Power system safety assessment method and system considering multiple meteorological sensitive factors

Abstract

The invention discloses an electric power system safety assessment method and system considering multiple meteorological sensitive factors, and relates to the technical field of electric power system safety assessment, and the method comprises the steps: building a fault model under the influence of multiple meteorological factors based on the coupling effect of different meteorological factors on electric power system elements; and calculating the fault probability of each element of the power system under the extreme weather condition by using a multi-modal data fusion and neural network technology. And evaluating the safety of the power system through a Monte Carlo method in combination with the fault probability of the system element, and outputting a reliability index. According to the method, the dependence of traditional safety assessment on single-factor hypothesis, a static model and low-efficiency calculation is broken through, and a full-link innovation path of multi-factor dynamic perception-deep learning fusion modeling-global state probability assessment is realized; finally, the technical effects of improving the evaluation accuracy, adapting to extreme weather evolution and enhancing the emergency response capability of the power system are achieved, and the method has remarkable social value and engineering application prospects.

Description

Technical Field

[0001] The present invention relates to the technical field of power system security assessment, and in particular to a power system security assessment method and system taking multiple meteorological sensitive factors into consideration. Background Art

[0002] At present, the development of the power system shows that the proportion of highly uncertain new energy sources such as wind and light is increasing, and the power load is becoming more complex and diversified. When extreme weather conditions occur, the impact of uncertainty in multiple links of power supply, load demand and equipment failure will be further amplified.

[0003] The interplay of global climate change and its attendant insecurity is one of the most pressing social security issues of our time. As a result, extreme natural disasters such as storms, floods, and typhoons are occurring more frequently than ever before, with greater destructive power and increased uncertainty. More seriously, the trend of increasing severity and worsening natural disasters is expected to persist for the foreseeable future. The normal functioning of modern society is highly dependent on a reliable power supply. In recent years, the complexity of global climate change and the diverse causes of major power outages have led to a recurring pattern of major blackouts worldwide, resulting in severe social impacts and economic losses.

[0004] Power grid companies need to reasonably assess the safety of grid operation under extreme weather conditions to ensure a safe and reliable power supply. Under the influence of the current complex environmental factors, different meteorological factors will have a coupled impact on various components in the power system, which in turn will affect the safety and reliability of the power system. It is necessary to ensure reliable power supply under the coupling of multiple meteorological factors. Summary of the Invention

[0005] In view of the above-mentioned problems, the present invention is proposed.

[0006] Therefore, the technical problems solved by the present invention are: the existing power system safety assessment methods ignore the superimposed influence of multiple meteorological factors, resulting in large deviations in the assessment results, lack of refined modeling of the failure risks of key components under extreme climatic conditions, and the inability of safety index calculations to dynamically respond to meteorological changes. There is also the problem of how to accurately assess the operational safety and reliability of the power system in extreme weather scenarios under the influence of multiple meteorological sensitive factors.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: A method for power system security assessment that takes into account multiple meteorological sensitive factors, including establishing a fault model under the influence of multiple meteorological factors based on the coupling effect of different meteorological factors on power system components. Multimodal data fusion and neural network technology are used to calculate the failure probability of each component of the power system under extreme weather conditions. Combined with the failure probability of system components, the safety of the power system is evaluated by the Monte Carlo method, and a reliability index is output. Evaluating the safety of the power system by the Monte Carlo method includes, based on the failure probability of each component of the power system under extreme weather conditions, simulating the power system by the Monte Carlo sampling method, randomly sampling the status of each component in the power system, and generating the system status distribution by combining the system's flow and power dispatch model. The system security assessment result is obtained by statistical analysis of multiple system states.

[0008] As a preferred embodiment of the power system security assessment method considering multiple meteorologically sensitive factors described in the present invention, establishing a fault model under the influence of multiple meteorological factors includes obtaining historical fault data for key power system components and analyzing their impact in combination with meteorological factors. For these key components, a specific failure probability model is established using AI deduction and multimodal data fusion methods to determine the failure risk of each power system component under different extreme weather conditions.

[0009] As a preferred embodiment of the power system security assessment method considering multiple meteorologically sensitive factors described in the present invention, calculating the failure probability of each power system component under extreme weather conditions includes, after determining the meteorological factors and component failure risk, using a long-short-term memory neural network model to analyze the component failure probability under multiple meteorological factors. The long-short-term memory neural network model processes time series data through input gates, forget gates, and output gates, capturing the long-term dependency between meteorological factors and failure probability, and outputting the predicted failure probability of each power system component under future extreme weather conditions.

