A power grid technical improvement optimization method fusing failure spatiotemporal pattern and resilience enhancement

Abstract

The application discloses a power grid technical improvement optimization method fusing a fault space-time mode and a resilience enhancement, and relates to the technical field of power grid improvement.The method comprises the following steps: collecting power grid historical operation data, and constructing a space-time comprehensive risk function; constructing a resilience index system and dividing the resilience level of system area operation according to the state transition probability of equipment operation state in the historical operation data; fusing multi-dimensional risk information based on the space-time comprehensive risk function and the resilience level, obtaining a high-risk technical improvement line and a list of to-be-transformed demand, and generating a technical improvement scheme set; evaluating each technical improvement scheme, constructing a multi-objective optimization model, and solving an optimal technical improvement scheme set.The application obtains the optimal technical improvement scheme set through the integrated closed-loop process of "recognition-evaluation-planning-optimization", provides a scientific basis for decision-making, and significantly improves the scientificity, implementability and engineering benefit of the technical improvement combined scheme.

Description

Technical Field

[0001] This invention relates to the field of power grid upgrading technology, and more specifically, to a power grid upgrading optimization method that integrates fault spatiotemporal patterns and resilience enhancement. Background Technology

[0002] In recent years, with the large-scale integration of new energy sources, frequent extreme weather events, and rapid growth in urban electricity load, the safety of power grid operation has faced increasingly severe challenges. Traditional power transmission and transformation systems, during the planning and design phase, primarily focus on stability and economic efficiency, relying heavily on human experience and static models, making it difficult to dynamically adapt to environmental changes and fault evolution. Especially in disaster-prone areas, sudden faults can easily spread along spatial chains, triggering multi-level cascading blackouts and widespread load interruptions, severely impacting the safe and stable operation of the power system.

[0003] Currently, the following shortcomings are commonly found in the implementation of power grid technological upgrades:

[0004] (1) Fault risk identification relies heavily on experience-based judgment and lacks precise modeling methods driven by multi-source data, making it difficult to accurately reflect the spatial distribution characteristics of risks and their temporal evolution trend.

[0005] (2) Existing resilience assessments mostly focus on the level of a single device and have not yet established a system-level resilience classification standard and quantitative assessment model, which makes it difficult to support the decision-making of overall resilience improvement;

[0006] (3) The route planning process often overlooks the actual geographical construction difficulty and the coupling characteristics of multiple risk factors, resulting in unreasonable configuration of protective measures and low resource utilization efficiency;

[0007] (4) In the strategy selection stage, a multi-objective decision-making model that comprehensively considers transformation costs, operational risks and system flexibility has not yet been built, making it difficult to provide scientific and differentiated planning basis for different engineering scenarios.

[0008] In summary, to meet the urgent needs of modern power grid development, it is imperative to construct a power grid technological upgrading planning methodology with dynamic risk perception and system resilience enhancement capabilities. Summary of the Invention

[0009] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a power grid technical upgrade optimization method that integrates fault spatiotemporal patterns and resilience enhancement, addressing the problem of how to comprehensively consider fault risks and economic costs to obtain a comprehensive power grid technical upgrade scheme that meets the needs of the power grid under the premise of stable operation of the power grid system.

[0010] To achieve the above objectives, embodiments of the present invention provide a power grid technical upgrade optimization method that integrates fault spatiotemporal modes and resilience enhancement, comprising:

[0011] Collect historical power grid operation data to construct a spatiotemporal integrated risk function;

[0012] Based on the state transition probability of equipment operating status in the historical operating data, a resilience index system is constructed and the resilience level of system regional operation is divided.

[0013] Based on the spatiotemporal integrated risk function and the resilience level, multi-dimensional risk information is integrated to obtain a list of high-risk technical transformation routes and transformation needs, and a set of technical transformation schemes is generated.

[0014] Each of the proposed technical improvement schemes is evaluated, and a multi-objective optimization model is constructed to solve for the optimal set of technical improvement schemes.

[0015] In a preferred embodiment, the spatiotemporal integrated risk function integrates the fault spatial density distribution and fault seasonal fluctuation trend of the power grid's historical operation.

[0016] In a preferred embodiment, historical power grid operation data is collected, and a spatiotemporal integrated risk function is constructed, including:

[0017] Collect historical operational data to build a multi-dimensional basic sample library;

[0018] Based on the geographical coordinates and sample size of historical fault points in the basic sample library, a spatial fault density function with continuous spatial distribution of faults is constructed using the kernel density estimation method.

[0019] Based on the fault occurrence time in the basic sample library, and taking into account the seasonal fault trend of power grid operation, Fourier series is introduced to perform periodic modeling of historical fault time series to obtain Fourier seasonal fluctuation function.

[0020] Based on the spatial fault density function and the Fourier seasonal fluctuation function, a spatiotemporal integrated risk function is constructed.

[0021] In a preferred embodiment, after constructing the spatial fault density function, the method further includes:

[0022] Based on the aforementioned spatial fault density function, the spatial density distribution of high-risk points is clustered and identified to obtain spatial high-risk blocks.

[0023] Extract the center coordinates and fault density index of the high-risk blocks in the space to construct a preliminary fault clustering layer;

[0024] After obtaining the Fourier seasonal fluctuation function, the method further includes:

[0025] Based on the Fourier seasonal fluctuation function, the seasonal risk fluctuation curve of the power grid is obtained and the annual high-risk period is identified.

[0026] In a preferred embodiment, the state transition probability is obtained specifically by:

[0027] The operating states of power grid equipment are classified, and based on the frequency of occurrence and transition of the operating states in historical operating data and the influence of weather, a state transition probability matrix of the operating states of power grid equipment is constructed to obtain the state transition probabilities.

[0028] In a preferred embodiment, the step of constructing a resilience index system and classifying the resilience levels of system regional operations includes:

[0029] Based on the output characteristics of wind power and photovoltaics, power models for wind power and photovoltaics are constructed respectively;

[0030] Based on the state transition probability matrix and the power model, a resilience index system for evaluating the regional operational resilience of the power grid system is constructed using Bayesian networks and stochastic power flow simulation.

[0031] Based on the aforementioned resilience index system, the resilience levels of the system's regional operations are classified.

[0032] The toughness levels include high vulnerability, medium toughness, and high toughness.

[0033] In a preferred embodiment, the power model includes a wind power model constructed using a Weibull distribution and a photovoltaic power model constructed using a normal distribution;

[0034] The resilience index system includes four core indicators: renewable energy power deficit rate, recovery time window, system volatility index, and average path length of fault chains.

