operation log will be recorded.

[0045] Optionally, construct an intelligent decision engine to analyze the power grid status and the recovery operation log in real time, automatically generate a fault recovery strategy, and update the control instructions; adjust the operation of the power system according to the updated control instructions, and monitor the operation status of the power system, specifically including:

[0046] Collect recovery operation logs, including operation time, operation type, devices involved, and operation results;

[0047] Combining the real-time grid status data set and the recovery operation log data set, the intelligent decision engine uses machine learning algorithms to analyze the data and obtain a grid status analysis report, including the status of the fault area, the recovery status of the non-fault area, and potential risk prediction;

[0048] Based on the grid status analysis report, the intelligent decision engine uses operations research methods to automatically generate fault recovery strategies, including recommended grid equipment operating parameter adjustments and power supply path optimization solutions;

[0049] Based on the generated fault recovery strategy, the intelligent decision engine updates the control instructions to guide further adjustments of the power grid equipment;

[0050] The updated control instructions are sent to the corresponding power grid equipment and their execution is monitored.

[0051] Optionally, the following verification steps are also included:

[0052] After executing the updated control instructions, the power system is monitored again using the smart sensor network to obtain a verification data set to verify whether the power supply in the fault area and the non-fault area has been restored as expected;

[0053] Compare and analyze the obtained verification data set with the expected normal operation parameters of the power system to obtain a verification analysis report;

[0054] If the verification and analysis report shows that the power system has not been fully restored or there are new anomalies, the intelligent decision engine will automatically adjust the fault recovery strategy based on the report and generate and execute new control instructions;

[0055] Continuously monitor the operating status of the power system and repeat fault recovery operations and verification steps as needed until the power system is fully restored to normal operating conditions.

[0056] A power system fault prediction and self-healing control system, comprising:

[0057] Intelligent sensor networks are used to monitor the electrical quantities of the power system in real time, including voltage, current, power, frequency, and the operating status of equipment and environmental parameters, and obtain monitoring data sets;

[0058] Machine learning models are used to predict potential fault points and fault types based on monitoring data sets, and to set up long-term monitoring plans to focus on monitoring areas where potential fault points are located;

[0059] Artificial intelligence algorithm module, used to diagnose faults, determine fault locations, and generate fault diagnosis reports once abnormalities or fault signs are found during key monitoring;

[0060] A control instruction generation module is used to dynamically adjust the operating parameters of the power grid equipment according to the fault diagnosis report, isolate the fault area, and generate control instructions;

[0061] An automatic recovery module is used to execute control instructions, automatically restore power supply to non-faulty areas, and record recovery operation logs;

[0062] Intelligent decision-making engine for real-time analysis of grid status and recovery operation logs, automatic generation of fault recovery strategies, and updating of control instructions;

[0063] The monitoring module is used to adjust the operation of the power system according to the updated control instructions and monitor the operating status of the power system.

[0064] Optionally, the monitoring module is further used to verify whether the power supply to the fault area and the non-fault area has been restored as expected after executing the updated control instructions, and to generate a verification analysis report;

[0065] The intelligent decision-making engine is also used to automatically adjust the fault recovery strategy according to the verification and analysis report, and generate and execute new control instructions if the verification and analysis report shows that the power system has not been fully restored or there are new abnormalities.

[0066] The present invention includes at least one of the following beneficial technical effects:

[0067] The present invention uses an intelligent sensor network to monitor the electrical quantities of the power system in real time and uses a machine learning model to analyze the monitoring data set. This method can accurately predict potential fault points and fault types. Compared with traditional empirical judgment, this data-driven prediction method greatly improves the accuracy of fault prediction.

[0068] The present invention focuses on monitoring potential fault points by setting up a long-term monitoring plan, and when abnormalities or signs of faults are found, it uses artificial intelligence algorithms to quickly diagnose and locate faults and generate fault diagnosis reports. This not only improves the fault response speed, but also enhances the system's self-healing ability and reduces dependence on manual intervention.

[0069] According to the fault diagnosis report, the present invention can dynamically adjust the operating parameters of the power grid equipment, isolate the fault area, and generate control instructions. This real-time parameter adjustment and fault isolation strategy helps to reduce the impact of faults on the power grid and improve the stability and reliability of the power grid.

[0070] After the control instruction is executed, the method can automatically restore the power supply of the non-fault area and record the restoration operation log. This automated restoration operation not only improves the restoration speed, but also reduces the risk of secondary failures caused by human operation errors.

[0071] By building an intelligent decision-making engine, the present invention can analyze the power grid status and recovery operation log in real time, automatically generate fault recovery strategies, and update control instructions. This intelligent decision support system significantly improves the decision-making efficiency and response speed of power grid operation.