[0010] As a preferred embodiment of the power system security assessment method considering multiple meteorologically sensitive factors described in the present invention, the Monte Carlo method for assessing power system security includes performing multiple samplings using the Monte Carlo sampling method based on the calculated predicted failure probability of each power system component under future extreme weather conditions to generate different system states. In each sampling, the fault state of each key component in the power system is randomly extracted and applied to the power system's power flow and power dispatch model. Each sampling result generates a corresponding system state, indicating whether each component in the system is in a fault state at that moment, and the corresponding power flow distribution, load distribution, and voltage stability are calculated. The results of each sampling are used to analyze the overall system operation, including calculating load reduction and assessing whether the system experiences islanding effects and power imbalance. Statistical analysis of the multiple sampling results yields power system security indicators, including load loss and reliability indicators under different extreme weather conditions.

[0011] As a preferred embodiment of the power system security assessment method considering multiple meteorologically sensitive factors described in the present invention, the load loss amount includes, in each Monte Carlo sampling, calculating the load reduction amount using an optimal power flow algorithm based on the current system state, analyzing the system state to determine whether islanding effects and power imbalances occur, and calculating the load loss amount of the power system under extreme meteorological conditions by statistically analyzing the results of multiple samplings.

[0012] As a preferred embodiment of the power system security assessment method considering multiple meteorological sensitive factors described in the present invention, the reliability index includes the sample mean of the system reliability index obtained by combining mathematical statistics with the power system's state and load loss at each sampling point through multiple Monte Carlo sampling cycles. The reliability index includes the power shortage probability, power shortage expectation, and power shortage expectation.

[0013] As a preferred embodiment of the power system security assessment method considering multiple meteorologically sensitive factors described in the present invention, outputting the reliability index includes determining whether the system security index has converged after the calculation is completed. If converged, the final reliability assessment result is output. If not, meteorological state sampling and sampling calculation are repeated until a converged reliability index result is obtained, ensuring the accuracy of the assessment and the safety of the system.

[0014] Another object of the present invention is to provide a power system security assessment system that takes into account multiple meteorological sensitive factors. The system can establish a fault model under the influence of multiple meteorological factors based on the coupling effect of different meteorological factors on power system components, thereby solving the problems of the current existing power system security assessment methods, such as ignoring the superimposed influence of multiple meteorological factors, resulting in large deviations in the assessment results, lack of refined modeling of the failure risks of key components under extreme climatic conditions, and the inability of safety index calculations to dynamically respond to meteorological changes. In addition, the system can solve the problem of how to accurately assess the operational safety and reliability of the power system in extreme weather scenarios under the influence of multiple meteorological sensitive factors.

[0015] As a preferred solution of the power system security assessment system considering multiple meteorological sensitive factors described in the present invention, it includes: a power system component fault model establishment module, a power system security assessment module, and a reliability index output and convergence judgment module.

[0016] The power system component fault model establishment module is used to establish a fault model under the influence of multiple meteorological factors based on the coupling effects of different meteorological factors on power system components.

[0017] The power system security assessment module is used to calculate the failure probability of each component of the power system under extreme weather conditions by using multimodal data fusion and neural network technology.

[0018] The module for outputting reliability index and judging convergence is used to evaluate the safety of the power system by Monte Carlo method based on the failure probability of system components and output reliability index.

[0019] A computer device includes a memory and a processor, wherein the memory stores a computer program and the processor executes the computer program to implement a method for power system security assessment considering multiple meteorological sensitive factors.

[0020] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a power system security assessment method considering multiple meteorological sensitive factors.

[0021] Beneficial effects of the present invention: The power system security assessment method considering multiple meteorological sensitive factors provided by the present invention constructs a coupling relationship model between multiple meteorological factors such as wind speed, lightning strikes, floods and other meteorological factors and key components of the power system. It not only covers typical extreme meteorological events, but also can reflect the impact mechanism of local differences and meteorological evolution processes on power system components, improves the physical reality and adaptability of fault modeling, and lays a solid data and logical foundation for subsequent system reliability assessment based on meteorological field input.

[0022] By utilizing multimodal data fusion and neural network technology to calculate failure probabilities, and by integrating multi-source time-series meteorological data with historical equipment failure data using an LSTM (Long Short-Term Memory) neural network model, the system can output predictive failure probabilities for equipment in future extreme weather scenarios. This significantly improves the timeliness and accuracy of risk assessments, adapting to the decision-making needs of rapidly changing meteorological events and providing a reliable source of risk warning information for power grid dispatchers.

