Running state risk assessment and early warning method, system and equipment of electric power system under extreme weather condition and medium

Abstract

The invention discloses an electric power system operation state risk assessment early warning method under extreme weather conditions, and relates to the technical field of electric power system safety operation and risk early warning, and the method comprises the steps: inputting global meteorological data to a load prediction architecture, and outputting a load prediction result; constructing a power system fault probability model according to a load prediction result; determining an initial fault operation scene result of the power system in the extreme weather based on the power system fault probability model; based on the N-k cascading failure propagation path, constructing an operation evaluation scheme according to an initial failure operation scene result of the power system in extreme weather; the operation scene high risk in the current extreme weather is determined based on the operation evaluation scheme, and early warning is performed according to the operation scene high risk in the extreme weather, so that the evaluation and early warning capability of the electric power system operation state risk quantification is improved, and the accuracy of the electric power system operation state risk evaluation and early warning in the extreme weather is improved.

Description

Technical Field

[0001] The present application relates to the technical field of safe operation and risk warning of power systems, and in particular to a method, system, equipment and medium for risk assessment and warning of the operation status of power systems under extreme weather conditions. Background Art

[0002] In modern society, the stability of the power system is crucial to ensuring social operation and economic development. Extreme weather events, such as typhoons, heavy rains, and high temperatures, have an increasingly significant impact on the power system, causing damage to power equipment, a sharp increase in load demand, or a decrease in power supply capacity, triggering chain failures, and seriously threatening the safe operation of the power system. Therefore, how to accurately assess the impact of extreme weather on the system and provide timely warnings and responses has become an urgent problem for the power system.

[0003] Traditional power system risk assessment methods often rely on historical data and empirical judgment, and are difficult to adapt to the complex and changeable operating environment brought about by extreme weather.

[0004] In order to meet these challenges, the risk assessment and early warning technology of power systems needs to be further improved and optimized. Existing research and practice have shown that data-driven methods combined with advanced machine learning algorithms can provide more accurate and real-time predictions of power system operation status, thereby providing decision support for the safe operation of power systems. However, these methods are not accurate enough in dealing with the risk assessment and early warning of power system operation status under extreme weather conditions. There are often problems such as insufficient consideration of the impact of extreme weather, limited ability to identify the propagation path of cascading faults, and insufficient quantitative assessment of the risk of power system operation status, which leads to inaccurate risk assessment and early warning of power system operation status under extreme weather conditions. Summary of the invention

[0005] The present application provides a method, system, device and medium for risk assessment and early warning of the operating status of an electric power system under extreme weather conditions, so as to solve the problem of low accuracy of risk assessment and early warning of the operating status of the electric power system caused by the impact of extreme weather on the operating status of the electric power system.

[0006] To achieve the above objectives, this application provides the following solutions:

[0007] In a first aspect, the present application provides a method for risk assessment and early warning of the operation status of a power system under extreme weather conditions, comprising:

[0008] The global meteorological data is input into the load forecasting framework, and the load forecasting result is output; the load forecasting framework is constructed based on a graph neural network; the load forecasting framework includes a feature extraction layer, a feature transformation layer and a prediction output layer connected in sequence, wherein the feature extraction layer is used to extract the first high-dimensional meteorological features of the global meteorological data in the power supply area, and mine the short-term time series features and long-term time series features implicit in the load data of the global meteorological data to obtain the second high-dimensional meteorological features and the third high-dimensional meteorological features; the feature transformation layer is used to perform linear transformation on the first high-dimensional meteorological features, the second high-dimensional meteorological features and the third high-dimensional meteorological features respectively to obtain the fourth high-dimensional meteorological features, the fifth high-dimensional meteorological features and the sixth high-dimensional meteorological features; the prediction output layer is used to perform weighted fusion on the fifth high-dimensional meteorological features and the sixth high-dimensional meteorological features to output the load forecasting results under extreme weather conditions.

[0009] A power system failure probability model is constructed according to the load forecasting results.

[0010] Based on the power system failure probability model, the initial failure operation scenario results of the power system under extreme weather conditions are determined.

[0011] Based on the Nk cascading fault propagation path, an operation evaluation plan is constructed according to the initial fault operation scenario results of the power system under the extreme weather; the operation evaluation plan includes an operation evaluation plan for the cascading fault propagation stage and an operation evaluation plan for the optimization scheduling stage.

[0012] The high-risk operation scenarios under the current extreme weather conditions are determined according to the operation assessment plan, and warnings are issued according to the high-risk operation scenarios under the extreme weather conditions.

[0013] In a second aspect, the present application provides an operation status risk assessment and early warning system for a power system under extreme weather conditions, comprising:

[0014] The load forecasting module is used to input global meteorological data into the load forecasting framework and output the load forecasting results.

[0015] The power system initial fault operation scenario result output module is used to determine the power system initial fault operation scenario result under extreme weather conditions based on the power system fault probability model.

[0016] The high-risk warning module is used to determine the high-risk of the operation scenario under the current extreme weather according to the operation evaluation plan, and to issue a warning based on the high-risk of the operation scenario under the extreme weather.

[0017] In a third aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of risk assessment and early warning of the operating status of the power system under any of the extreme weather conditions described above.

[0018] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of risk assessment and early warning of the operating status of the power system under any of the extreme weather conditions described above.