[0035] In a preferred embodiment, the step of integrating multi-dimensional risk information to obtain a list of high-risk technical upgrade routes and upgrade requirements includes:

[0036] By integrating multi-dimensional risk information such as historical fault density, geological risk and construction difficulty, an improved A* path search algorithm is introduced. Taking into account the distance and risk from the fault source node to the target point, the path with the lowest comprehensive risk is obtained as the high-risk technical transformation route.

[0037] Extract high-risk key node equipment from each of the high-risk technical upgrade routes, establish the operation information of the key node equipment, and construct a structured path-equipment-state triplet;

[0038] Based on the aforementioned triples, identify the objects to be upgraded and obtain a list of upgrade requirements;

[0039] The operational information of key node equipment includes the equipment's years of operation, health index, and current operating status.

[0040] In a preferred embodiment, generating a set of technical improvement schemes and evaluating each of the technical improvement schemes includes:

[0041] Based on the list of requirements to be upgraded, multiple optional technical upgrade methods are configured for each key node device, and a constraint satisfaction model is constructed to generate a preliminary set of technical upgrade solutions.

[0042] Based on the initial set of technical improvement schemes, an evaluation vector is constructed for each technical improvement scheme to obtain a comprehensive score for each scheme and to preliminarily screen and obtain a set of high-quality technical improvement schemes.

[0043] The evaluation vectors include economic indicators, risk mitigation indicators, construction feasibility indicators, and system compatibility indicators.

[0044] In a preferred embodiment, the step of constructing a multi-objective optimization model and solving for the optimal set of technical improvement schemes includes:

[0045] A multi-objective optimization function is constructed with the objectives of minimizing total cost and maximizing risk reduction.

[0046] With the goal of improving the operational flexibility of the power grid system, optimization constraints are established;

[0047] Based on the set of high-quality technical improvement schemes, the improved NSGA-II non-dominated sorting genetic algorithm is used to solve for the Pareto optimal solution set as the output of the optimal technical improvement scheme set.

[0048] The beneficial effects of this invention are:

[0049] (1) By constructing a spatiotemporal integrated risk function based on multi-source historical operation data, a dynamic risk map based on historical operation data is formed, which improves the accuracy and foresight of power grid fault identification and provides accurate data support for subsequent power grid technical transformation;

[0050] (2) Based on the equipment operating status and the transition probability between each operating status in the historical operating data, a resilience index system for evaluating the regional operating resilience of the power grid system is constructed, and the regional operation of the power grid system is classified into resilience levels to identify high-vulnerability areas of the system operation area, providing a precise basis for subsequent path optimization and strategy formulation.

[0051] (3) By establishing an integrated closed-loop process of “identification-evaluation-planning-optimization”, the optimal set of technical transformation schemes is obtained by solving and optimizing, which significantly improves the scientificity, feasibility and engineering benefits of the technical transformation combination schemes, provides a scientific basis for decision-making and helps to obtain the optimal technical transformation combination schemes that achieve the greatest benefits. Attached Figure Description

[0052] Figure 1A flowchart illustrating a power grid upgrade optimization method that integrates fault spatiotemporal patterns and resilience enhancement.

[0053] Figure 2 A flowchart illustrating the process of constructing a spatiotemporal integrated risk function;

[0054] Figure 3 Example of a risk map for a spatiotemporal integrated risk function;

[0055] Figure 4 A flowchart illustrating the process of constructing a resilience index system and classifying the resilience levels of system regions.

[0056] Figure 5 A flowchart illustrating the process of generating a set of technical improvement solutions;

[0057] Figure 6 A flowchart illustrating the process of finding the optimal set of technical improvement solutions. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0059] Example

[0060] Please see Figure 1 A method for optimizing power grid technical upgrades that integrates spatiotemporal fault patterns and resilience enhancement includes: collecting historical power grid operation data and constructing a spatiotemporal comprehensive risk function; constructing a resilience index system and classifying the resilience levels of system regions based on the state transition probabilities of equipment operating states in the historical operation data; integrating multi-dimensional risk information based on the spatiotemporal comprehensive risk function and the resilience levels to obtain a list of high-risk technical upgrade lines and upgrade requirements, and generating a set of technical upgrade schemes; evaluating each technical upgrade scheme and constructing a multi-objective optimization model to solve for the optimal set of technical upgrade schemes.

[0061] By collecting historical power grid operation data, a spatiotemporal comprehensive risk function is constructed based on multi-source historical operation data, forming a dynamic risk map based on historical operation data. This improves the accuracy and foresight of power grid fault identification, providing precise data support for subsequent power grid technical upgrades. Furthermore, based on the equipment operating status and transition probabilities between different operating states in the historical operation data, a resilience index system for assessing the regional operational resilience of the power grid system is constructed. This resilience index system is then used to classify the regional operation of the power grid system into resilience levels, identifying high-vulnerability, medium-resilience, and high-resilience areas, providing a precise basis for subsequent path optimization and strategy formulation. Based on the spatiotemporal comprehensive risk function and the resilience levels classified by the power grid system operation, multi-dimensional risk information is integrated to obtain high-risk technical upgrade lines and identify a list of upgrade needs, generating a preliminary set of technical upgrade schemes. Each technical upgrade scheme is evaluated to obtain a preliminary set of high-quality technical upgrade schemes. Finally, a multi-objective optimization model is constructed to solve for the optimal set of technical upgrade schemes, providing a scientific basis for decision-making and helping to select the optimal combination of technical upgrade schemes that achieves maximum benefit under limited resource conditions. In other words, by establishing an integrated closed-loop process of "identification-evaluation-planning-optimization", the scientific nature, feasibility and engineering benefits of the technical transformation combination scheme have been significantly improved.

[0062] Please see Figure 2 To improve the accuracy of basic data for power grid technological upgrading and optimization, this embodiment further limits the collection of historical power grid operation data and the construction of the spatiotemporal comprehensive risk function. Specifically:

[0063] S1. Collect historical power grid operation data and construct a spatiotemporal integrated risk function; wherein, the spatiotemporal integrated risk function integrates the fault spatial density distribution and fault seasonal fluctuation trend of historical power grid operation;

[0064] S11. Collect historical operational data and construct a multi-dimensional basic sample library:

[0065] To comprehensively reflect the causes and distribution characteristics of power grid faults, historical operation data of the power grid in recent years is collected and standardized to construct a multi-dimensional basic sample library. This historical operation data includes historical fault data, which contains multi-dimensional features such as time, space, fault type, and environmental factors. For example, historical operation data from the past 5-10 years is collected, standardized into a standard modeling format, and a structured basic sample library is formed.