[0023] Combining failure probabilities with Monte Carlo methods, this approach assesses power system security and outputs reliability indicators. By constructing a system state generation mechanism based on a non-sequential Monte Carlo method and conducting multiple rounds of sampling simulation using the failure probabilities of each component, combined with power flow calculations and load reduction analysis models, this approach enables a comprehensive assessment of system-level operating status, islanding effects, power imbalances, and other issues. This complete evaluation loop, from components to systems, from probabilities to indicators, and from data to simulation, enhances the scientific nature, applicability, and reliability of the assessment results. It also provides a solid basis for power companies to formulate operational plans and allocate dispatch resources in response to extreme weather conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 This is an overall flow chart of a power system security assessment method considering multiple meteorological sensitive factors provided by the first embodiment of the present invention.

[0026] Figure 2 This is an overall flow chart of a power system security assessment system considering multiple meteorological sensitive factors provided in the third embodiment of the present invention. DETAILED DESCRIPTION

[0027] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0028] Example 1, reference Figure 1 , which is an embodiment of the present invention, provides a power system security assessment method considering multiple meteorological sensitive factors, including:

[0029] S1: Based on the coupling effect of different meteorological factors on power system components, a fault model under the influence of multiple meteorological factors is established.

[0030] Obtain historical fault data of key components of the power system and analyze the impact of meteorological factors on the components; for key components, use AI deduction and multimodal data fusion methods to establish specific failure probability models to obtain the failure risks of each component of the power system under different extreme weather conditions.

[0031] One preferred approach to building a specific failure probability model involves first modeling the failure probability of the line. The failure of any conductor or tower on the line will cause the entire line to shut down. The line can only operate normally if both the conductors and towers are functioning properly. Therefore, for an overhead transmission line with m1 conductors and m2 towers, the wind-induced outage model is as follows:

[0032]

[0033] in, represents the outage probability of the lth line at time t, and are the wind speed v of the kth line and the kth tower at time t respectively. i,k,t The failure probability under the action of m1 is m1. m2 is m2 transmission towers on the line. Indicates that the k-th tower is subjected to wind speed v at time t i,k,t Probability of failure at impact. Indicates that the k-th tower is subjected to wind speed v at time t i,k,t Failure probability at impact. i,k,t represents the wind speed of the kth tower at time t.

[0034] Combining historical line wind-induced outage data and adopting data-driven and artificial intelligence methods, the outage probability of transmission lines under different wind speeds is obtained.

[0035] For multi-span ice-covered lines, the series model is used to calculate the average failure rate λ of the line in the period [t, t+Δt]:

[0036]

[0037] Where λ represents the average failure rate of the line in the time interval [t, t+Δt], which is used to evaluate the failure risk of the entire line section in this time period. i represents the unit section failure rate of the i-th section of the line in the time interval. n represents the number of span sections in the multi-span line, that is, the line consists of several sections connected in series.

[0038] Assuming that the probability of a line outage within a time interval Δt follows a Poisson distribution, the reliability and failure probability of the line in the period [t, t+Δt] are:

[0039]

[0040] P f (Δt)=1-P R (Δt)=1-exp(-λΔt)

[0041] Among them, P R (Δt) represents the reliability of the line in the time interval [t, t+Δt], that is, the probability that the line maintains normal operation during this time period. Δt represents the evaluation time length, that is, the span of the time interval under investigation. exp(-λΔt) is the probability that no fault occurs per unit time under the Poisson process. P f (Δt) represents the probability of a line failure within the time interval [t, t+Δt]. This formula is based on the failure modeling of the Poisson distribution and represents the cumulative probability of at least one failure event occurring.

[0042] The [0, t] time period is considered to be composed of several time periods of length Δt. The series model is used to calculate the reliability and failure probability of the line in [0, t]:

[0043]

[0044] Among them, P R (t) represents the reliability of the line in the time interval [0, t], that is, the cumulative probability that the line maintains normal operation during the entire evaluation period. Represents the reliability of the line in the i-th time slice (length is Δt). f (t) represents the failure probability of the line in the time interval [0, t], that is, the cumulative probability of at least one failure occurring during the entire period. n' represents the total time period [0, t] divided into n' time intervals, each of length Δt, that is, t = n'·Δt.

[0045] The wind turbine failure model under the influence of multiple meteorological factors is as follows: when the wind speed in the wind farm is higher than the cut-out wind speed, wind power cannot provide reliable power output. Therefore, only the failure risk scenarios within the wind speed range that can be used by wind power are considered. When the wind speed in the wind farm is lower than the rated wind speed of the wind turbine, the operation status of the wind turbine is relatively reliable and the probability of failure is very small. It can be considered that only accidental random failures will occur. The probability is expressed as This can be obtained from historical data on wind turbine failure rates. When the wind speed is between the rated wind speed and the cut-out wind speed, wind turbine failures primarily manifest as malfunctions or failures caused by overheating of components such as the motor windings, gearbox, and inverter under high load at high wind speeds. The failure risk increases with increasing wind speed. The probability of wind turbine failure within this wind speed range can be approximately modeled as:

[0046]

[0047] in, It indicates the probability of accidental failure of the fan under normal operating conditions, which can be obtained from the historical data of the fan operation failure rate. r Indicates the rated wind speed of the fan. cutin Indicates the cut-in wind speed. v cutoff Indicates the cut-out wind speed. v w It indicates the real-time wind speed at the geographical location of the wind farm, which can be calculated based on the distance between the wind farm and the typhoon center. Indicates that the fan is at a wind speed of v w The probability of failure in the event of a failure.