[0019] According to the specific embodiments provided in this application, this application has the following technical effects:

[0020] The present application provides a method for risk assessment and early warning of the operating status of an electric power system under extreme weather conditions, which inputs global meteorological data into a load forecasting architecture and outputs load forecasting results; the load forecasting architecture is constructed based on a graph neural network; the load forecasting architecture includes a feature extraction layer, a feature transformation layer and a prediction output layer connected in sequence, wherein the feature extraction layer is used to extract the first high-dimensional meteorological features of the global meteorological data in the power supply area, and mine the short-term time series features and long-term time series features implicit in the load data of the global meteorological data to obtain the second high-dimensional meteorological features and the third high-dimensional meteorological features; the feature transformation layer is used to linearly transform the first high-dimensional meteorological features, the second high-dimensional meteorological features and the third high-dimensional meteorological features, respectively, to obtain the fourth high-dimensional meteorological features, the fifth high-dimensional meteorological features and the sixth high-dimensional meteorological features; the prediction output layer is used to transform the fifth high-dimensional meteorological features into the first high-dimensional meteorological features, the second high-dimensional meteorological features and the third high-dimensional meteorological features, respectively. The characteristics are weightedly integrated with the sixth meteorological characteristic to output the load forecast results under extreme weather; based on the power system fault probability model, the initial fault operation scenario results of the power system under extreme weather are determined; based on the Nk cascading fault propagation path, an operation evaluation plan is constructed according to the initial fault operation scenario results of the power system under extreme weather; the operation evaluation plan includes the operation evaluation plan for the cascading fault propagation stage and the operation evaluation plan for the optimization scheduling stage; according to the operation evaluation plan, the high-risk operation scenarios under the current extreme weather are determined, and early warnings are issued according to the high-risk operation scenarios under extreme weather. By fully considering the impact of extreme weather and improving the ability to identify the cascading fault propagation path, the ability to quantify the risk assessment and early warning of the power system operation status is improved, and the accuracy of the risk assessment and early warning of the power system operation status under extreme weather is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 A schematic diagram of a process flow of a method for risk assessment and early warning of the operation status of a power system under extreme weather conditions provided in an embodiment of the present application;

[0023] Figure 2 A schematic diagram of the IEEE 39-node line topology provided for an embodiment of the present application;

[0024] Figure 3 A schematic diagram of the probability density distribution function of the line failure rate as a function of temperature provided in an embodiment of the present application;

[0025] Figure 4 A schematic diagram of load data prediction results under normal weather conditions provided in an embodiment of the present application;

[0026] Figure 5 A schematic diagram of the prediction results of load nodes under high temperature weather conditions provided in an embodiment of the present application. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0028] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0029] In modern society, the stability of the power system is crucial to ensuring social operation and economic development. Extreme weather events, such as typhoons, heavy rains, and high temperatures, have an increasingly significant impact on the power system, causing damage to power equipment, a sharp increase in load demand, or a decrease in power supply capacity, triggering chain failures, and seriously threatening the safe operation of the power system. With the intensification of global climate change, the frequency and intensity of extreme weather events are increasing, and the risks faced by the power system are also increasing. Therefore, how to accurately assess the impact of extreme weather on the power system and provide timely warnings and responses has become an urgent problem to be solved in the field of power systems.

[0030] Traditional power system risk assessment methods often rely on historical data and empirical judgments, and are difficult to adapt to the complex and changeable operating environment brought about by extreme weather. In addition, with the rapid development of new energy power generation technology, the proportion of intermittent energy such as wind power and photovoltaics in the power system continues to increase. The output of these energy sources has great uncertainty, which is superimposed on the fluctuations in the load demand of the power system, resulting in frequent imbalances in the supply and demand of active power in the interconnected power grid, causing frequency deviation problems. Extreme weather conditions will further aggravate this output fluctuation, posing greater challenges to the power balance and frequency stability of the power system.

[0031] In order to meet these challenges, the risk assessment and early warning technology of power systems needs to be further improved and optimized. Existing research and practice have shown that data-driven methods combined with advanced machine learning algorithms can provide more accurate and real-time predictions of power system operation status, thereby providing decision support for the safe operation of power systems. However, when dealing with the risk assessment of power system operation under extreme weather conditions, these methods often have problems such as insufficient consideration of the impact of extreme weather, limited ability to identify the propagation path of cascading faults, and insufficient quantitative assessment of power system operation risks. Therefore, developing a new high-risk early warning method for power system operation to meet the needs of safe operation of power systems under extreme weather conditions has become a research hotspot in this field.

[0032] like Figure 1 As shown, the embodiment of the present application provides a method for risk assessment and early warning of the operation status of a power system under extreme weather conditions, specifically comprising:

[0033] Step 101: Input global meteorological data into the load forecasting framework and output load forecasting results.

[0034] Step 102: constructing a power system failure probability model according to the load forecasting result.

[0035] Step 103: Based on the power system fault probability model, determine the initial fault operation scenario results of the power system under extreme weather conditions.

[0036] Step 104: Based on the Nk cascading fault propagation path, an operation evaluation plan is constructed according to the initial fault operation scenario results of the power system under the extreme weather.

[0037] Step 105: Determine the high-risk operation scenarios under the current extreme weather conditions according to the operation assessment plan, and issue an early warning based on the high-risk operation scenarios under the extreme weather conditions.

[0038] The present application provides a method for risk assessment and early warning of the operating status of an electric power system under extreme weather conditions. The method aims to improve the operating resilience of the electric power system under extreme weather conditions, and provide decision support for the planning and scheduling of the electric power system by accurately predicting load changes, simulating power equipment failures, and evaluating the risk of cascading failures, so as to reduce the losses caused by extreme weather to the power grid and enhance the overall resilience of the power grid.

[0039] First, this application first constructs a load forecasting architecture based on graph neural network feature fusion, which introduces global meteorological data to improve forecasting accuracy. Through this architecture, load changes under extreme weather conditions can be more accurately predicted, providing key input for the operational risk assessment of the power system.

[0040] The load forecasting architecture uses graph neural networks to extract meteorological information of each load node in the power system for feature extraction. GNN updates the meteorological features of the load node by aggregating the information of meteorological stations close to the load node, and achieves deeper feature learning through multi-layer stacking.