[0072] According to the updated control instructions, the method can adjust the operation of the power system and continuously monitor the operation status of the power system. This continuous monitoring and optimization helps to timely discover and deal with potential problems and ensure the long-term stable operation of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is a flow chart of an embodiment of a method for predicting and self-healing control of a power system fault according to the present invention.

[0074] Figure 2 It is a module schematic diagram of an embodiment of a power system fault prediction and self-healing control system of the present invention. DETAILED DESCRIPTION

[0075] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present invention can be combined with each other without conflict.

[0076] The following detailed description is an exemplary description, which is intended to provide further detailed description of the present invention. Unless otherwise specified, all technical terms used in the present invention have the same meaning as those generally understood by those skilled in the art to which the present invention belongs. The terms used in the present invention are only for describing specific embodiments, and are not intended to limit the exemplary embodiments according to the present invention.

[0077] Example 1

[0078] like Figure 1 As shown, a method for predicting and self-healing control of power system faults includes the following steps:

[0079] 1. Data collection and preprocessing:

[0080] Smart sensors, including voltage sensors, current sensors, power sensors, frequency sensors, and environmental parameter sensors, are deployed at key nodes of the power system’s transmission lines, substations, and distribution networks to monitor the electrical quantities and equipment status of the power system in real time.

[0081] The voltage data, current data, power data, frequency data of the power system, and the operating status and environmental parameter data of the equipment are collected in real time through intelligent sensors. Specifically, the voltage data of the power system is collected from the voltage sensor, the current data is collected from the current sensor, the power data is collected from the power sensor, the frequency data is collected from the frequency sensor, and the operating status and environmental parameter data of the equipment are collected from the environmental parameter sensor.

[0082] The collected data is integrated into an initial data set, which contains the overall picture of the power system at the moment of monitoring, providing a basis for subsequent data analysis and fault prediction. Edge computing devices are used to clean the received raw data to remove noise and outliers.

[0083] Feature extraction is performed on the cleaned data to identify key parameters and patterns, and data compression is performed on the extracted features to obtain the preprocessed data set as the monitoring data set.

[0084] 2. Fault prediction and key monitoring:

[0085] Deploy smart sensor networks to collect real-time data on electrical quantities in the power system, including voltage, current, power, frequency, equipment operating status, and environmental parameters. These data are integrated into monitoring data sets.

[0086] The monitoring datasets are preprocessed, including data cleaning (removing noise and outliers), feature extraction (identifying key parameters and patterns), and data compression to facilitate subsequent analysis.

[0087] Train a machine learning model that can analyze trends and abnormal patterns in monitoring data sets based on historical fault data and grid operation patterns.

[0088] Using supervised learning algorithms such as random forests, support vector machines, or neural networks, the model is trained to identify signs that may lead to failure. Labeled historical failure data is used during the model training process to ensure prediction accuracy.

[0089] The preprocessed monitoring data set is input into the trained machine learning model, which analyzes the data and predicts potential failure points and failure types.

[0090] Model outputs include the probability, location and type of failure, as well as the time window in which the failure may occur.

[0091] Based on the predictions from the machine learning model, a long-term monitoring plan is developed, which details the key monitoring strategies for the areas where potential failure points are located, including increasing the monitoring frequency, adjusting sensor parameters, or deploying additional monitoring resources.

[0092] The long-term monitoring program also includes specific monitoring plans for predicted fault types. For example, for predicted overheating problems, it may be necessary to increase the monitoring frequency of temperature sensors.

[0093] Implement a long-term monitoring program to focus on areas with potential failure points. This includes real-time tracking and recording of monitoring data, as well as regular evaluation of monitoring results to ensure that signs of failure are detected in a timely manner.

[0094] Monitoring data is analyzed in real time through edge computing devices to quickly respond to any abnormal changes.

[0095] Regularly conduct in-depth analysis of data from key monitoring areas to verify the predictive accuracy of machine learning models, and adjust the predictive models and monitoring plans based on the latest monitoring results.

[0096] The analysis results are fed back to the intelligent decision-making engine to optimize fault recovery strategies and control instructions.

[0097] 3. Fault diagnosis and report generation:

[0098] Once an abnormality or fault sign is found during key monitoring, artificial intelligence algorithms are used to diagnose the fault, determine the fault location, and generate a fault diagnosis report, which details the nature and location of the fault and the recommended initial response measures.

[0099] Specifically, based on the monitoring data set, the key features are extracted using adaptive threshold and pattern recognition technology. In this step, the monitoring data set is first preprocessed, including normalization and outlier removal, to enhance the accuracy of feature extraction. Then, the local three-value pattern adaptive threshold technology is used to divide the image into blocks according to the local three-value pattern (LTP), and the ε-LTP feature in each block area is calculated.