[0048] Lightning discharges can cause short-term overvoltage in electronic components within the wind turbine, leading to insulation failure and short circuits in internal components, which can cause wind turbine failures. Analysis of historical data on wind turbine lightning failures shows that the probability of a wind turbine shutting down due to lightning strikes is approximately linearly related to the frequency of lightning strikes on the wind turbine, as expressed by:

[0049]

[0050] Wherein, k represents a lightning strike correlation parameter greater than zero, which can be obtained by fitting based on actual lightning strike data. In the present invention, k is set to 0.2186. is the number of lightning strikes on the wind turbine. Indicates the probability of failure of the wind turbine due to lightning strike. N L Indicates the number of lightning strikes on the wind turbine, expressed as:

[0051]

[0052] Among them, h s Indicates the height of the fan in meters (m). N g Indicates the ambient lightning ground flash density under meteorological observation.

[0053] Therefore, the failure probability of a wind turbine affected by lightning in severe weather is expressed as:

[0054]

[0055] Energy storage equipment may also fail during flooding. At the beginning of an extreme rainstorm, the failure rate of energy storage equipment is comparable to that in normal weather. However, water accumulation continues. When the water level in the area where the energy storage equipment is located exceeds the flood control height, the failure rate of the energy storage equipment increases rapidly. Therefore, based on the grid division, an exponential function fitting method is used to establish a relationship between the failure rate of energy storage equipment and the dynamically changing water level. The failure rate of the energy storage equipment at time t is:

[0056]

[0057] Among them, d w (t) represents the degree of water accumulation in the grid where device w is located at time t. w represents the designed flood-proof height of the energy storage device. ζ represents the attenuation coefficient. γ represents the damping coefficient. It represents the failure rate of the energy storage device at position w in the grid at time t, that is, the probability density of failure per unit time. Bw Indicates the flood prevention tolerance boundary of the equipment, which is used to control the lower limit of the judgment range to prevent misjudgment.

[0058] When Δt is small enough, the failure probability of the energy storage device within [t, t+Δt] is:

[0059]

[0060] in, represents the failure probability of the energy storage device in the time period [t, t+Δt].

[0061] Several time periods of length Δt can be combined into the [0, T] period. The failure probability within [0, T] can be calculated using the series model:

[0062]

[0063] Where T is the evaluation duration. represents the failure probability of the energy storage device in the time period [0, t].

[0064] S2: Utilize multimodal data fusion and neural network technology to calculate the failure probability of each component of the power system under extreme weather conditions.

[0065] After determining the meteorological factors and component failure risks, a long-short-term memory neural network model is used to analyze the failure probability of components under multiple meteorological factors. The long-short-term memory neural network model processes time series data through input gates, forget gates, and output gates to capture the long-term dependency between meteorological factors and failure probability, and outputs the predicted failure probability of each component in the power system under future extreme weather conditions.

[0066] Based on the failure probability of each component under different meteorological factors and combined with historical data on component failures under multiple meteorological factors, multimodal data fusion and a long short-term memory (LSTM) neural network are used to calculate the failure probability of components under the combined action of multiple meteorological factors. LSTM uses a gating mechanism to address the gradient vanishing and gradient exploding problems in traditional recurrent neural networks, effectively capturing long-term dependencies in time series data. The LSTM unit consists of an input gate, a forget gate, an output gate, and a cell state. The specific model is represented as follows:

[0067]

[0068] in, represents the weight parameter to be learned, is the bias vector of the lth layer network. In this model, the input of LSTM is the historical data of various meteorological factors and failure rates of key components in the power system, and the output is the predicted failure probability of the component. represents the input hidden state of the l-1th layer at the current time step t. Represents the hidden state output of this layer at the previous time step t-1. Represents the hidden state output of this layer at the previous time step t. Represents the cell state at the current time step t. represents the cell state at the previous time step t-1. Represents the input gate, which controls the extent to which the current input information is written into the cell state. Represents the forget gate, which controls how much of the state at the previous moment is retained in the current state. Represents the output gate, which controls how much information in the current cell state is output to the current hidden state. Represents the candidate state (candidate memory), which is the result of processing the raw input information at the current time step without gating. σ(·) represents the sigmoid activation function, with output values ​​in the range [0, 1]. tanh(·) represents the hyperbolic tangent activation function, used to calculate candidate states and final output.