[0041] The extracted meteorological features are combined with historical load data to form a comprehensive feature set. This step is achieved through a linear transformation module, which maps high-dimensional features to the same feature space to facilitate the subsequent load forecasting model training. Based on the fused features, a load forecasting model is then trained. The model is able to predict the load demand at a specific point in the future based on the input meteorological information and historical load data.

[0042] Secondly, this application uses a non-sequential Monte Carlo sampling simulation method to construct a set of initial fault operation scenarios for power systems under extreme weather conditions. This method provides comprehensive initial conditions for subsequent risk assessment by simulating a variety of possible fault conditions that may occur in power systems under the influence of extreme weather.

[0043] Thirdly, this application designs an operation evaluation scheme based on the identification of Nk cascading fault propagation paths to screen out high-risk scenarios. This scheme can effectively identify the cascading fault propagation paths that may occur under extreme weather conditions, thereby identifying the vulnerable links in the power system and providing decision support for the operation risk management of the power system.

[0044] On this basis, an optimization dispatch model is provided to optimize the adjustment of the remaining generator nodes or load nodes after the cascading fault propagation to ensure the safety of the power system under the new operation mode. The model takes into account the economic factors of generator output adjustment and aims to minimize the impact of extreme weather on the power grid.

[0045] Fourthly, the present application verifies the effectiveness of the proposed method through simulation results. The simulation results show that the method of the present application can effectively identify high-risk fault scenarios under extreme weather conditions, provide decision-making support for reducing power grid losses, and enhance the resilience and stability of the power grid under extreme weather conditions. By implementing the above technical solutions, the present application can provide a powerful risk warning and management strategy for the safe operation of the power system under extreme weather conditions.

[0046] By implementing the above technical solutions, the present application can effectively assess and warn of high-risk operation risks of the power system under extreme weather conditions, reduce the negative impact on the stability of the power grid, and improve the overall resilience and operating performance of the power grid.

[0047] In this way, the embodiment of the present application not only accurately predicts the load results, but also enhances the sensitive monitoring capability of cascading failures caused by extreme weather while ensuring the safe operation of the power grid. The present application fully considers the impact of extreme weather on the power system and the vulnerability of power equipment components under abnormal working conditions, and describes the high-risk operation risk warning method of the power system as a comprehensive model combining data-driven and Nk cascading failure propagation path identification. On this basis, the present application adopts a non-sequential Monte Carlo sampling simulation method to strictly construct a set of initial fault operation scenarios of the power system under extreme weather. These criteria are applicable to complex conditions such as power constraints and failure probability uncertainty caused by meteorological factors. Through these criteria, the present application can achieve accurate early warning of high-risk operation risks of the power system under extreme weather, and can ensure the stability and reliability of the power grid even in the case of cascading failures caused by extreme weather. In addition, the early warning method proposed in the present application is highly adaptable and flexible, and can cope with the challenges of power grid operation under extreme weather conditions, especially when the uncertainty of new energy output is high, it can still maintain the stable operation of the power grid. Through simulation verification, the early warning method of this application not only reduces power grid losses and improves decision-making efficiency, but also improves the power grid's ability to respond to extreme weather and risk management level, proving its effectiveness and superiority in actual power grid operation.

[0048] Further, in an exemplary embodiment, step 101 may be replaced by the following steps.

[0049] Step 1011: The load prediction architecture is constructed based on a graph neural network, and the load prediction architecture includes a feature extraction layer, a feature transformation layer and a prediction output layer connected in sequence.

[0050] Step 1012: The feature extraction layer includes a GNN module (Graph Neural Network), a TCN module (Temporal Convolutional Network) and an LSTM module (Long Short-Term Memory). The GNN module (Graph Neural Network) is used to extract the first high-dimensional meteorological feature of the global meteorological data in the entire power supply area based on the meteorological site information. The TCN module (Temporal Convolutional Network) is used to capture the change pattern of the load data within a preset time range based on the first high-dimensional meteorological feature in combination with multi-layer causal convolution and dilated convolution. The LSTM module (Long Short-Term Memory) is used to mine the short-term time series features and long-term time series features implicit in the load data based on the change pattern of the load data within a preset time range to obtain the second high-dimensional meteorological feature and the third high-dimensional meteorological feature. Time series features refer to features in the data that change over time. They can be short-term (such as intraday changes) or long-term (such as seasonal changes). High-dimensional meteorological features are more complex feature representations that contain more information through feature extraction and transformation based on time series features. That is, high-dimensional meteorological features are an extension and deepening of time series features, and they increase in both dimension and complexity.

[0051] This layer contains three modules: GNN (Graph Neural Network), TCN (Temporal Convolutional Network) and LSTM (Long Short-Term Memory). Among them, the meteorological information of each load node has spatial correlation, so GNN (Graph Neural Network) is considered to extract high-dimensional meteorological features in the entire power supply area; while TCN (Temporal Convolutional Network) and LSTM modules (Long Short-Term Memory) are used to mine the short-term and long-term time series features implicit in the load data, respectively.

[0052] (1) Weather feature extraction module based on GNN (Graph Neural Network).

[0053] Assume that the historical records of different meteorological stations are V = [V 1,V 2 ,...,V Ξ ] T .in (φ=1,2,…,Ξ) is the historical meteorological data of the φth load node, is the meteorological data vector at the mth moment in the past, m = 1, 2, ..., M; the geographical distance set between the meteorological station and the load node is E = {e id |i=1,2,...,Ξ; d=1,2,...,N D},e id is the distance between meteorological station i and load node d; the historical load data of each load node is N D is the number of load nodes, and Ω is the load segment length.