[0066] The historical sample set built based on historical operational data, i.e., the multi-dimensional basic sample library, is as follows:

[0067]

[0068] In the formula, Indicates the time of the fault occurrence, in hours; , Represents geographic coordinates; Indicate the fault type; , , These represent wind speed, temperature, and humidity, respectively. Indicates the service life of the equipment; Indicates the number of samples.

[0069] This basic sample library includes failure time, geographical coordinates, failure type, wind speed, temperature, humidity, equipment service life, and sample quantity. It not only contains single-point failure information, but also characterizes the combined failure risk under the interaction of equipment aging and extreme weather, which is the basis for subsequent model calculations.

[0070] S12. Based on the aforementioned basic sample library, construct the spatial fault density function:

[0071] To identify areas of concentrated fault distribution in the power grid, a smooth model of the spatial distribution of fault points is constructed using the kernel density estimation method. Specifically, based on the geographical coordinates and sample size of historical fault points in the basic sample library, the kernel density estimation method is used to construct a continuous spatial distribution function of faults, i.e., a spatial fault density function.

[0072] The spatial fault density function is as follows:

[0073]

[0074] In the formula, Indicates kernel bandwidth, controlling the smoothness level; Representing coordinates The risk density value; Indicates the first in history The spatial location of each fault point; Indicates the number of samples.

[0075] By constructing a spatial fault density function using the kernel density estimation method, the problem of difficulty in analyzing point-like discrete data can be solved, and a continuous risk surface can be formed on the map for visualization. At the same time, this spatial fault density function can well reflect the geographical clusters of faults in the power grid.

[0076] S13. Based on the spatial fault density function, the improved DBSCAN clustering algorithm is used to cluster and identify the spatial density distribution of high-risk points to obtain spatial high-risk blocks.

[0077] The center coordinates and fault density index of the high-risk blocks in the space are extracted to construct a preliminary fault clustering layer.

[0078] To classify high-risk points into spatial high-risk blocks, an improved DBSCAN clustering algorithm is used to cluster the spatial density distribution results based on the spatial fault density function; that is, density clustering identifies high-risk areas. The DBSCAN clustering algorithm is a density-based spatial clustering algorithm that defines a cluster as the largest set of density-connected points. It identifies clusters by using the concept of density, dividing areas with sufficiently high density into clusters, and can discover clusters of arbitrary shapes in noisy spatial databases.

[0079] The DBSCAN clustering algorithm mainly relies on neighborhood radius. and minimum points Two parameters, based on the radius of the data point in its neighborhood. The number of points contained within a node is used to determine whether it is a core point, a boundary point, or a noise point, thereby achieving clustering.

[0080] Each cluster center and density index is defined, and the coordinates of the cluster center are:

[0081] ,

[0082] The cluster density index is:

[0083]

[0084] In the formula, , Indicates the center coordinates of the clustered region; This represents the number of fault points contained in the c-th cluster; Indicates the area of ​​the clustered region; This represents the density index; the higher the value, the more concentrated the faults are in that area.

[0085] The DBSCAN clustering algorithm is used to classify and divide high-risk points in the power grid, and the spatial distribution results of high-risk points are clustered to identify spatial high-risk blocks. By extracting the center coordinates and fault density index of the clustering region of each high-risk block, a preliminary fault clustering layer is formed.

[0086] S14. Based on the fault occurrence time in the basic sample library, and taking into account the seasonal fault trend of power grid operation, Fourier series is introduced to perform periodic modeling on the historical fault time series to obtain the Fourier seasonal fluctuation function.

[0087] Based on the Fourier seasonal fluctuation function, the seasonal risk fluctuation curve of the power grid is obtained and the annual high-risk period is identified.

[0088] Based on the time series of historical fault points in the aforementioned basic sample library, and considering the obvious seasonal fault trends in power system operation, such as lightning strikes in summer and icing in winter, Fourier series are introduced to model the historical fault time series periodically, constructing an annual Fourier seasonal fluctuation function. This annual seasonal fluctuation function is an infinite series that represents a periodic function as a combination of sine and cosine functions.

[0089] The Fourier seasonal fluctuation function is defined as follows:

[0090]

[0091] In the formula, Indicates time Seasonal risk value; A time variable measured in days; "Tian" indicates a cycle of one year. This indicates the harmonic order, which is typically taken as 2-3. , Indicates the first The regression coefficients of the first harmonic; The DC component represents the baseline or average level of seasonal risk throughout the year.

[0092] By introducing Fourier series to perform periodic modeling of historical failure time series, i.e. seasonal risk trend modeling, Fourier seasonal function is obtained to output seasonal risk curve; through the seasonal risk curve, seasonal risk fluctuation patterns can be identified, and then annual high-risk periods can be identified, such as typhoon season and icing period, typical high-risk periods can be extracted.

[0093] S15. Based on the spatial fault density function and the Fourier seasonal fluctuation function, construct a spatiotemporal integrated risk function: wherein the spatiotemporal integrated risk function integrates the fault spatial density distribution and fault seasonal fluctuation trend of the power grid's historical operation.

[0094] Please see Figure 3 To achieve a unified assessment of spatial fault density and seasonal fault fluctuations, a spatiotemporal integrated risk function is constructed based on the spatial fault density function and the Fourier seasonal fluctuation function. as follows:

[0095]

[0096] In the formula, , These are weighting coefficients used to balance the contributions of spatial risk and seasonal risk; Indicates time and location The overall failure risk; Representing coordinates The risk density value, i.e., the spatial fault density function; This represents the cluster density, used in the weighted spatial fault density function to characterize whether the location is in a high-risk area; Indicates time The seasonal risk value is the Fourier seasonal fluctuation function. It is the cluster density index obtained based on the DBSCAN clustering algorithm in S13, when If it is located inside a spatially high-risk cluster block c identified by the improved DBSCAN clustering algorithm in S13, then ;when When it does not belong to any identified high-risk cluster block, The higher the cluster density, the greater the risk value of the region.

[0097] By integrating the spatial fault density function and the Fourier seasonal fluctuation function—that is, based on the spatial fault density function and its weighting coefficients, and the Fourier seasonal fluctuation function and its weighting coefficients—a spatiotemporal integrated risk function is constructed to calculate the spatiotemporal fault integrated risk weight. This achieves a unified assessment of spatial fault density and seasonal fault fluctuations, providing accurate indicator inputs for subsequent path optimization and resource allocation. The spatiotemporal integrated risk function is also known as the spatiotemporal fault integrated risk function.