[0069] S3: Combined with the failure probability of system components, the Monte Carlo method is used to evaluate the safety of the power system and output reliability indicators.

[0070] Based on the calculated predicted failure probability of each component of the power system under future extreme weather conditions, the Monte Carlo sampling method is used to perform multiple sampling to generate different system states. In each sampling, the fault status of each key component in the power system is randomly extracted and applied to the power system's flow and power scheduling model. Each sampling result produces a corresponding system state, which is whether each component in the system is in a fault state at that moment, and the corresponding power flow distribution, load distribution and voltage stability are calculated. The results of each sampling are used to analyze the overall operation of the system, including calculating the amount of load reduction and evaluating whether the system has islanding effects and power imbalance problems. By performing statistical analysis on the results of multiple samplings, the safety indicators of the power system are obtained, including the amount of load loss and reliability indicators under different extreme weather conditions.

[0071] The load loss amount includes, in each Monte Carlo sampling, calculating the load reduction amount through the optimal power flow algorithm according to the current system status, analyzing the system status to determine whether islanding effect and power imbalance occur; statistically analyzing multiple sampling results to obtain the load loss amount of the power system under extreme meteorological conditions.

[0072] Reliability indicators include calculating the state and load loss of the power system at each sampling through multiple Monte Carlo sampling, and combining mathematical statistics methods to obtain the sample mean of the system reliability indicator as the expected value; reliability indicators include power shortage probability, power shortage expectation and power shortage expectation.

[0073] Outputting reliability indicators includes, after the calculation is completed, judging whether the system safety indicators have converged; if converged, outputting the final reliability assessment results; if not, re-executing meteorological state sampling and sampling calculations until a converged reliability indicator result is obtained to ensure the accuracy of the assessment and the safety of the system.

[0074] It should be noted that probabilistic assessment methods for power system security are primarily divided into two categories: analytical and Monte Carlo methods. While analytical methods offer precise mathematical models and highly accurate reliability indicators, they are difficult to apply to large-scale power system security assessments because the number of system states to analyze increases exponentially with system size. Therefore, given these insurmountable shortcomings of analytical methods, the Monte Carlo method has gained widespread application.

[0075] The non-sequential Monte Carlo method mainly samples the state of each component in the system and obtains the state of the system by combining the states of all components. Assuming that the system consists of N components, the system state can be represented as an N-dimensional vector X = [x1, x2, 1 / 4, x N ], the operating state of any component is X iUse a random number r uniformly distributed in the interval [0,1] i To simulate, that is:

[0076]

[0077] Among them, p i is the failure probability of component i. i Indicates the operating status of the i-th component, which is a binary variable of 0 or 1 (fault or not). i represents a random variable uniformly distributed on the interval [0,1], used to simulate whether a component has failed. X represents an N-dimensional vector of system status, representing the operating or failure status of all components in the power system.

[0078] If F(X) represents the system state function and P(X) represents the probability of occurrence of the system state, the system safety index can be expressed as:

[0079]

[0080] Where W is the system state space. E(F(X)) represents the expected value of the system state function, which represents the weighted average of all possible indicators (such as load loss and failure probability). F(X) represents the system state function, which is used to evaluate a certain performance indicator (such as whether the power supply demand is met) when the system is in state X.

[0081] According to the relevant knowledge of mathematical statistics, the sample mean of F(X) can be used as an unbiased estimate of its expected value, that is:

[0082]

[0083] in: represents the estimated value of E(F(X)). i ) represents the system state function obtained by the i-th sampling. It represents the unbiased estimate of the expected value of the system performance index E(F(X)), that is, the value calculated approximately by sampling. n" represents the number of samples (that is, the number of samples taken in the Monte Carlo simulation). X i Represents the system state vector obtained by the i-th sampling.

[0084] Specific safety indicators are as follows:

[0085] Loss of Load Probability (LOLP) indicates the probability that the system cannot meet the load demand (i.e., load shedding occurs), which is expressed as:

[0086]

[0087] Where S is the set of all system states that cannot meet the load demand, and P(X) is the probability of occurrence of system state X.

[0088] The expected value of the system load reduction due to various reasons is expressed as:

[0089]

[0090] Among them: F EDNS (X) is the load reduction when the system is in state X.

[0091] The expected energy shortage (Excepted Energy Not Supplied, EENS) indicates the expected value of energy loss due to load reduction in a given time period. EENS and EDNS have the following relationship:

[0092] EENS=EDNS*T

[0093] Where T is the evaluation duration.