[0054] In this application, GNN (Graph Neural Network) aggregates the information of meteorological stations close to the load node to update the meteorological characteristics of the load node, and achieves deeper feature learning through multi-layer stacking. The meteorological feature extraction process based on GNN (Graph Neural Network) is as follows:

[0055]

[0056] Where z∈[1,Z] is the number of layers of GNN (Graph Neural Network), is the aggregation result of the meteorological station φ at the zth layer, σ is the activation function, N(d) is the set of neighboring meteorological stations of load node d; R is the preset distance; |N(d)| is the number of neighboring meteorological stations; w (z) is the weight matrix; b (z) For bias.

[0057] (2) Load feature extraction module based on TCN and LSTM.

[0058] TCN (Temporal Convolutional Network) combines multi-layer causal convolution and dilated convolution to capture the change pattern of load data in a short time range. By adjusting the size of the convolution kernel and the step size of the expansion, TCN (Temporal Convolutional Network) can achieve the sensitivity of local features. The process of extracting local time series features by TCN (Temporal Convolutional Network) is shown in the formula

[0059]

[0060] Where k∈[1,Π] is the number of layers of TCN (Temporal Convolutional Network); is the output of the middle kth layer; C TCN is the size of the convolution kernel; is the convolution kernel weight; Φ (k) is the expansion rate; β (k) For bias.

[0061] LSTM (Long Short-Term Memory) uses a gating mechanism to achieve continuous retention of implicit features during the forward propagation process, thereby capturing long-term load characteristics.

[0062]

[0063] η t-ρ =o t-ρ tanh(C t-ρ ).

[0064] h LSTM =[η t-Ω ,η t-Ω+1 ,...,η 1 ].

[0065] In the formula, f t-ρ , t-ρ and t-ρ are the outputs of the forget gate, input gate, and output gate respectively; w F 、w I 、w C and w O is the weight matrix; v F 、v I 、v C and v O is bias; is the intermediate memory state; C t-ρ is the memory state at time t-ρ; η t-ρ is the implicit state at time t-ρ, h LSTM It is the output sequence of LSTM (Long Short-Term Memory).

[0066] Step 1013: The fourth high-dimensional meteorological feature is:

[0067] h' GNN =w GNN h GNN +b GNN .

[0068] The fifth high-dimensional meteorological characteristics are:

[0069] h' TCN =w TCN h TCN +b TCN .

[0070] The sixth high-dimensional meteorological characteristics are:

[0071] h' LSTM =w LSTM h LSTM +b LSTM .

[0072] Among them, h GNN is the first high-dimensional meteorological feature, h TCN is the second highest dimensional meteorological feature, h LSTM is the third high-dimensional feature, h' GNN is the fourth highest dimensional meteorological feature, h' TCN is the fifth high-dimensional meteorological feature, h' LSTM is the sixth high-dimensional meteorological feature, w GNN 、w TCN and w LSTM is the weight matrix, b GNN , b TCN and b LSTM For bias.

[0073] This layer contains three linear transformation modules, which respectively transform the high-dimensional features h obtained by the feature extraction layer. GNN 、h TCN and h LSTM Perform linear transformation. Among them, h GNN After linear transformation, a linear combination of meteorological characteristics is obtained to facilitate the prediction of the probability of extreme weather at a certain node; h TCN and h LSTM After linear transformation, they are mapped to the same feature space, which is conducive to the fusion of the two.

[0074] h′ GNN =w GNN h GNN +b GNN .

[0075] h T ' CN =w TCN h TCN +b TCN .

[0076] h L ' STM =w LSTM h LSTM +b LSTM .

[0077] In the formula, w GNN 、w TCN and w LSTM is the weight matrix, b GNN , b TCN and b LSTM For bias.

[0078] Step 1014: The prediction output layer specifically includes:

[0079] Using the formula Determine the probability of extreme weather occurring in load data; where λ d represents the probability of extreme weather occurring in the dth load data; h' GNN Represents the fourth high-dimensional meteorological characteristics.

[0080] Step 1015: Based on the probability of extreme weather occurrence of the dth load data, use the formula η d =λ d ·h′ TCN +(1-λ d )·h′ LSTM The fifth high-dimensional meteorological feature and the sixth meteorological feature are weightedly fused to determine the fused high-dimensional meteorological feature, wherein η d Represents the high-dimensional meteorological features after fusion; h' TCN Represents the fifth high-dimensional meteorological characteristics; h' LSTM Represents the sixth high-dimensional meteorological feature.

[0081] Step 1016: Based on the fused high-dimensional meteorological features, use formula P d,t =w Θ η d +b Θ Determine the load forecast result, where p d，t represents the load forecast result, w Θ represents the weight matrix, b Θ Indicates bias.

[0082] In order to realize load forecasting under extreme weather conditions, the two-way time series features h T ' CN and h L ' STM Weighted fusion is performed, and the size of the weight depends on the probability of extreme weather, which can be obtained by the transformed meteorological information feature h′ GNN It is obtained through logistic regression calculation. The specific process is as follows:

[0083] η d =λ d ·h T ' CN+(1-λ d )·h′ LSTM .

[0084] In the formula, λ d represents the probability of extreme weather occurring at the dth load node; η d Represents the fused time series features. Finally, the fused features are linearly transformed to output the load forecast results.

[0085] P d,t =w Θ η d +b Θ .

[0086] In the formula, w Θ is the weight matrix, b Θ For bias.

[0087] Further, in an exemplary embodiment, step 102 may be replaced by the following steps.

[0088] Step 1021: constructing a power system fault probability model according to the load forecast result, specifically including:

[0089] Based on the load forecast results, using the formula Determine the failure probability dependent on meteorological factors.

[0090] In the formula, f n,W is the failure probability of the nth line in the power system when the extreme weather intensity is X, n = 1, 2, ..., N L , N L is the number of lines; n,X , n,X are the mean and standard deviation of the failure probability of the ith line under extreme weather conditions.