[0100] The obtained key feature set is input into a pre-trained deep learning model. The model may adopt a graph convolutional neural network (GCN) structure, which can extract spatial features from an undirected graph and obtain time series features through a temporal convolutional neural network. These features are fused to form spatiotemporal features for building a fault prediction model.

[0101] The deep learning model outputs fault diagnosis results, including fault type and possible causes. In this step, the model can identify abnormal patterns in the power system and predict potential fault points and fault types by learning spatiotemporal features.

[0102] According to the fault diagnosis results, the graph theory algorithm is combined with the grid topology to locate the fault location. In this step, the shortest path algorithm and minimum cut algorithm in graph theory are used, combined with the real-time topology and power flow distribution of the grid, to intelligently identify the fault area and isolation plan that need to be isolated.

[0103] The fault diagnosis results and the located fault location are integrated as a fault diagnosis report to guide the adjustment of operating parameters of power grid equipment and the isolation of fault areas. The report describes in detail the nature, location, possible causes and recommended response measures of the fault, providing a basis for the subsequent generation and execution of control instructions.

[0104] Through the above implementation, the method can realize accurate fault diagnosis and location of the power system, generate detailed fault diagnosis reports, and provide scientific and accurate decision support for the self-healing control of the power system. This method improves the automation level of the power system, reduces the cost and risk of fault handling, and enhances the reliability and stability of the power grid.

[0105] 4. Control instruction generation and execution:

[0106] According to the fault diagnosis report, the operating parameters of the power grid equipment are dynamically adjusted, the fault area is isolated, and control instructions are generated. These control instructions include adjusting the operating status of the power grid equipment, such as automatic switching of circuit breakers and automatic distribution of loads.

[0107] Execute control instructions, automatically restore power supply to non-faulty areas, and record recovery operation logs. These logs include operation time, operation type, equipment involved, and operation results.

[0108] Specifically, it receives a fault diagnosis report that includes the fault type, location and possible cause. Based on this report, it assesses the impact of the fault on grid operation, including potential chain reactions. This assessment process involves analyzing the grid area that the fault may affect, predicting the scope of power supply interruption that may be caused by the fault, and assessing the impact on grid stability.

[0109] The evaluation results are summarized into a fault evaluation report, which provides a basis for subsequent adjustments to the operating parameters of power grid equipment.

[0110] Combined with fault diagnosis reports, fault assessment reports, dynamic characteristics of the power grid and historical fault data, the adaptive algorithm is applied to calculate the optimal operating parameters of the power grid equipment in real time. The adaptive algorithm can dynamically adjust the operating parameters according to the real-time status and historical data of the power grid to maintain the stability and efficiency of the power grid.

[0111] The algorithm takes into account the load changes of the power grid, the health status of the equipment and environmental factors, and calculates the optimal operating parameters of the power grid equipment under the current fault conditions.

[0112] Based on the fault location information in the fault diagnosis report, combined with the real-time topology and power flow distribution of the power grid, the shortest path algorithm and minimum cut algorithm in graph theory are used to intelligently identify the fault area that needs to be isolated and the isolation plan. This step involves determining the boundary between the fault area and the non-fault area, and determining the best path to isolate the fault area.

[0113] The isolation plan includes the specific equipment that needs to be isolated and the sequence of operations to ensure that the faulty area can be isolated quickly and effectively while minimizing the impact on non-faulty areas.

[0114] According to the intelligently identified fault area and isolation scheme, the protection devices in the power grid, including relay protection and automatic reclosing devices, are automatically configured. This step involves adjusting the set values ​​of the protection devices to adapt to the new power grid operating status and isolation scheme.

[0115] The configuration of the protection device ensures that it can act quickly when a fault occurs, isolate the fault area, and protect the power grid equipment from damage.

[0116] Combined with the optimized grid equipment operating parameters, fault areas, isolation schemes and protection device configuration schemes, multi-objective optimization algorithms, such as genetic algorithms or particle swarm optimization algorithms, are used to generate control instructions for automatic control of grid equipment. These algorithms can find the best balance between multiple objectives, such as power restoration speed, system stability and resource consumption.

[0117] The control instructions include adjusting the operating status of grid equipment, such as closing and opening of switchgear, and redistribution of loads, to achieve rapid isolation of fault areas and restoration of power supply to non-fault areas.