[0094] The power system security assessment process considering multiple meteorological factors is as follows:

[0095] Determine the initial parameters of the power system. Read the initial parameters of the power system, including the power output of coal-fired units and distributed power generation, the load of each node, the initial parameters of energy storage equipment, the network structure parameters of the power system, etc. Set the sampling frequency K = 1.

[0096] Initial climate factors are determined. Randomly sample the climate state of the system to obtain meteorological parameters, including temperature, strong winds, snowfall and other meteorological disaster factors, to determine the climate state of all components in the system.

[0097] Calculate the failure probability of each component. Based on the aforementioned power system component failure model that considers meteorological factors and the climate conditions of each component, the failure probability of each component in the power system under given initial weather parameters is obtained.

[0098] Monte Carlo sampling determines the state of each component. Based on the calculated failure probability of transmission lines, distributed generation (DGs), and other components, Monte Carlo sampling is applied to randomly sample the state of all components in the power system, generating a deterministic system state.

[0099] System status analysis and determination: Analyze whether the power system has islanding and power imbalance, apply optimal power flow to calculate the load reduction of the power system under the current sampling period, and calculate system reliability indicators.

[0100] Set the sampling times K=K+1 and check whether the reliability index converges. If the reliability index converges, output the reliability index of the power system; otherwise, redetermine the initial climate factors.

[0101] Example 2 is an embodiment of the present invention, which provides a power system security assessment method considering multiple meteorological sensitive factors. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0102] The experiments involved various power system components, including transmission lines, wind turbines, and energy storage devices. These components were all subject to meteorological factors such as wind speed, lightning strikes, precipitation, and water accumulation. All test components were considered for their behavior under extreme weather conditions, such as typhoons and heavy rain, which can impact component reliability. The failure probability of each component was modeled using real-world data.

[0103] Using a multimodal data fusion method based on AI inference, component failure models are established by combining various meteorological factors (such as wind speed, precipitation, and lightning frequency). Each component's failure probability model is constructed by calculating its failure risk under different meteorological conditions to obtain the predicted failure probability of each component under extreme weather conditions. A long short-term memory (LSTM) neural network model is then trained on this time series data, using the LSTM's gating mechanism to capture the long-term dependency between meteorological factors and component failure probability. This allows for more accurate predictions of the failure probability of power system components under future extreme weather conditions, while also being able to handle time-series fluctuations in meteorological factors.

[0104] The Monte Carlo method is used to assess the safety of the power system. Based on the calculated failure probabilities of each component, multiple random samplings are performed to generate different system states. Each sampling randomly selects the failure state of each component and, combined with the power system's power flow and power dispatch models, calculates the corresponding power flow distribution, load distribution, and voltage stability, thereby generating system reliability indicators and safety assessment results. In particular, the assessment further verifies the system's safety through the calculation of load reduction, islanding effects, and power imbalance.

[0105] To ensure the accuracy of the evaluation results, this embodiment also provides a reliability indicator convergence judgment mechanism. If the reliability indicators of multiple sampling results converge, the final evaluation result is output. If not, meteorological state sampling and sampling calculations are continued until a converged reliability result is obtained.

[0106] Through this method, this embodiment can accurately assess the reliability and stability of the power system when facing extreme weather conditions caused by different meteorological factors, and provide scientific data support for power dispatch and emergency response. The experimental data table is shown in Table 1.

[0107] Table 1 Experimental data table

[0108]

[0109] Through the above experimental data, the safety performance of different power system components under extreme meteorological conditions is compared, thereby verifying the beneficial effects of the power system security assessment method that considers multiple meteorological sensitive factors.

[0110] Failure probability data reveals the failure risks of different components under the influence of meteorological factors. For example, the failure probability of component C is 6.1%, significantly higher than the 2.9% of component D. This suggests that component C may be more vulnerable, particularly in high wind speeds or other extreme weather conditions, where the risk of failure is greater. Implementing predictions based on multimodal data fusion and neural network technology allows us to more precisely capture these differences, providing a more accurate failure risk assessment for component C.

[0111] The Underpower Probability (LOLP) and Expected Energy Shortage (EENS) metrics also provide crucial information about power system security. For example, component B has a 2.3% underpower probability. This higher underpower probability indicates a higher likelihood of power outages during extreme weather conditions. This data analysis can effectively identify vulnerable links in the system that could potentially fail, allowing for the proactive development of emergency response measures.

[0112] Load reduction directly reflects the amount of load the system would be unable to provide in the event of a fault. For example, Component C's load reduction of 18.5 MW indicates that the power system could experience a significant load reduction in the event of a fault, potentially impacting power supply stability. Further calculations combined with system state stability (such as Component D's 99.0% stability) can better assess system operational safety under various extreme weather conditions.