[0091] Step 1022: Based on the load forecast result, use the formula Determine the failure probability of the road flow dependence.

[0092] In the formula, f n,0 is the failure probability of line n during normal operation, P n,r and P n,max are the rated power and maximum allowable power of line n respectively, P n,s is the power flow size of line n during the sth round of cascading failure propagation.

[0093] Step 1023: construct a power system fault probability model based on the fault probability dependent on meteorological factors and the fault probability dependent on power flow.

[0094] 1) Failure probability dependent on meteorological factors: The cumulative log-normal distribution function is widely used to describe the vulnerability of components under abnormal working conditions.

[0095] 2) Failure probability dependent on line flow: When the line operates normally, the probability of power line failure is often low; when the line flow is between the rated power and the maximum allowable power, the failure probability and the line flow are approximately linearly related; when the line flow exceeds the maximum allowable power, the line will inevitably be overloaded and disconnected.

[0096] Further, in an exemplary embodiment, step 103 may be replaced by the following steps.

[0097] Step 1031: Based on the power system fault probability model, using formula f n =min{δ n,W f n,W +δ n,F f n,F ,1}Determine the comprehensive failure probability of the line under extreme weather conditions.

[0098] Step 1032: Based on the comprehensive fault probability of the line, use the non-sequential Monte Carlo sampling simulation method to determine the initial fault operation scenario results of the power system under extreme weather conditions.

[0099] Under extreme weather failures, there is a risk of multiple lines failing simultaneously in the power system. To this end, this section uses a non-sequential Monte Carlo sampling simulation scheme to generate a series of operating scenarios. Specifically, at the kth sampling, for the i-th power line, a random number γ is randomly generated from the interval [0, 1]. i,k , so the state ε of power line i in one sampling ik satisfy:

[0100]

[0101] In the formula, 0 represents normal state and 1 represents fault state. For all lines in the power system, the operating state after one sampling is

[0102] E k ={ε 1,k ,ε 2,k ,...,ε N,k}.

[0103] By repeating the sampling operation K times, we can get K initial fault scenario sets, denoted as {E 1 , E 2 ,…,E K}.

[0104] Further, in an exemplary embodiment, step 104 may be replaced by the following steps.

[0105] Step 1041: Operation evaluation scheme for the cascading fault propagation stage: In the cascading fault propagation stage, the line power flow under the initial fault scenario is calculated based on the DC power flow model, and it is determined whether the power flow of the remaining lines exceeds the limit, and a first judgment result is obtained.

[0106] Step 1042: If the first judgment result is that the power flow of the remaining lines exceeds the limit, disconnect the faulty line of the power system, update the power line topology and perform cascading fault propagation.

[0107] Step 1043: If the first judgment result is that the power flow in the remaining lines does not exceed the limit, the cascading fault propagation ends.

[0108] Step 1044: Operation evaluation plan for the optimization dispatching stage: In the optimization dispatching stage, based on the operation status of the power system after the cascading fault propagation stage, an optimization dispatching plan is formulated according to the simulation dispatching center.

[0109] Step 1045: Determine the load loss amount and the engine output adjustment amount according to the optimized scheduling plan.

[0110] Design an operation evaluation scheme based on Nk cascading fault propagation path identification to screen out high-risk scenarios;

[0111] For the fault scenario set generated in step 103, it is obvious that there is a possibility that multiple lines may be disconnected at the same time under extreme weather conditions; and further, the large-scale flow transfer caused by the disconnection of multiple lines may cause some lines to exceed the flow limit and trip, causing cascading failures. For this reason, this section establishes a high-risk operation risk warning model that takes into account the propagation of Nk cascading faults, accurately identifies the fault propagation process of the power system under extreme weather conditions, identifies vulnerable lines, and provides a decision-making basis for the planning and dispatching department. The following is a detailed introduction.

[0112] Cascading failure propagation phase:

[0113] When multiple lines in the power system fail, it will trigger power flow transfer and subsequent line overload tripping risks, forming a chain fault propagation. To this end, in the fault propagation stage, the line power flow under the initial fault scenario is first calculated based on the DC power flow model, and then it is determined whether there are any remaining lines with power flow exceeding the limit. If the power flow exceeds the limit, the corresponding fault line is further disconnected, and the power line topology is updated and the next round of fault propagation is entered; conversely, if the power flow does not exceed the limit, the chain fault propagation ends. After each round of chain fault propagation is completed, the set of fault lines in that round is updated, that is,

[0114]

[0115] Where, L sis the set of faulty lines, s∈Z + Indicates the number of fault propagation times; l n is the line number, It is a 0-1 variable. When the value is 0, it means that line n has a power flow over-limit accident in the sth round of faults and is out of operation.

[0116] The following constraints must be met during the power flow calculation process:

[0117] (1) Node power balance constraints

[0118] For any node, the power balance constraint needs to be satisfied in the sth round of fault propagation, that is,

[0119] A s P s =B G,s P G,s -C D,s P D,s .

[0120] In the formula, A s is N×N L The node-line matrix, A s middle element a in,s (i=1,2,…,N; n=1,2,…,N L ) is 1, indicating that node i is connected to line n, and the connection direction is from line n to node i; -1, indicating that the connection direction is from node i to line n; and 0, indicating that node i is not connected to line n; B G is N×N G Order node-generator matrix, B G,s Element b ig,s (i=1,2,…,N; g=1,2,…,N G ) takes a value of 1, indicating that node i is connected to generator node g, and a value of 0 indicates that they are not connected; C D,s is N×N D Order node-load matrix, C D,s Element C id,s (i=1,2,…,N; d=1,2,…,N D ) takes a value of 1, indicating that node i is connected to load node d, and a value of 0 indicates that they are not connected; s is the line power flow vector during the sth round of cascading fault propagation, with a dimension of N L ×1 order; P G,s is the generator output vector during the sth round of cascading fault propagation, with a dimension of N G ×1 order; P D,s is the load vector of the sth round of cascading failure propagation, with dimension N D×1 level.