[0098] By accessing multi-source historical operational data and integrating elements such as time, space, equipment operating status, and environmental meteorology, a multi-dimensional feature system is constructed, and key influencing factors are identified. At the same time, by combining density clustering and seasonal analysis techniques, a dynamic risk map is formed, which significantly improves the accuracy and foresight of fault identification and provides data support for technical transformation projects.

[0099] Please see Figure 4 To enhance the scientific rigor and precision of power grid technological upgrades, this application further elaborates on constructing a resilience index system for assessing the regional operational resilience of power grid systems and classifying the resilience levels of regional system operations. Specifically:

[0100] S2. Based on the state transition probability of the power grid equipment operating status in the historical operating data, and based on the characteristics of new energy output, construct a resilience index system to evaluate the regional operating resilience of the power grid system and classify the regional operating resilience level of the system.

[0101] S21. Classify the operating states of power grid equipment, and based on the frequency of occurrence and transition of the operating states in historical operating data and the impact of weather, construct a state transition probability matrix for the operating states of power grid equipment to obtain the state transition probabilities:

[0102] The operating status of power grid equipment can be divided into the following five states: This indicates that the system is functioning normally. Indicates a minor abnormality; Indicates a protection trip; Indicates a complete failure; This indicates that the work was done by someone to isolate or repair the area.

[0103] Based on the operating status of power grid equipment, and according to the number of transitions between the various operating states in historical power grid operating data, the total number of occurrences of each operating state, the weighting coefficient of weather influence, and the intensity of weather influence, the state transition probability matrix of power grid equipment during operation from the first state to the second state is constructed as follows:

[0104]

[0105] Wherein, each element in the state transition probability matrix for:

[0106]

[0107] In the formula, Indicates the state of power grid equipment Transition to state The probability of; This indicates the status in the historical operation data of the power grid. Transition to state The number of times; This indicates the status in the historical operation data of the power grid. Total number of occurrences; These are weighting coefficients used to measure the degree of influence of weather on this type of state transition; Denotes the correction factor, where, Indicates the first The normalized value of the intensity of extreme weather events, such as the intensity of typhoons and ice storms.

[0108] The first state and the second state of the power grid equipment are both one of the five states of the power grid equipment operation, and the first state is different from the second state.

[0109] By dividing the operating states of power grid equipment and considering the frequency of transitions between these states and the occurrence of each state in historical operating data, along with the impact of weather disturbances, a state transition probability matrix for power grid equipment operating states is constructed, forming the basic mechanism for power grid equipment fault evolution. This state transition probability matrix yields the state transition probabilities of power grid equipment, which can then be used to calculate load losses during system faults and adjust power output based on wind and solar power output. .

[0110] S22. Based on the output characteristics of wind power and photovoltaic power, construct power models for wind power and photovoltaic power respectively; wherein, the power models include a wind power power model constructed using Weibull distribution and a photovoltaic power model constructed using normal distribution;

[0111] S221, Wind Power Modeling:

[0112] wind speed The probability density function:

[0113]

[0114] Corresponding wind power curve:

[0115]

[0116] In the formula, , The shape and scale parameters of the Weibull distribution; , , These represent the cut-in, rated, and cut-out wind speeds, respectively. This indicates the rated power of the wind turbine.

[0117] S222, Photovoltaic power modeling:

[0118] Light intensity The distribution is as follows:

[0119]

[0120] Photovoltaic power output is:

[0121]

[0122] In the formula, , This represents the mean and standard deviation of light intensity. Indicates photovoltaic conversion efficiency; Indicates standard illumination intensity; This indicates the rated power of the photovoltaic unit.

[0123] By comprehensively considering the statistical characteristics of wind power and photovoltaic power output, the uncertainty of new energy power is modeled using Weibull distribution and normal distribution respectively, and multi-scenario input sequences are generated through power mapping, thereby providing dynamic disturbance sources for the operation of the power grid system.

[0124] S23. Based on the state transition probability matrix and the power model, a resilience index system for evaluating the regional operational resilience of the power grid system is constructed using Bayesian networks and stochastic power flow simulation. The resilience index system includes four core indicators: renewable energy power deficit rate, recovery time window, system fluctuation index, and average path length of fault chains.

[0125] Under random disturbances, the S21 grid equipment status and the S22 renewable energy output power are used as input conditions. Through system evolution and stochastic power flow simulation, the operation fluctuations of the grid system are simulated and tracked.

[0126] S231. Based on the aforementioned state transition probability model, model the propagation of cascading failures:

[0127] Based on Bayesian networks, the conditional probability model for cascading failure propagation is constructed as follows:

[0128]

[0129] In the formula, This indicates an indicator function; it returns 1 if the condition is met, and 0 otherwise. Indicates the first The system frequency at any given time; This indicates the frequency tolerance threshold.

[0130] The conditional probability model is an expansion of the state transition probability matrix. The probability of a fault occurring is calculated using the state transition probability matrix, and the power output fluctuations of new energy sources such as wind power and photovoltaics are incorporated into the calculation of load loss.

[0131] S232. Construct a resilience index system for assessing the regional operational resilience of power grid systems.

[0132] S2321, Core Indicator 1: New Energy Power Shortage Rate :

[0133] The threshold calculation is adjusted based on the impact of wind and solar power on the grid load. During a fault, the system state transition probability affects the load loss calculation. When the system transitions from normal state N to fault state F, insufficient wind and solar power will lead to higher load losses.

[0134] By combining the output fluctuations of wind power and photovoltaic power with the system state transition probability, the renewable energy power deficit rate is obtained. for:

[0135]

[0136] In the formula, and These are wind power and solar power at each moment. Power output; This refers to the rated power of wind and solar power. It is the total load of the power grid; This refers to the rated power of the photovoltaic system.

[0137] S2322, Core Indicator 2: Calculate the recovery time window based on system recovery time and disturbance occurrence time. for:

[0138]

[0139] In the formula, Indicates system recovery time; Indicates the time when the disturbance occurred.

[0140] S2323, Core Indicator 3: Calculate the system fluctuation index, i.e., the amount of unstable transitions, based on the number of state transitions in the operating state of power grid equipment. for:

[0141]

[0142] In the formula, Indicates at time step The change in state; This represents the total number of observation periods.

[0143] S2324, Core Indicator Four: Based on the... Calculate the average path length of each fault chain based on its length and the total number of fault chains. for:

[0144]

[0145] In the formula, Indicates the first Length of the fault chain; This indicates the total number of fault chains.