[0118] Through the above specific implementation methods, the method can achieve rapid response and self-healing control of the power system when a fault occurs, reduce the impact of the fault on the operation of the power grid, and improve the reliability and stability of the power grid. This method provides strong technical support for the safe operation of the power system through accurate fault diagnosis, intelligent fault area identification, automatic protection device configuration and optimized control instruction generation.

[0119] 5. Intelligent decision-making and system adjustment:

[0120] Build an intelligent decision-making engine to analyze the power grid status and recovery operation logs in real time, automatically generate fault recovery strategies, and update control instructions. The intelligent decision-making engine combines machine learning and operations research methods to analyze the power grid status in real time, automatically generate fault recovery strategies, and update control instructions to optimize the self-healing response of the power system.

[0121] Adjust the operation of the power system according to the updated control instructions, and continuously monitor the operating status of the power system to ensure the reliability and stability of power supply.

[0122] Specifically, it first involves collecting restoration operation logs, which detail the time, type, equipment involved, and results of each operation. This data provides the intelligent decision engine with detailed information on historical and current grid operations.

[0123] The intelligent decision engine combines the real-time grid status data set and the recovery operation log data set, and uses machine learning algorithms for in-depth data analysis. This step uses big data technology and artificial intelligence algorithms, such as deep learning, to monitor and analyze the operating status of the power grid in real time to identify potential risks and failure modes.

[0124] Through data analysis, the intelligent decision engine generates a power grid status analysis report, which includes the status of the fault area, the recovery of the non-fault area, and potential risk prediction. The report provides scientific decision support for the operation and maintenance of the power system.

[0125] Based on the grid status analysis report, the intelligent decision engine automatically generates fault recovery strategies using operations research methods. These strategies include recommended grid equipment operating parameter adjustments and power supply path optimization solutions to achieve rapid power supply recovery and stable system operation.

[0126] The intelligent decision engine updates the control instructions based on the generated fault recovery strategy, which guides the further adjustment of the power grid equipment. The updated control instructions are sent to the corresponding power grid equipment to achieve rapid isolation of the fault area and power supply restoration in the non-fault area.

[0127] Finally, the intelligent decision engine monitors the execution of the updated control instructions to ensure that the grid equipment correctly adjusts the operating parameters according to the instructions and that the fault recovery strategy is effectively implemented. During the monitoring process, any deviations or anomalies will be recorded and used to further optimize the performance of the decision engine.

[0128] Through the above implementation, the method can achieve rapid response and self-healing control of the power system when a fault occurs, reduce the impact of the fault on the operation of the power grid, and improve the reliability and stability of the power grid. This method provides strong technical support for the safe operation of the power system through accurate fault diagnosis, intelligent fault area identification, automatic protection device configuration and optimized control instruction generation.

[0129] The method also includes the following steps: after executing the updated control instructions, the power system is monitored again using a smart sensor network. These sensors are deployed at various key nodes of the power system, including transmission lines, substations, and distribution networks, to collect data such as voltage, current, power, frequency, and environmental parameters.

[0130] The data collected through the smart sensor network are integrated into a validation dataset that reflects the real-time status of the power system after the control instructions are executed.

[0131] Compare the validation data set with the expected normal operation parameters of the power system. This step involves comparing the real-time monitoring data with the historical normal operation data of the power system and the preset normal operation thresholds to verify that the power supply has been restored as expected.

[0132] The results of the comparative analysis are collated into a verification analysis report that details the effectiveness of the power system restoration operations, including which areas have successfully restored power supply, which areas have not, and whether there are new anomalies or potential risks.

[0133] If the verification and analysis report shows that the power system has not been fully restored or there are new anomalies, the intelligent decision engine will automatically adjust the fault recovery strategy based on the report content. This step involves re-evaluating the power grid status, identifying unresolved issues, and generating new fault recovery strategies based on the machine learning and operations research algorithms of the intelligent decision engine.

[0134] The new fault recovery strategy aims to address the issues pointed out in the verification analysis report, optimize the power supply path, and adjust the operating parameters of power grid equipment to achieve more effective fault isolation and power restoration.

[0135] Based on the automatically adjusted fault recovery strategy, the intelligent decision engine generates new control instructions. These instructions include specific operation steps, such as reconfiguring grid equipment and adjusting power supply paths.

[0136] New control instructions are sent to the corresponding power grid equipment and executed to achieve further recovery of the power system.

[0137] The operating status of the power system is continuously monitored, and the intelligent decision engine continuously evaluates the progress of the power grid restoration based on real-time data. If necessary, the fault recovery operation and verification steps are repeated until the power system is fully restored to normal operation.

[0138] This continuous monitoring and iterative process ensures that the power system can recover quickly and effectively after a fault occurs, while improving the system's self-healing capabilities and overall reliability.