[0121] For line power flow P s The calculation results are obtained using the DC model:

[0122] P s =B L,s θ s .

[0123] In the formula, B L,s is the admittance matrix of the sth round of cascading fault propagation, with dimension N L ×N L B L,s Element b n,s (n=1,2,…,N L ) is the admittance of line n. When line n is disconnected due to extreme weather or cascading fault propagation, the admittance of the line becomes 0, otherwise the line admittance remains unchanged; θ s is the phase angle vector of the line, with dimension N L ×1 level.

[0124] When the sth round of cascading fault propagation occurs, whether any line has a power flow over-limit trip can be judged by the following formula:

[0125]

[0126] When the sth round of cascading fault propagation occurs, the set of faulty lines can be recorded as

[0127] L s ={n|n=1,2,...,N L ; ε n,s =1}.

[0128] In addition, it should be noted that if there is an extreme weather failure or the s-1 round of cascading failure propagation, If it is determined that the limit is exceeded and all lines connected to the generator node or load node are disconnected, the power of the corresponding generator node or load node needs to be set to 0 before the sth round of cascading failure determination. When the generator node or load node exits and causes the node active power to be unbalanced, it violates the formula P s =B L,s θ s When the active power of the balancing node is adjusted, the convergence of the DC power flow calculation can be guaranteed.

[0129] After the cascading fault propagates, the simulation dispatch center often needs to optimize and adjust the remaining generator nodes or load nodes to ensure the safety of the power system under the new operating mode. To this end, in this stage, the simulation dispatch center formulates a dispatch plan based on the grid operation status after the cascading fault propagates. Ultimately, the propagation path of the cascading fault and the load loss and generator output adjustment in the optimized dispatch process are formed. For the kth simulated fault scenario:

[0130] Cascading failure propagation path:

[0131] E k →L 1 →L 2 →...→L S .

[0132] Load loss:

[0133]

[0134] Power generation adjustment:

[0135]

[0136] Where ΔP d,D , ΔP g,G are load and generator adjustments respectively.

[0137] The simulation dispatch center generally aims to minimize load loss to reduce the impact of extreme weather on the power grid. This application further considers the economic factors of generator output adjustment and establishes an optimization dispatch model. The objective function is the operating cost of the fault propagation stage and the optimization dispatch stage, as shown in the following formula

[0138] J=minμ D P loss,D +μ G P G,∑ .

[0139] Where, J is the power grid operation loss after extreme weather occurs, μ D , μ G is the cost coefficient.

[0140] Assume that the number of rounds of cascading failure propagation is S, then the relevant constraints are:

[0141] a) Power balance constraint: Any node must satisfy the power supply and demand conservation, that is,

[0142]

[0143] In the formula, They are the line flow, generator output and load vector after optimized scheduling. Obviously,

[0144]

[0145] b) Line flow constraint: After optimization and adjustment, the flow of each line does not exceed the flow limit and meets the DC flow calculation conditions.

[0146]

[0147] In the formula, is the power angle of node n, θ max is the maximum power angle.

[0148] c) Phase angle limit constraint: To ensure the stability of the power line, the power angle of any node must meet a certain range.

[0149]

[0150] e) Generator output limit constraints: The active output of the generator has upper and lower limits.

[0151]

[0152] Where P g,min , P g,max are the minimum and maximum allowable outputs of the g-th generator respectively.

[0153] f) Load adjustment constraint: Under extreme weather conditions, the load can be adjusted from 0 to the load demand at the initial moment, that is,

[0154]

[0155] For the optimization scheduling model, the objective function contains the absolute value term |ΔP g,G |, it is impossible to directly use linear programming tools to solve it. Therefore, an auxiliary variable is introduced

[0156]

[0157] Therefore, the optimization scheduling model can be transformed into a linear programming model and solved using the same scheme.

[0158] Further, in an exemplary embodiment, step 105 may be replaced by the following steps.

[0159] Step 1051: Repeat the cascading fault propagation stage and the optimization scheduling stage K times, and you can get the cascading fault propagation path and corresponding adjustment cost caused by extreme weather on the power grid under the initial fault scenario set. Then, arrange them in descending order according to the adjustment cost, and you can get the danger level of the operation scenario under extreme weather, which is used by the operation planning department to formulate relevant disaster prevention and mitigation measures.

[0160] The effectiveness of the proposed method is verified in a simulation environment, and the superior performance of the proposed solution in extreme high temperature weather is demonstrated by comparing different control schemes. The present invention constructs a simulation model in the MATLAB / Simulink environment, and selects a city in the northwest for testing the operating performance in extreme high temperature weather. The line topology is as follows: Figure 2 As shown, the red numbers 1-39 represent the line numbers, the black numbers 1-39 represent the node numbers, and the probability density distribution function of the line failure rate changing with temperature is as follows Figure 3 As shown, Table 1 is a parameter table of the simulation model for setting simulation parameters. The parameters of the simulation model are set according to the simulation parameters in Table 1, including a fault probability weight coefficient, a preset distance, a cost coefficient, a generator output coefficient, etc.