[0113] In terms of system stability, most components (such as components D and E) exhibit high system stability (nearly 99%), indicating that they can effectively withstand the impact of various weather factors and ensure the overall safety of the system. However, the stability of components C and F is slightly lower, which may indicate that these components are less reliable in extreme weather conditions. Therefore, their operating status should be closely monitored and timely maintenance or replacement should be carried out.

[0114] Through this experiment and data analysis, it is clear that by implementing a power system security assessment method that considers multiple meteorological sensitivity factors, it is possible to identify potential risks to the power system in advance under different meteorological conditions and provide effective decision support. Compared with existing technologies, this invention can effectively address the assessment errors caused by changes in meteorological factors in existing methods and provide more accurate and comprehensive assessment results, showing significant innovation and advantages.

[0115] Example 3, reference Figure 2 , which is an embodiment of the present invention, provides a power system security assessment system that considers multiple meteorological sensitive factors, including a power system component fault model establishment module 100, a power system security assessment module 200, and a reliability index output and convergence judgment module 300.

[0116] S4: Establishing a power system component fault model The module 100 is used to establish a fault model under the influence of multiple meteorological factors based on the coupling effects of different meteorological factors on power system components.

[0117] The power system component fault model building module 100 includes a meteorological sensing data processing submodule 101 and a component fault probability modeling submodule 102 .

[0118] Furthermore, the meteorological sensing data processing submodule 101 is used to acquire regional meteorological data, including multidimensional meteorological factors such as wind speed, precipitation, lightning strikes, and flooding, and then performs rasterization processing and time series normalization to construct a standardized meteorological feature input set. The component failure probability modeling submodule 102 is used to construct dynamic failure probability models based on the interaction mechanism of meteorological factors for key components such as transmission lines, wind turbines, and energy storage devices, combining their structural characteristics with regional meteorological data through methods such as cascade failure models, Poisson process modeling, and wind speed segment risk functions.

[0119] It should be noted that the meteorological sensing data processing submodule 101 is the starting point for establishing the power system component failure model module 100, providing standardized input data for the subsequent modeling process. The component failure probability modeling submodule 102 is the core of the module, integrating meteorological factors with the physical characteristics of power components for coupled analysis to form a precise meteorological risk mapping relationship.

[0120] It should also be noted that the power system component fault model module 100 is a basic module of the system. Its output directly affects the accuracy of subsequent fault prediction and system safety assessment, and provides accurate data support for system reliability analysis.

[0121] S5: The power system security assessment module 200 is used to calculate the failure probability of each component of the power system under extreme weather conditions by using multimodal data fusion and neural network technology.

[0122] The power system security assessment module 200 includes a multimodal fusion and time series modeling submodule 201 and a fault probability prediction submodule 202 .

[0123] Furthermore, the multimodal fusion and time series modeling submodule 201 performs multimodal alignment and fusion of meteorological data from different time scales and spatial resolutions with historical power component operating data, and employs an LSTM neural network model to construct a time series learning structure. The fault probability prediction submodule 202, based on the trained LSTM model, inputs a sequence of multiple meteorological factors and outputs the predicted failure probability of the corresponding component at a future moment.

[0124] It should be noted that the multimodal fusion and time series modeling submodule 201 is responsible for building a unified data input framework and utilizing the input, forget, and output gates in the LSTM model to process complex nonlinear relationships. The fault probability prediction submodule 202, acting as the model execution unit, dynamically outputs the failure probability of each component in the power system based on the fused meteorological behavior data.

[0125] It should also be noted that the power system security assessment module 200 is an important transition link from static modeling to dynamic prediction. Its performance depends on the quality of the fault model established in module 100, and also determines the system's risk response capabilities under future meteorological evolution.

[0126] S6: Outputting reliability index and judging convergence The module 300 is used to evaluate the safety of the power system by Monte Carlo method based on the failure probability of system components and output the reliability index.

[0127] The reliability index output and convergence judgment module 300 includes a system state simulation submodule 301 and a safety index output and convergence judgment submodule 302 .

[0128] The system state simulation submodule 301 is used to perform multiple rounds of Monte Carlo sampling on the power system components based on the fault probability calculated by the previous module, generate different operating state combinations, and perform power flow calculation and load matching processing based on the power grid topology structure to form a system operating state set covering different meteorological scenarios.

[0129] The safety indicator output and convergence judgment submodule 302 is used to calculate reliability indicators such as the under-power probability (LOLP), expected under-power situation (EDNS), and expected energy shortage situation (EENS) after obtaining sufficient simulation status, and evaluate their convergence. If the convergence criteria are not met, the next round of sampling simulation iteration is automatically triggered until the indicators stabilize, and the evaluation results are finally output.