[0139] Through the above implementation, the method provides a closed-loop power system fault prediction and self-healing control solution, which forms a complete operation process from fault detection, fault diagnosis, control instruction generation, to power supply restoration and system verification. This method ensures that the power system can respond quickly when a fault occurs, automatically restore power supply, and continuously monitor the system status until the system is fully stable.

[0140] Example 2

[0141] It also includes simulation verification and optimization steps:

[0142] Based on the monitoring data set and fault diagnosis report, the hybrid simulation mode and dynamic simulation technology are used to simulate various fault conditions in the power system and verify the implementation effect of the control instructions.

[0143] Create fault simulations that simulate the types and conditions of faults that can occur in the power system, using the initial data set as simulation input.

[0144] Apply fault simulation to the power system simulation environment, execute control instructions, and record system responses and self-healing operations during the simulation.

[0145] Analyze the simulation results and verify the response speed and recovery efficiency of control instructions to different fault types to optimize the control strategy.

[0146] Specifically, it involves developing a hybrid simulation platform that combines digital and physical simulation techniques to simulate various fault conditions in the power system. The digital simulation part uses computer models to simulate the dynamic behavior of the power system, while the physical simulation part may include scaled-down power system hardware to simulate actual physical responses.

[0147] Based on the monitoring data set and fault diagnosis report, a series of possible fault types and conditions are defined. These fault types include but are not limited to short circuit, overload, equipment aging, etc.

[0148] Initial datasets containing grid parameters under normal operation and historical fault conditions are used as simulation inputs to generate realistic fault scenarios.

[0149] The created fault simulation is applied to the power system simulation environment. In the simulation environment, the fault occurrence process is simulated and control instructions are executed according to the preset fault type and conditions.

[0150] Record system responses during simulation, including changes in voltage, current, power, and frequency, as well as the execution of self-healing operations such as circuit breaker operation and load redistribution.

[0151] The simulation results are analyzed in detail to verify the response speed and recovery efficiency of the control instructions to different fault types. The analysis includes fault detection time, fault isolation time, power restoration time, and the state of the system after recovery.

[0152] The effectiveness of control instructions can be evaluated by comparing simulation results with expected goals, which may include minimizing power outage time, maximizing power restoration speed, and ensuring safe and stable system operation.

[0153] Based on the simulation results, verify whether the control instructions can quickly and effectively respond to different types of faults and restore power supply. If the simulation results show that the control instructions fail to achieve the expected effect, adjust and optimize the control instructions.

[0154] Optimization may involve tuning the parameters of control algorithms, improving fault detection logic, or increasing the robustness of self-healing strategies.

[0155] Simulation verification is regarded as a continuous improvement process. Each iteration updates the fault simulation based on new monitoring data and fault diagnosis reports to ensure that the simulation environment can reflect the latest power grid status and potential risks.

[0156] Through continuous simulation verification and optimization, the practical application effect of power system fault prediction and self-healing control methods can be improved, and the self-healing ability and overall reliability of the power grid can be enhanced.

[0157] Through the above implementation, the method can verify the effectiveness of the control instructions through simulation before they are actually deployed, ensuring that faults can be responded to quickly and accurately in the real power system, minimizing the impact of faults on power grid operation. This method improves the automation level of the power system and reduces the cost and risk of fault handling.

[0158] Example 3

[0159] like Figure 2 As shown, a power system fault prediction and self-healing control system specifically includes the following components and their operations:

[0160] Deployment and data collection of smart sensor networks:

[0161] Deploy a smart sensor network consisting of multiple types of sensors, including voltage sensors, current sensors, power sensors, frequency sensors, and environmental parameter sensors. These sensors are installed at key nodes of the power system's transmission lines, substations, and distribution networks to monitor the electrical quantities and equipment status of the power system in real time.

[0162] The data collected by the sensor network includes voltage, current, power, frequency, as well as the operating status and environmental parameters of the equipment. These data are integrated into a monitoring data set to provide basic information for subsequent fault prediction and self-healing control.

[0163] Application of Machine Learning Models:

[0164] Implement machine learning models that can analyze monitoring data sets, identify abnormal patterns and trends, and predict potential failure points and types of failures.

[0165] Based on the prediction results of the machine learning model, a long-term monitoring plan is set up to focus on monitoring areas where potential fault points are located so that signs of faults can be detected in a timely manner.

[0166] Fault diagnosis of artificial intelligence algorithm module:

[0167] Once the smart sensor network detects abnormalities or signs of faults during key monitoring, the artificial intelligence algorithm module immediately intervenes to perform fault diagnosis.

[0168] The module uses advanced algorithms such as deep learning and pattern recognition to determine the fault location and generate a detailed fault diagnosis report containing the fault type, location and possible cause.