[0146] By employing Bayesian networks and stochastic power flow simulation methods to model fault propagation paths and grid cascading responses, the load loss, recovery time, state fluctuations, and fault chain structure of the system under different disturbance scenarios are recorded in real time. Four core indicators of the system under different disturbance scenarios are calculated: renewable energy power deficit rate, recovery time window, system fluctuation index, and average fault chain path length. This provides a scientific basis for the subsequent classification of the resilience level of the system's operating areas.

[0147] S24. Based on the resilience index system for assessing the regional operational resilience of the power grid system, classify the regional operational resilience levels of the system; wherein, the resilience levels include high vulnerability level, medium resilience level and high resilience level.

[0148] Please refer to Table 1 below. Based on the four core indicators obtained from simulation—new energy power deficit rate, recovery time window, system fluctuation index, and average path length of the fault chain—the system operating area is divided into three resilience levels as follows:

[0149] Table 1: Statistical Table of Resilience Levels for System Operating Area Division

[0150]

[0151] Based on four core indicators obtained from simulations—new energy power deficit rate, recovery time window, system volatility index, and average fault chain path length—a resilience index system is constructed to assess the resilience of the system's operating areas. This system quantitatively evaluates and classifies the operational resilience of each area, dividing the system's operating areas into high-vulnerability, medium-resilience, and high-resilience zones. The high-vulnerability zone is identified, providing a precise basis for subsequent path optimization and strategy formulation. Specifically, the high-vulnerability zone is a high-risk area for priority technological upgrades, the medium-resilience zone is a key monitoring area, and the high-resilience zone is a stable operating area.

[0152] Unlike traditional approaches that focus only on individual power grid equipment or local indicators, this embodiment combines the characteristics of the power grid structure and its operational response capabilities, and considers the output characteristics of new energy sources. It constructs a multi-dimensional resilience index system that includes four core indicators: new energy power deficit rate, recovery time window, system fluctuation index, and average path length of fault chains. This system supports the classification of resilience levels for areas affected by faults by region, greatly improving the scientific and precise nature of power grid technical upgrades.

[0153] Please see Figure 5 To construct a multi-level set of technical improvement solutions, this embodiment further elaborates on the generation and preliminary screening of the technical improvement solution set, specifically:

[0154] S3. Based on the spatiotemporal integrated risk function and the resilience level, multi-dimensional risk information is integrated, and an improved A* path search algorithm is introduced to obtain high-risk technical modification routes.

[0155] Extract high-risk key node equipment from the high-risk technical upgrade route, obtain a list of upgrade requirements, and generate a set of technical upgrade solutions by combining the operating information of the key node equipment.

[0156] Each of the aforementioned technical improvement schemes was evaluated, and a set of high-quality technical improvement schemes was initially selected.

[0157] In short, based on the spatiotemporal integrated risk function and the classified resilience level, and integrating multi-dimensional information such as historical fault density, geological risk, and construction difficulty, an improved A* path search algorithm is introduced to search for the optimal technical upgrade path through high-risk nodes in the power grid. Combining the operational information of key node equipment along the path, status diagnosis is performed on key nodes and lines in each path to form a technical upgrade requirement list. Based on this technical upgrade requirement list, multiple selectable technical upgrade methods are matched using rule-driven or expert experience bases, and combined according to strategic logic to form a set of technical upgrade solutions.

[0158] S31. By integrating multi-dimensional risk information such as historical fault density, geological risk, and construction difficulty, an improved A* path search algorithm is introduced. This algorithm comprehensively considers the distance and risk from the fault source node to the target point to obtain the path with the lowest overall risk as the high-risk technical upgrade route. That is:

[0159] By incorporating multi-dimensional information such as historical fault density, geological risk, and construction difficulty, the traditional A* heuristic function is improved to form a path optimization mechanism for power grid scenarios. In the power grid topology, starting from the fault source node and targeting key load centers or important substations, a path with the minimum weight is constructed as the priority line for technical upgrades. Specifically, based on the geographical coordinates of historical fault points, the area of ​​each cluster region, and the number of fault points, the coordinates of the cluster centers and the cluster density index are calculated to identify fault risk areas.

[0160] S311, Define Path Overall Weight

[0161] By integrating multi-dimensional risk information such as historical fault density, geological risk, and construction difficulty, an improved A* path search algorithm is introduced to establish a comprehensive path weight for each route. for:

[0162]

[0163] In the formula, The historical fault density representing the edge position is obtained from the spatial fault density function estimated by the S1 kernel density; Geological risk level; Due to construction difficulty; , , These are the weighting coefficients, satisfying... .

[0164] S312. Based on the comprehensive weight of each edge, the Euclidean distance from the node to the target point, and the balance factor, calculate the priority evaluation value in the A* search. That is, the improved A* path search algorithm is guided by both the distance to the target point and the accumulated risk of the searched paths, searching for the optimal solution path that combines distance and risk, and defining the node guided by the combined distance and risk. The heuristic function is:

[0165]

[0166] In the formula, Let e ​​be the actual minimum cost from the starting node to the current node n, where e is the path length. One edge on; Let n be the estimated minimum cost from the current node n to the target node, where n is the minimum cost. For nodes Euclidean distance to the target point; As a balancing factor, it controls the weights of distance and risk; This is the priority evaluation value in A* search, which is the estimated total cost of the path starting from the current node n.

[0167] S313, Intelligent Path Search and Avoidance

[0168] To find the optimal path from the fault source to the target node, an obstacle avoidance mechanism was implemented during the search process. This mechanism automatically avoids black hole nodes with extremely high risk values ​​and very poor regional operational resilience, ultimately obtaining the path with the lowest overall risk as the high-risk technical upgrade route. Specifically:

[0169] Initialization: Initialize the starting node and the target node, that is: set the starting point and the target point.

[0170] Expanding the path: Starting from the initial node, traverse adjacent nodes to compute the priority evaluation value in the A* search. That is: traverse adjacent nodes and calculate .

[0171] Obstacle avoidance mechanism: If the current adjacent node satisfies the following conditions: the comprehensive risk weight of spatiotemporal faults is greater than the threshold of the comprehensive risk weight of spatiotemporal faults, and the resilience level is 1, then the current adjacent node is skipped. satisfy If the toughness level is 1, then skip this step. In the formula, It is the spatiotemporal fault comprehensive risk weight threshold, used to delineate high-risk and low-risk areas in risk assessment; a resilience level of 1 refers to highly vulnerable areas in the system. When a fault occurs in these areas, the power grid has poor recovery capabilities and is prone to causing large-scale faults. Therefore, in the calculation of risk propagation paths, areas with a resilience level of 1 may have more severe risk propagation and should be prioritized for improvement and reinforcement.