[0130] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0131] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0132] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0133] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having logic gate circuits for implementing logical functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc. It should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention and are not intended to be limiting. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced with equivalents without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications should be encompassed by the claims of the present invention.

Claims

1. A power system security assessment method considering multiple meteorological sensitive factors, characterized by: include: Based on the coupling effect of different meteorological factors on power system components, a fault model under the influence of multiple meteorological factors is established; Utilize multimodal data fusion and neural network technology to calculate the failure probability of various power system components under extreme weather conditions; Combined with the failure probability of system components, the Monte Carlo method is used to evaluate the safety of the power system and output reliability indicators; The Monte Carlo method is used to evaluate the safety of power systems. This involves simulating the power system using the Monte Carlo sampling method based on the failure probability of each component in the power system under extreme weather conditions, randomly sampling the status of each component in the power system, and combining the system's power flow and power scheduling model to generate the system's state distribution; and obtaining the system's safety assessment results through statistical analysis of multiple system states.

2. The power system security assessment method considering multiple meteorological sensitive factors according to claim 1, characterized in that: The establishment of a fault model under the influence of multiple meteorological factors includes: Obtain historical fault data of key components in the power system and analyze the impact of meteorological factors on the components; For key components, AI deduction and multimodal data fusion methods are used to establish specific failure probability models to obtain the failure risks of various components in the power system under different extreme weather conditions.

3. The power system security assessment method considering multiple meteorological sensitive factors according to claim 1 or 2, characterized in that: The calculation of the failure probability of each component of the power system under extreme weather conditions includes: After determining the meteorological factors and component failure risks, a long short-term memory neural network model is used to analyze the failure probability of components under multiple meteorological factors. The long short-term memory neural network model processes time series data through input gates, forget gates and output gates, captures the long-term dependency between meteorological factors and failure probability, and outputs the predicted failure probability of each component of the power system under future extreme weather conditions.

4. The power system security assessment method considering multiple meteorological sensitive factors according to claim 3 is characterized in that: The Monte Carlo method for evaluating the security of power systems includes: Based on the calculated predicted failure probability of each power system component under future extreme weather conditions, the Monte Carlo sampling method is used to perform multiple sampling to generate different system states. In each sampling, the fault status of each key component in the power system is randomly extracted and applied to the power system's power flow and power dispatch model. Each sampling result generates a corresponding system state, indicating whether each component in the system is in a fault state at the corresponding moment, and the corresponding power flow distribution, load distribution and voltage stability are calculated; The results of each sampling are used to analyze the overall operation of the system, including calculating the load reduction and evaluating whether the system has islanding effects and power imbalance problems; By statistically analyzing the results of multiple samplings, we can obtain the safety indicators of the power system, including the load loss and reliability indicators under different extreme weather conditions.

5. The power system security assessment method considering multiple meteorological sensitive factors according to claim 4 is characterized in that: The load loss includes: In each Monte Carlo sampling, the load reduction amount is calculated using the optimal power flow algorithm based on the current system state, and the system state is analyzed to determine whether islanding effects and power imbalances occur. By statistically analyzing multiple sampling results, we can obtain the load loss of the power system under extreme meteorological conditions.

6. The power system security assessment method considering multiple meteorological sensitive factors according to claim 5 is characterized in that: The reliability indicators include: Through multiple Monte Carlo sampling, the state and load loss of the power system at each sampling are calculated, and the sample mean of the system reliability index is obtained as the expected value by combining mathematical statistics methods; Reliability indicators include power shortage probability, power shortage expectation and power shortage expectation.

7. The power system security assessment method considering multiple meteorological sensitive factors according to any one of claims 1, 2, 4 or 6, characterized in that: The output reliability indicators include: After the calculation is completed, determine whether the system security index has converged; If converged, the final reliability assessment result is output; if not, the meteorological state sampling and sampling calculation are re-executed until a converged reliability index result is obtained to ensure the accuracy of the assessment and the safety of the system.

8. A power system security assessment system considering multiple meteorological sensitive factors, characterized by: It includes a power system component fault model establishment module (100), a power system security assessment module (200), and a reliability index output and convergence judgment module (300); The power system component fault model establishment module (100) is used to establish a fault model under the influence of multiple meteorological factors based on the coupling effects of different meteorological factors on power system components; The power system security assessment module (200) is used to calculate the failure probability of each component of the power system under extreme weather conditions using multimodal data fusion and neural network technology; The module (300) for outputting reliability index and judging convergence is used to evaluate the safety of the power system by Monte Carlo method in combination with the failure probability of system components, and output the reliability index.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the power system security assessment method considering multiple meteorological sensitive factors according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the power system security assessment method considering multiple meteorological sensitive factors according to any one of claims 1 to 7 are implemented.

Images