[0169] Control the operation of the instruction generation module:

[0170] The control instruction generation module dynamically adjusts the operating parameters of the power grid equipment, such as the output voltage of the transformer and the status of the circuit breaker, according to the fault diagnosis report.

[0171] The module is also responsible for isolating the fault area and generating control instructions to guide the automatic control of power grid equipment, achieving rapid isolation of the fault area and power supply restoration in the non-fault area.

[0172] Functions of the automatic recovery module:

[0173] The automatic recovery module executes control instructions, automatically adjusts the operating status of power grid equipment, and restores power supply to non-fault areas.

[0174] During the execution process, the module records a detailed log of each operation, including the operation time, operation type, equipment involved and operation results, providing data support for subsequent fault analysis and system maintenance.

[0175] Real-time analysis and strategy generation of intelligent decision-making engine:

[0176] The intelligent decision-making engine analyzes the grid status and restoration operation logs in real time, uses machine learning algorithms to perform data analysis, and generates a grid status analysis report.

[0177] Based on the analysis results, the intelligent decision-making engine uses operations research methods to automatically generate fault recovery strategies, including recommended adjustments to grid equipment operating parameters and power supply path optimization plans, and updates control instructions.

[0178] System adjustment and status monitoring of monitoring modules:

[0179] The monitoring module adjusts the operation of the power system according to the updated control instructions and continuously monitors the operating status of the power system.

[0180] This module ensures that the power system is adjusted as directed by the intelligent decision engine and repeats fault recovery operations and verification steps when necessary until the power system is fully restored to normal operation.

[0181] Through the above implementation methods, the system can realize real-time monitoring, fault prediction, self-healing control and continuous monitoring of the power system, improve the reliability and stability of the power system, reduce the power outage time caused by faults, and enhance the self-healing ability of the power grid. This method provides a comprehensive power system fault prediction and self-healing control solution, from data collection, fault prediction, fault diagnosis, control command generation, to power supply restoration and system monitoring, forming a complete closed-loop control process, significantly improving the self-healing ability and operating efficiency of the power system.

[0182] A power system fault prediction and self-healing control system further includes the following operations:

[0183] Monitoring module power supply verification operation:

[0184] After executing the updated control instructions, the monitoring module starts the power supply verification process. The module uses a smart sensor network to collect power supply data in faulty and non-faulty areas, including key parameters such as voltage, current, power and frequency.

[0185] By comparing real-time monitoring data with the normal operating parameters of the power system, the monitoring module verifies whether the power supply has been restored as expected. This step ensures that the fault area has been effectively isolated and the power supply in the non-fault area has been restored to normal levels.

[0186] Verification analysis report generation:

[0187] Based on the data collected by the monitoring module, a verification analysis report is generated. The report records in detail the status of power supply restoration, including the power supply quality in the restored areas, the power supply status in the unrestored areas, and any new abnormalities or potential risks.

[0188] The data in the report is analyzed by an intelligent decision engine to determine the overall restoration status of the power system and any issues that require further action.

[0189] Adjustment of the fault recovery strategy of the intelligent decision engine:

[0190] If the verification and analysis report shows that the power system has not been fully restored or there are new anomalies, the intelligent decision-making engine will automatically intervene and adjust the fault recovery strategy based on the report content.

[0191] The intelligent decision engine uses machine learning algorithms and operations research methods to analyze and verify the data in the analysis report, identify unresolved issues, and generate new fault recovery strategies. These strategies are designed to solve the problems pointed out in the report, optimize the power supply path, and adjust the operating parameters of the power grid equipment.

[0192] Generation and execution of new control instructions:

[0193] Based on the new fault recovery strategy generated by the intelligent decision-making engine, the system automatically generates new control instructions. These instructions include further adjustments to the power grid equipment, such as reconfiguring the power supply path and adjusting the equipment operating parameters.

[0194] The new control instructions are sent to the corresponding power grid equipment and executed to achieve further recovery of the power system. During the execution process, the monitoring module continuously tracks the execution of the instructions to ensure that the new control instructions are correctly implemented.

[0195] Continuous monitoring and iterative cycles:

[0196] The system continuously monitors the operating status of the power system, and the intelligent decision-making engine continuously evaluates the progress of the power grid restoration based on real-time data. If necessary, the fault recovery operation and verification steps are repeated until the power system is fully restored to normal operation.

[0197] This continuous monitoring and iterative process ensures that the power system can respond quickly after a fault occurs, automatically restore power supply, and continue to monitor the system status until the system is fully stable.