[0172] Completion Judgment: Once the target node is accessed, the high-risk propagation backbone path is output as the lowest comprehensive risk path, i.e., the high-risk technical upgrade line. .

[0173] S32. Based on the obtained high-risk technical transformation lines, extract the high-risk key node equipment in each high-risk technical transformation line and establish the operation information of the key node equipment;

[0174] Construct a structured "path-device-status" triplet to identify objects to be upgraded and obtain a list of upgrade requirements; among them, the operational information of key node equipment includes the equipment's service life, health index and current operating status.

[0175] Based on the obtained high-risk technical upgrade routes, high-risk key node equipment in each high-risk technical upgrade route is extracted. Combining information such as equipment operating years, health index and current operating status, the "path-equipment-status" triple is constructed to identify the objects to be upgraded, and the abstract high-risk technical upgrade routes are transformed into a specific list of upgrade requirements.

[0176] Define the path-device-state triplet, i.e., the identification unit. for:

[0177]

[0178] In the formula, For the first The high-risk technical upgrade route is determined by the improved A* path search algorithm in step S31; For path Upper Key node devices; For path Upper The operational information of key node devices, which can be represented as:

[0179]

[0180] In the formula, For equipment Operating lifespan; As a health index; This indicates the current status of the device.

[0181] S33. Based on the list of requirements to be modified, configure multiple optional technical modification methods for each key node device, and construct a constraint satisfaction model to generate a preliminary set of technical modification schemes; wherein, the constraint satisfaction model includes mandatory constraints, economic constraints and reliability constraints;

[0182] S331. Based on each identified high-risk technical upgrade route, extract all key node equipment on the high-risk technical upgrade route. For each device node Provides a set of multiple optional technical modification methods :

[0183]

[0184] In the formula, For device nodes The nth optional technical modification method.

[0185] S332. Construct the constraint satisfaction model as follows:

[0186] (1) Decision variables: Represents device node Among the selected technological upgrade methods, For device nodes The chosen technological improvement method is the decision variable.

[0187] (2) The constraints include:

[0188] S3321, Mandatory Constraints:

[0189] Equipment that is not in good working order must undergo technical upgrades:

[0190]

[0191] S3322, Economic Constraints:

[0192] The total cost of technological upgrading is lower than the threshold for technological upgrading costs, meaning the total cost of technological upgrading is limited by:

[0193]

[0194] in, Indicates the first The cost of a technical upgrade plan typically includes expenses related to the upgrade, such as equipment replacement and maintenance costs. This is the total budget, meaning that the total cost of all technological upgrades must not exceed this budget.

[0195] S3323, Reliability Constraints:

[0196] Adjacent devices must not be shut down simultaneously to prevent system-wide interruptions.

[0197]

[0198]

[0199] In the formula, A represents a set of equipment pairs that are electrically or functionally coupled and are not allowed to shut down simultaneously; Each technical improvement method Required downtime.

[0200] S333. Output the set of all technical improvement schemes that satisfy the above constraints:

[0201]

[0202] In the formula, This represents the set of all eligible technological upgrading schemes, where each scheme... It includes a set of technological transformation decisions; m is the total number of equipment.

[0203] S34. Based on the preliminary set of technical improvement schemes, an evaluation vector is constructed for each technical improvement scheme to obtain a comprehensive score for each scheme and to preliminarily select a set of high-quality technical improvement schemes. The evaluation vector includes economic indicators, risk mitigation indicators, construction feasibility indicators, and system compatibility indicators. That is, an evaluation vector for each technical improvement scheme is constructed based on the economic indicators, risk mitigation indicators, construction feasibility indicators, and system compatibility indicators. A comprehensive score function is constructed based on the evaluation vector to calculate the comprehensive score for each technical improvement scheme. The top K high-quality technical improvement schemes are selected based on the obtained comprehensive scores.

[0204] For each scheme Construct the evaluation vector:

[0205]

[0206] In the formula, For economic indicators, namely ; This is a risk mitigation indicator, representing the cumulative reduction in risk along the path. For construction feasibility indicators; This is a system compatibility indicator.

[0207] The comprehensive scoring function is:

[0208]

[0209] In the formula, It is a comprehensive score for each technical transformation plan, which takes into account four factors: economic efficiency, risk, wind energy reduction rate, and construction period. These are the weighting coefficients for each indicator, reflecting the importance of each factor in the final evaluation.

[0210] Select the top scorers based on their overall scores. The best technical improvement solutions are selected and combined to form a set of best technical improvement solutions, which are used as input for S4 multi-objective optimization and then sorted or screened for implementation in S4.

[0211] By introducing an improved A* path search method and applying it to power grid technical upgrade path planning, the method comprehensively considers risk density, construction environment and cost weight, identifies key risk protection paths, and links them with risk level assessment results to construct a multi-level set of technical upgrade solutions, thus realizing a closed-loop design from planned path to specific strategy.

[0212] Please see Figure 6 To obtain a feasible technological upgrading solution that balances economic considerations with risk reduction, this example further elaborates on how to solve for the optimal set of technological upgrading solutions, based on the existing set of high-quality solutions. Specifically:

[0213] S4. Based on the aforementioned set of high-quality technical improvement schemes, a multi-objective optimization model is constructed, and an improved NSGA-II non-dominated sorting genetic algorithm is used to solve for the Pareto optimal solution set, which is then output as the optimal technical improvement scheme set.

[0214] S41. Construct a multi-objective optimization model

[0215] The construction of the multi-objective optimization model includes:

[0216] A multi-objective optimization function is constructed with the objectives of minimizing total cost and maximizing risk reduction.

[0217] With the goal of improving the operational flexibility of the power grid system, optimization constraints are established.

[0218] S411. Construct a multi-objective optimization function with the objectives of minimizing total cost and maximizing risk reduction:

[0219] It has One candidate high-quality technological upgrading scheme, decision variables The objective function is defined as follows:

[0220]

[0221] The minimized part is:

[0222]

[0223] In the formula, This represents the total investment cost; It is a decision variable; indicating whether to choose the first option. One option; This represents the sum of risk downgrades; similarly, It is a decision variable; it represents the first... Risk downgrade value for each option.