[0161] Table 1 Simulation parameter settings Parameters of the simulation model

[0162]

[0163] In order to verify the prediction performance of the proposed model under extreme weather conditions, this application selects LSTM (Long Short-Term Memory), TCN (Temporal Convolutional Network), LSTM-TCN (LT) and other methods to predict the load value of a single node within a day under normal weather, high temperature weather and extremely high temperature weather, and compares them with the GLT method proposed in this application. In order to evaluate the performance of the algorithm, this application uses the root mean square error (RMSE) and mean absolute error (MAE) to quantify the deviation of the model prediction. The calculation formulas for RMSE and MAE are as follows:

[0164]

[0165] The results are shown in Tables 2 to 4, where Table 2 is a comparison of prediction results under normal weather conditions, Table 3 is a comparison of prediction results under high temperature weather conditions, and Table 4 is a comparison of prediction results under extreme high temperature weather conditions.

[0166] By analyzing the results in the table, we can see that: 1) Under normal weather conditions, users' electricity consumption habits remain stable, and the long-term time series characteristics of the load are more significant. At this time, the accuracy of LSTM (Long Short-Term Memory) is better than TCN (Temporal Convolutional Network), RMSE decreases by 32.43%, and MAE decreases by 25.44%; 2) In extremely hot weather, users' electricity consumption behavior changes significantly, and the short-term time series characteristics of the load are more significant. At this time, the accuracy of TCN (Temporal Convolutional Network) is better than LSTM (Long Short-Term Memory), RMSE decreases by 47.04%, and MAE decreases by 40.87%; 3) Under high temperature weather conditions, the influence of long-term and short-term time series characteristics of the load on the results is relatively balanced. At this time, the accuracy of LSTM (Long Short-Term Memory) and TCN (Temporal Convolutional Network) is better than that of LSTM (Long Short-Term Memory). The proposed GLT method introduces weather information, so that the prediction model can dynamically adjust the fusion ratio of long-term and short-term time series features according to weather conditions. Compared with LSTM (Long Short-Term Memory), TCN (Temporal Convolutional Network) and LT methods, the RMSE of the proposed method decreased by 35.36%, 32.44%, and 33.66%, respectively, and the MAE decreased by 34.82%, 26.71%, and 36.91%, respectively.

[0167] In order to more intuitively reflect the prediction effect of the proposed method, this application selects load nodes to compare the prediction results of LSTM (Long Short-Term Memory), TCN (Temporal Convolutional Network), LT and GTL methods under different weather conditions. Figure 4 and Figure 5 .

[0168] Table 2 Comparison of forecast results under normal weather conditions

[0169] LSTM TCN LT GLT RMSE 6.73 9.96 7.84 4.35 MAE 5.57 7.47 6.91 3.63

[0170] Table 3 Comparison of prediction results under high temperature weather conditions

[0171] LSTM TCN LT GLT RMSE 10.37 8.07 6.95 4.61 MAE 8.00 7.12 5.13 3.76

[0172] Table 4 Comparison of prediction results under extreme high temperature weather conditions

[0173] LSTM TCN LT GLT RMSE 14.84 7.86 12.44 5.31 MAE 11.5 6.80 9.35 4.29

[0174] The present application provides a risk assessment and early warning system for the operation status of a power system under extreme weather conditions, specifically comprising:

[0175] The load forecasting module is used to input global meteorological data into the load forecasting framework and output the load forecasting results.

[0176] The power system initial fault operation scenario result output module is used to determine the power system initial fault operation scenario result under extreme weather conditions based on the power system fault probability model.

[0177] The high-risk warning module is used to determine the high-risk of the operation scenario under the current extreme weather according to the operation evaluation plan, and to issue a warning based on the high-risk of the operation scenario under the extreme weather.

[0178] The present application provides a computer device, which specifically includes:

[0179] A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the above-mentioned power system load forecasting and fault propagation risk warning methods. The present application also provides a computer-readable storage medium, specifically comprising:

[0180] The computer program instructions for causing the processor to execute any of the above methods are stored.

[0181] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0182] The present application uses specific examples to illustrate the principles and implementation methods of the present application. The above examples are only used to help understand the method and core ideas of the present application. At the same time, for those skilled in the art, according to the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present application.

Claims

1. A method for risk assessment and early warning of the operation status of a power system under extreme weather conditions, characterized in that: The power system operation status risk assessment and early warning method comprises: The global meteorological data is input into the load forecasting framework, and the load forecasting result is output; the global meteorological data includes meteorological observation data, meteorological product data and Internet meteorological data; the load forecasting framework is constructed based on a graph neural network; the load forecasting framework includes a feature extraction layer, a feature transformation layer and a prediction output layer connected in sequence, wherein the feature extraction layer is used to extract the first high-dimensional meteorological features of the global meteorological data in the power supply area, and mine the short-term time series features and long-term time series features implicit in the load data of the global meteorological data to obtain the second high-dimensional meteorological features and the third high-dimensional meteorological features; the feature transformation layer is used to perform linear transformation on the first high-dimensional meteorological features, the second high-dimensional meteorological features and the third high-dimensional meteorological features respectively to obtain the fourth high-dimensional meteorological features, the fifth high-dimensional meteorological features and the sixth high-dimensional meteorological features; the prediction output layer is used to perform weighted fusion on the fifth high-dimensional meteorological features and the sixth high-dimensional meteorological features to output the load forecasting results under extreme weather conditions; Constructing a power system failure probability model according to the load forecasting results; Based on the power system failure probability model, determining the initial failure operation scenario results of the power system under extreme weather conditions; Based on the Nk cascading fault propagation paths, an operation evaluation plan is constructed according to the results of the initial fault operation scenario of the power system under the extreme weather; the operation evaluation plan includes an operation evaluation plan for the cascading fault propagation stage and an operation evaluation plan for the optimization scheduling stage; N is the number of normally operating components in the power system operation, and k is the number of faulty components in the system; The high-risk operation scenarios under the current extreme weather conditions are determined according to the operation assessment plan, and warnings are issued according to the high-risk operation scenarios under the extreme weather conditions.