[0198] Through the above implementation, the system provides a closed-loop power system fault prediction and self-healing control solution, which forms a complete operation process from fault detection, fault diagnosis, control command generation, to power supply restoration and system verification. This method ensures that the power system can respond quickly when a fault occurs, automatically restore power supply, and continuously monitor the system status until the system is fully stable.

[0199] It is known from common technical knowledge that the present invention can be implemented by other embodiments that do not deviate from its spirit or essential features. Therefore, the above disclosed embodiments are only illustrative in all respects and are not exclusive. All changes within the scope of the present invention or within the scope equivalent to the present invention are included in the present invention.

Claims

1. A method for predicting and self-healing faults in a power system, characterized in that: include, Use smart sensor networks to monitor the electrical quantities of the power system in real time and obtain monitoring data sets; Based on the monitoring data set, use machine learning models to predict potential fault points and fault types; set up long-term monitoring plans and focus on monitoring areas where potential fault points are located; once abnormalities or signs of faults are found in key monitoring, use artificial intelligence algorithms to diagnose faults, determine the fault location, and generate fault diagnosis reports; According to the fault diagnosis report, the operating parameters of the power grid equipment are dynamically adjusted, the fault area is isolated, and control instructions are generated; the control instructions are executed to automatically restore the power supply to the non-fault area and record the recovery operation log; an intelligent decision-making engine is built to analyze the power grid status and the recovery operation log in real time, automatically generate a fault recovery strategy, and update the control instructions; according to the updated control instructions, the operation of the power system is adjusted and the operating status of the power system is monitored.

2. A method for predicting and self-healing faults in a power system according to claim 1, characterized in that: The following steps are also included: Based on the monitoring data set and fault diagnosis report, the hybrid simulation mode and dynamic simulation technology are used to simulate various fault conditions in the power system and verify the implementation effect of the control instructions, including: Create fault simulations that simulate the types and conditions of faults that may occur in the power system, using the initial data set as simulation input; Apply fault simulation to the power system simulation environment, execute control instructions, and record system responses and self-healing operations during the simulation; Analyze the simulation results and verify the response speed and recovery efficiency of control instructions to different fault types.

3. A method for predicting and self-healing faults in a power system according to claim 1, characterized in that: The steps of using smart sensor networks to monitor the electrical quantities of the power system in real time and obtain monitoring data sets include: Deploy smart sensors at key nodes of power transmission lines, substations and distribution networks in the power system, including voltage sensors, current sensors, power sensors, frequency sensors and environmental parameter sensors, to monitor the electrical quantities and equipment status of the power system in real time; The following data are collected in real time through smart sensors: collecting voltage data of the power system from voltage sensors; collecting current data of the power system from current sensors; Collecting power data of the power system from power sensors; collecting frequency data of the power system from frequency sensors; Collect equipment operation status and environmental parameter data from environmental parameter sensors; Integrate the collected data into an initial dataset; Perform preliminary processing on the initial data set. The specific steps include: Use edge computing devices to clean the received raw data to remove noise and outliers; Perform feature extraction on the cleaned data to identify key parameters and patterns; The extracted features are compressed to obtain a preprocessed data set as a monitoring data set.

4. A method for predicting and self-healing power system faults according to claim 1, characterized in that: The steps of using artificial intelligence algorithms to diagnose faults, determine fault locations, and generate fault diagnosis reports include: Based on the monitoring data set, key features are extracted through adaptive threshold and pattern recognition technology to obtain key feature sets; Input the obtained key feature set into the pre-trained deep learning model; The deep learning model outputs fault diagnosis results, including fault type and possible causes; According to the fault diagnosis results, the graph theory algorithm is combined with the power grid topology structure to locate the fault location; The fault diagnosis results and the located fault location are integrated as a fault diagnosis report to guide the adjustment of operating parameters of power grid equipment and the isolation of fault areas.

5. A method for predicting and self-healing power system faults according to claim 1, characterized in that: According to the fault diagnosis report, the operation parameters of the power grid equipment are dynamically adjusted, the fault area is isolated, and the control instructions are generated, specifically including: According to the fault diagnosis report, evaluate the impact of the fault on the grid operation, including the fault type, location and possible chain reaction, and obtain a fault assessment report; Based on fault diagnosis reports, fault assessment reports, dynamic characteristics of the power grid and historical fault data, an adaptive algorithm is applied to calculate the optimal operating parameters of power grid equipment in real time; Based on the fault location information in the fault diagnosis report, combined with the real-time topology and power flow distribution of the power grid, the shortest path algorithm and minimum cut algorithm in graph theory are used to intelligently identify the fault area that needs to be isolated and the isolation plan. The isolation plan includes the specific equipment that needs to be isolated and the operation sequence; Automatically configure protection devices in the power grid, including relay protection and automatic reclosing devices, based on intelligently identified fault areas and isolation schemes; Combined with the optimized grid equipment operating parameters, fault areas, isolation schemes and protection device configuration schemes, a multi-objective optimization algorithm including a genetic algorithm or a particle swarm optimization algorithm is used to generate control instructions for automatic control of grid equipment to achieve rapid isolation of fault areas and power supply restoration in non-fault areas.