[0224] The minimization part means that when selecting a technological upgrading plan, efforts should be made to keep investment costs and risks at a low level, with the goal of reducing investment and lowering risks.

[0225] The maximized part is:

[0226]

[0227] In the formula, Representing a path Maximum residual risk, relative to the maximum risk of the system Perform normalization; This represents the weighted sum of the system's fault tolerance metrics; It is the fault tolerance indicator for each technical improvement plan.

[0228] The goal of maximizing the remaining risk and system fault tolerance is to maximize the system's resilience and adaptability in the face of failures.

[0229] also, and The weighting is determined by experts based on experience. If investment cost is very important, a larger weight can be assigned to it. If fault tolerance is important for the system, then it can be given Larger weight.

[0230] Each technological improvement plan is modeled as an optional decision variable, and a set of alternative plans is formed by combining different strategy terms.

[0231] S412. To improve the operational flexibility of the power grid system, establish optimization constraints:

[0232] S4121. Budget Constraints:

[0233]

[0234] By establishing budget constraints, we ensure that total costs do not exceed the total budget.

[0235] S4122, Critical Node Coverage Constraints:

[0236]

[0237] In the formula, For the total budget, For the set of key nodes, This represents the minimum number of nodes required for coverage.

[0238] By establishing key node coverage constraints, the technical upgrade plan must cover at least These key nodes are designed to ensure enhanced resilience in the core operating area.

[0239] By constructing optimized constraints that integrate budget constraints and critical node coverage constraints, we can not only ensure the economic feasibility of the selected technical upgrade scheme, but also effectively improve the system's operational security.

[0240] S42. The optimal set of technical improvement schemes is obtained by using the improved NSGA-II non-dominated sorting genetic algorithm:

[0241] An improved NSGA-II non-dominated sorting genetic algorithm is used to output multiple Pareto optimal combinations of technical improvement schemes for decision-makers to choose from.

[0242] S421. Initialize the population

[0243] ,

[0244] Each individual This indicates the selection of a set of technical improvement schemes.

[0245] S422, Elite Preservation Strategy: In each evolutionary step, high-quality solutions are selected to enter the next generation based on the following two indicators:

[0246] S4221, Non-dominated sorting

[0247] individual Dominate ,like:

[0248]

[0249] and

[0250] After sorting, we get:

[0251]

[0252] S4222, Congestion Distance

[0253] Crowding is a method used in genetic algorithms to compare the sparsity of individual solutions. In multi-objective optimization, a higher crowding level indicates that the individual solution is more sparse in the solution space and is more likely to be selected.

[0254] By utilizing crowding density, genetic algorithms can select relatively independent solutions within a multi-objective space, thus avoiding local optima. For each optimization objective function... Calculate crowding degree in the sorting layer:

[0255]

[0256] The overall congestion level is:

[0257]

[0258] Different objectives of the optimization objective function will all be involved in the calculation of congestion. This indicates the target number.

[0259] S423, Iterative Evolution: Through genetic operations such as selection, crossover, and mutation, the biological evolution process is simulated to continuously generate new technological improvement solutions.

[0260] S4231, Selection Mechanism

[0261] Prioritize solutions with higher dominance levels; if dominance levels are the same, choose solutions with higher crowding.

[0262] During selection, the genetic algorithm uses the calculated value of the optimization objective function to determine the fitness. This objective function includes investment cost, risk degradation, residual risk, and fault tolerance. Therefore, the selection criteria are: the lower the investment cost and risk, the higher the fitness of the solution; the greater the residual risk and fault tolerance, the higher the fitness of the solution.

[0263] S4232, Genetic Operations

[0264] S42321. Cross-operation combines the characteristics of the parent solution to generate a child solution that has the advantages of both. That is, it combines the characteristics of the two different technical improvement schemes of the parent, inheriting the excellent characteristics of the parent in terms of investment cost and risk reduction, thereby potentially generating a better trade-off solution.

[0265]

[0266] During the cross-operation process, optimizing the investment cost and risk degradation in the objective function will affect the fitness of the sub-solution.

[0267] S42322. During the mutation operation, the mutated solution will be evaluated based on the calculated value of the optimization objective function. If the mutated solution can reduce investment costs, lower risk, increase residual risk, or improve fault tolerance within the optimization objective function, then it will have higher fitness and increase its chances of entering the next generation.

[0268] By randomly changing individual elements in the solution with a certain probability, such as randomly adding or removing a technical improvement project, new changes are introduced. This can be represented as:

[0269] (probability) )

[0270] Mutation operations make it less likely for individual solutions to get trapped in local optima during the genetic process, and may even lead to the discovery of better solutions.

[0271] S43. The solution after the mutation operation should satisfy the two constraints in S42.

[0272] To ensure that the solution does not violate the constraints, a penalty term is introduced into the objective function. The penalty function penalizes the objective function value based on the degree of constraint violation. The specific formula is as follows: If a constraint is violated, the fitness is adjusted using the penalty function:

[0273]

[0274] In the formula, For large numbers, To constrain the degree of violation.

[0275] By setting a penalty mechanism, the solution is to... If a constraint in S42 is violated, the algorithm will not discard the solution directly, but will instead add a large penalty value to the fitness of the solution, i.e., the objective function value of the solution. This penalty value ensures that solutions with extremely poor fitness are highly likely to be eliminated in subsequent screenings, thus effectively guiding the algorithm's search process to always remain within the region that satisfies the constraints.

[0276] S44. The set of Pareto optimal solutions is:

[0277] ,

[0278] In the formula, This represents a set of optimal solutions, i.e., the Pareto optimal solution set; in multi-objective optimization problems, The optimal solution set represents those solutions that cannot simultaneously optimize all objectives; These solutions achieve an optimal balance among the various objectives and cannot be improved by other solutions. Among multiple technical improvement schemes, the optimal combination of technical improvement schemes that minimizes cost, reduces risk the most, and maximizes flexibility is selected. That is, in this set of optimal solutions, each solution represents the best trade-off between minimizing total cost and maximizing risk reduction, which cannot be simultaneously surpassed. Decision-makers can make a final selection from this set of Pareto optimal solutions based on actual preferences, such as budget constraints and the degree of security pressure.

[0279] By optimizing and selecting technical improvement solutions, a multi-objective optimization model is designed, taking into account core dimensions such as investment cost, risk reduction effect, and improved operational flexibility. The NSGA-II non-dominated sorting genetic algorithm is used for solving the problem, and a penalty mechanism is set in the solution selection process to effectively guide the search process to always be carried out within the constrained region, thereby outputting multiple Pareto optimal solutions. At the same time, the multi-objective optimization model incorporates budget constraints and key node coverage requirements to ensure that the output set of optimal technical improvement solutions is feasible for engineering practice.