2. The method for risk assessment and early warning of the operation status of a power system under extreme weather conditions according to claim 1 is characterized in that: The feature extraction layer includes a GNN module, a TCN module and an LSTM module; Utilizing the GNN module, extracting the first high-dimensional meteorological features of the global meteorological data in the entire power supply area based on the meteorological station information; Using the TCN module, based on the first high-dimensional meteorological features, combined with multi-layer causal convolution and dilated convolution, to capture the change pattern of the load data within a preset time range; By utilizing the LSTM module, based on the change pattern of the load data within a preset time range, the short-term time series features and long-term time series features implicit in the load data are mined to obtain the second high-dimensional meteorological features and the third high-dimensional meteorological features.

3. The method for risk assessment and early warning of the operation status of a power system under extreme weather conditions according to claim 1 is characterized in that: The fourth high-dimensional meteorological characteristics are: h' GNN =w GNN h GNN +b GNN ; The fifth high-dimensional meteorological characteristics are: h' TCN =w TCN h TCN +b TCN ; The sixth high-dimensional meteorological characteristics are: h' LSTM =w LSTM h LSTM +b LSTM ; Among them, h GNN is the first high-dimensional meteorological feature, h TCN is the second highest dimensional meteorological feature, h LSTM is the third high-dimensional feature, h' GNN is the fourth highest dimensional meteorological feature, h' TCN is the fifth high-dimensional meteorological feature, h' LSTM is the sixth high-dimensional meteorological feature, w GNN 、w TCN and w LSTM is the weight matrix, b GNN , b TCN and b LSTM For bias.

4. According to claim 1, a method for risk assessment and early warning of the operation status of a power system under extreme weather conditions is characterized in that: The prediction output layer specifically includes: Using the formula Determine the probability of extreme weather occurring in load data; where λ d represents the probability of extreme weather occurring in the dth load data; h' GNN Represents the fourth highest dimensional meteorological characteristics; According to the probability of extreme weather occurrence of the dth load data, the formula η d =λ d ·h T ' CN +(1-λ d )·h L ' STM The fifth high-dimensional meteorological feature and the sixth meteorological feature are weightedly fused to determine the fused high-dimensional meteorological feature; wherein, η d Represents the high-dimensional meteorological features after fusion; h' TCN Represents the fifth high-dimensional meteorological characteristics; h' LSTM Represents the sixth high-dimensional meteorological feature; According to the fused high-dimensional meteorological features, using formula P d,t =w Θ η d +b Θ Determine the load forecast result; where p d，t Indicates the load forecast result; w Θ represents the weight matrix, b Θ Indicates bias.

5. The method for risk assessment and early warning of the operation status of a power system under extreme weather conditions according to claim 1 is characterized in that: Constructing a power system fault probability model according to the load forecast result specifically includes: Based on the load forecast results, using the formula Determine the probability of failure depending on meteorological factors; In the formula, f n,W is the failure probability of the nth line in the power system when the extreme weather intensity is X, n = 1, 2, ..., N L , N L is the number of lines; n,X , n,X are the mean and standard deviation of the failure probability of the ith line under extreme weather conditions; Based on the load forecast results, using the formula Determine the failure probability of the road flow dependence; In the formula, f n,0 is the failure probability of line n during normal operation; P n,r and P n,max are the rated power and maximum allowable power of line n respectively; P n,s is the power flow size of line n during the sth round of cascading fault propagation; A power system fault probability model is constructed based on the fault probability dependent on meteorological factors and the fault probability dependent on power flow.

6. The method for risk assessment and early warning of the operation status of a power system under extreme weather conditions according to claim 5 is characterized in that: Based on the power system fault probability model, the initial fault operation scenario results of the power system under extreme weather conditions are determined, specifically including: Based on the power system fault probability model, using formula f n =min{δ n,W f n,W +δ n,F f n,F ,1}Determine the comprehensive failure probability of the line; According to the comprehensive failure probability of the line, the non-sequential Monte Carlo sampling simulation method is used to determine the initial fault operation scenario results of the power system under extreme weather conditions.

7. The method for risk assessment and early warning of the operation status of a power system under extreme weather conditions according to claim 1 is characterized in that: Based on the Nk cascading fault propagation path, an operation evaluation scheme is constructed according to the initial fault operation scenario results of the power system under the extreme weather conditions, specifically including: Operation evaluation scheme in the cascading fault propagation stage: In the cascading fault propagation stage, the line power flow in the initial fault scenario is calculated based on the DC power flow model, and it is determined whether the power flow of the remaining lines exceeds the limit, and a first judgment result is obtained; If the first judgment result is that the power flow of the remaining lines exceeds the limit, disconnect the faulty line of the power system, update the power line topology and carry out cascading fault propagation; If the first judgment result is that the remaining lines do not have power flow exceeding the limit, the chain fault propagation ends; Operation evaluation scheme in the optimization dispatching stage: In the optimization dispatching stage, based on the operation status of the power system after the cascading fault propagation stage, an optimization dispatching plan is formulated according to the simulation dispatching center; The load loss amount and the engine output adjustment amount are determined according to the optimized scheduling plan.

8. A risk assessment and early warning system for the operation status of a power system under extreme weather conditions, characterized in that: include: The load forecasting module is used to input global meteorological data into the load forecasting framework and output the load forecasting results; An electric power system initial fault operation scenario result output module is used to determine the electric power system initial fault operation scenario result under extreme weather conditions based on the electric power system fault probability model; The high-risk warning module is used to determine the high-risk of the operation scenario under the current extreme weather according to the operation evaluation plan, and to issue a warning based on the high-risk of the operation scenario under the extreme weather.

9. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the power system load forecasting and fault propagation risk warning method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer program instructions for causing a processor to execute the method according to any one of claims 1 to 7 are stored.

Images