6. A method for predicting and self-healing faults in a power system according to claim 1, characterized in that: Execute control instructions, automatically restore power supply to non-faulty areas, and record the steps of the restoration operation log, including: Receive and execute generated control instructions, adjust grid equipment to new operating parameters, and isolate identified fault areas; According to the results of the isolation operation, the power supply status of the non-fault area is determined, and the power supply path and equipment parameters of the non-fault area are automatically adjusted to bypass the fault area and restore the power supply to the non-fault area; During execution, a detailed log is recorded for each operation, including the time of operation, type of operation, equipment involved, and results of the operation; Using smart sensor networks to monitor the status of the power system after restoration operations and verify whether power supply has been successfully restored to non-faulty areas; If the monitoring results show that the power supply has not been fully restored, the power supply to the non-fault area will be automatically restored. If the power supply is restored, the recovery operation log will be recorded.

7. A method for predicting and self-healing faults in a power system according to claim 1, characterized in that: The steps of building an intelligent decision-making engine, analyzing the power grid status and the recovery operation log in real time, automatically generating a fault recovery strategy, and updating the control instructions; adjusting the operation of the power system according to the updated control instructions, and monitoring the operation status of the power system, specifically include: Collect recovery operation logs, including operation time, operation type, devices involved, and operation results; Combining the real-time grid status data set and the recovery operation log data set, the intelligent decision engine uses machine learning algorithms to analyze the data and obtain a grid status analysis report, including the status of the fault area, the recovery status of the non-fault area, and potential risk prediction; Based on the grid status analysis report, the intelligent decision engine uses operations research methods to automatically generate fault recovery strategies, including recommended grid equipment operating parameter adjustments and power supply path optimization solutions; Based on the generated fault recovery strategy, the intelligent decision engine updates the control instructions to guide further adjustments of the power grid equipment; The updated control instructions are sent to the corresponding power grid equipment and their execution is monitored.

8. A method for predicting and self-healing power system faults according to claim 1, characterized in that: The following verification steps are also included: After executing the updated control instructions, the power system is monitored again using the smart sensor network to obtain a verification data set to verify whether the power supply in the fault area and the non-fault area has been restored as expected; Compare and analyze the obtained verification data set with the expected normal operation parameters of the power system to obtain a verification analysis report; If the verification and analysis report shows that the power system has not been fully restored or there are new anomalies, the intelligent decision engine will automatically adjust the fault recovery strategy based on the report and generate and execute new control instructions; Continuously monitor the operating status of the power system and repeat fault recovery operations and verification steps as needed until the power system is fully restored to normal operating conditions.

9. A power system fault prediction and self-healing control system, based on a power system fault prediction and self-healing control method according to any one of claims 1 to 8, characterized in that: include, Intelligent sensor networks are used to monitor the electrical quantities of the power system in real time, including voltage, current, power, frequency, and the operating status of equipment and environmental parameters, and obtain monitoring data sets; Machine learning models are used to predict potential fault points and fault types based on monitoring data sets, and to set up long-term monitoring plans to focus on monitoring areas where potential fault points are located; Artificial intelligence algorithm module, used to diagnose faults, determine fault locations, and generate fault diagnosis reports once abnormalities or fault signs are found during key monitoring; A control instruction generation module is used to dynamically adjust the operating parameters of the power grid equipment according to the fault diagnosis report, isolate the fault area, and generate control instructions; An automatic recovery module is used to execute control instructions, automatically restore power supply to non-faulty areas, and record recovery operation logs; Intelligent decision-making engine for real-time analysis of grid status and recovery operation logs, automatic generation of fault recovery strategies, and updating of control instructions; The monitoring module is used to adjust the operation of the power system according to the updated control instructions and monitor the operating status of the power system.

10. A power system fault prediction and self-healing control system according to claim 9, characterized in that: The monitoring module is further used to verify whether the power supply to the fault area and the non-fault area has been restored as expected after executing the updated control instructions, and to generate a verification analysis report; The intelligent decision-making engine is also used to automatically adjust the fault recovery strategy according to the verification and analysis report, and generate and execute new control instructions if the verification and analysis report shows that the power system has not been fully restored or there are new abnormalities.

Images