[0280] In summary, by constructing a multi-objective optimization function with the objectives of minimizing total cost and maximizing risk reduction, and establishing constraints that the technical upgrade scheme should satisfy, an improved NSGA-II non-dominated sorting genetic algorithm with a penalty mechanism is employed to obtain a Pareto optimal solution set. This provides a scientific basis for decision-making, helping to select the optimal technical upgrade combination scheme that achieves maximum benefit under limited resources. This optimal technical upgrade combination scheme, through comprehensive optimization of multiple factors such as cost, risk, and time, can reduce system failure risk, improve grid resilience, and ensure the project is completed smoothly within budget. Furthermore, based on this optimal technical upgrade combination scheme, a detailed implementation plan is developed, and resources, equipment, and personnel are allocated to ensure smooth project execution and guarantee the feasibility of the technical upgrade combination scheme.

[0281] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0282] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0283] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0284] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0285] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0286] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A power grid technical upgrade optimization method integrating fault spatiotemporal modes and resilience enhancement, characterized in that, include: Collect historical power grid operation data to construct a spatiotemporal integrated risk function; Based on the state transition probability of equipment operating status in the historical operating data, a resilience index system is constructed and the resilience level of system regional operation is divided. Based on the spatiotemporal integrated risk function and the resilience level, multi-dimensional risk information such as historical fault density, geological risk and construction difficulty is integrated, and an improved A* path search algorithm is introduced. The distance and risk from the fault source node to the target point are comprehensively considered to obtain the path with the lowest integrated risk as the high-risk technical transformation route. Extract high-risk key node equipment from each of the high-risk technical upgrade routes, establish the operation information of the key node equipment, and construct a structured path-equipment-state triplet; Based on the aforementioned triplet, the objects to be upgraded are identified, and a list of upgrade requirements is obtained; wherein, the operational information of key node equipment includes the equipment's service life, health index, and current operating status; A set of technical improvement schemes is generated, each of the schemes is evaluated, and a multi-objective optimization model is constructed to solve for the optimal set of technical improvement schemes.

2. The power grid upgrading and optimization method integrating fault spatiotemporal modes and resilience enhancement as described in claim 1, characterized in that, The spatiotemporal integrated risk function integrates the spatial density distribution of faults and the seasonal fluctuation trend of faults in the historical operation of the power grid.

3. The power grid upgrading and optimization method integrating fault spatiotemporal modes and resilience enhancement as described in claim 1 or 2, characterized in that, Collect historical power grid operation data and construct a spatiotemporal comprehensive risk function, including: Collect historical operational data to build a multi-dimensional basic sample library; Based on the geographical coordinates and sample size of historical fault points in the basic sample library, a spatial fault density function with continuous spatial distribution of faults is constructed using the kernel density estimation method. Based on the fault occurrence time in the basic sample library, and taking into account the seasonal fault trend of power grid operation, Fourier series is introduced to perform periodic modeling of historical fault time series to obtain Fourier seasonal fluctuation function. Based on the spatial fault density function and the Fourier seasonal fluctuation function, a spatiotemporal integrated risk function is constructed.

4. The power grid upgrading and optimization method integrating fault spatiotemporal modes and resilience enhancement as described in claim 3, characterized in that, After constructing the spatial fault density function, the method further includes: Based on the aforementioned spatial fault density function, the spatial density distribution of high-risk points is clustered and identified to obtain spatial high-risk blocks. Extract the center coordinates and fault density index of the high-risk blocks in the space to construct a preliminary fault clustering layer; After obtaining the Fourier seasonal fluctuation function, the method further includes: Based on the Fourier seasonal fluctuation function, the seasonal risk fluctuation curve of the power grid is obtained and the annual high-risk period is identified.

5. The power grid upgrading and optimization method integrating fault spatiotemporal modes and resilience enhancement as described in claim 1 or 4, characterized in that, The state transition probability is obtained specifically as follows: The operating states of power grid equipment are classified, and based on the frequency of occurrence and transition of the operating states in historical operating data and the influence of weather, a state transition probability matrix of the operating states of power grid equipment is constructed to obtain the state transition probabilities.

6. The power grid upgrading and optimization method integrating fault spatiotemporal modes and resilience enhancement as described in claim 5, characterized in that, The construction of a resilience index system and the classification of system regional operation resilience levels include: Based on the output characteristics of wind power and photovoltaics, power models for wind power and photovoltaics are constructed respectively; Based on the state transition probability matrix and the power model, a resilience index system for evaluating the regional operational resilience of the power grid system is constructed using Bayesian networks and stochastic power flow simulation. Based on the aforementioned resilience index system, the resilience levels of the system's regional operations are classified. The toughness levels include high vulnerability, medium toughness, and high toughness.

7. The power grid upgrading and optimization method integrating fault spatiotemporal modes and resilience enhancement as described in claim 6, characterized in that, The power models include a wind power model constructed using the Weibull distribution and a photovoltaic power model constructed using the normal distribution; The resilience index system includes four core indicators: renewable energy power deficit rate, recovery time window, system volatility index, and average path length of fault chains.

8. The power grid upgrading and optimization method integrating fault spatiotemporal modes and resilience enhancement as described in claim 1, characterized in that, The process of generating a set of technical improvement solutions and evaluating each of the technical improvement solutions includes: Based on the list of requirements to be upgraded, multiple optional technical upgrade methods are configured for each key node device, and a constraint satisfaction model is constructed to generate a preliminary set of technical upgrade solutions. Based on the initial set of technical improvement schemes, an evaluation vector is constructed for each technical improvement scheme to obtain a comprehensive score for each scheme and to preliminarily screen and obtain a set of high-quality technical improvement schemes. The evaluation vectors include economic indicators, risk mitigation indicators, construction feasibility indicators, and system compatibility indicators.

9. The power grid upgrading and optimization method integrating fault spatiotemporal modes and resilience enhancement as described in claim 8, characterized in that, The construction of a multi-objective optimization model to solve for the optimal set of technical improvement solutions includes: A multi-objective optimization function is constructed with the objectives of minimizing total cost and maximizing risk reduction. With the goal of improving the operational flexibility of the power grid system, optimization constraints are established; Based on the set of high-quality technical improvement schemes, the improved NSGA-II non-dominated sorting genetic algorithm is used to solve for the Pareto optimal solution set as the output of the optimal technical improvement scheme set.

Images