A method for evaluating the risk of source-load imbalance under extreme weather

Abstract

This invention relates to the fields of power system operation control and new energy grid power prediction, and discloses a method for assessing the risk of source-load imbalance under extreme weather conditions. The method employs the MCNN-Mamba wind power prediction model to extract multi-scale meteorological disturbance characteristics and wind power mutation time-series characteristics under extreme weather scenarios, obtaining wind power output prediction results for the target time period. Based on the predicted wind power sequence, a cross-sectional net loss measurement is constructed by combining the load power sequence, available photovoltaic output sequence, conventional unit baseline output sequence, and interconnection and exchange plan sequence. Furthermore, an available control resource pool is formed by combining the response boundaries of energy storage, photovoltaic, conventional units, interruptible loads, demand response resources, and pumped storage or flexible DC support resources, thus completing the assessment of multi-resource coordinated control capabilities under extreme weather impacts. The strategy proposed in this invention can provide support for the proactive emergency control resource pre-positioning and coordinated regulation of the power grid under extreme weather conditions in the face of short-term high-risk common-cause failures.

Description

Technical Field

[0001] This invention relates to the field of power system operation control and new energy grid power prediction technology, specifically to a method for assessing the risk of source-load imbalance under extreme weather conditions. Background Technology

[0002] With the continuous advancement of new power system construction, the proportion of new energy sources, represented by wind power and photovoltaics, in the power system is constantly increasing, leading to greater uncertainty and volatility in power supply structure and load characteristics. Simultaneously, extreme weather events such as typhoons, heavy rains, blizzards, cold waves, heat waves, and severe convection are frequent, easily affecting new energy output, load demand, and the operational status of multiple regulatory resources simultaneously. This can induce extreme source-load imbalances in the research section within a short period, forming typical short-term, high-risk common-cause failure scenarios. Therefore, conducting source-load imbalance scenario generation and multi-resource control capability assessments in response to extreme weather impacts is of great significance for improving the proactive defense level and proactive emergency control preparedness capabilities of new power systems.

[0003] Existing research on power system operation analysis under extreme weather conditions mostly focuses on single-variable prediction, selection of typical scenarios, or static reserve analysis. There is still a lack of integrated methods to address issues such as the scarcity of extreme weather samples, drastic fluctuations in wind power output, and dynamic changes in the coordinated regulation capacity of multiple resources. Under the impact of extreme weather, the limited number of historical observation samples means that directly relying on raw samples for modeling can easily lead to insufficient model learning of extreme conditions, making it difficult to effectively characterize the sudden changes in power output and the temporal evolution patterns during extreme weather processes.

[0004] Meanwhile, existing wind power prediction methods mostly employ recurrent neural networks, convolutional neural networks, or Transformer structures. Although these methods can achieve certain results under normal meteorological conditions, under extreme meteorological conditions, the input features are highly dimensional, the time-series dependencies are long, and local abrupt changes are significant. Traditional methods still have shortcomings in multi-scale time-series feature extraction and long-series dynamic modeling. In addition, existing studies on the multi-resource regulation capacity analysis mostly remain at the level of static superposition of resource capacity, lacking dynamic capacity boundary assessment based on the predicted source-load imbalance process, making it difficult to accurately reflect the system's actual available resources' ability to coordinate compensation for cross-sectional imbalances under extreme meteorological impacts.

[0005] To this end, this invention proposes a source-load imbalance risk assessment method under extreme weather conditions, in order to more accurately characterize extreme scenarios, improve the accuracy of wind power prediction, and achieve quantitative assessment of the boundary of multi-resource collaborative active emergency control capabilities. Summary of the Invention

[0006] The purpose of this invention is to address the difficulties in constructing source-load imbalance scenarios under extreme weather conditions, the insufficient accuracy of wind power output fluctuation prediction, and the challenge of uniformly assessing the multi-resource regulation capabilities. This invention proposes a method for assessing source-load imbalance risks under extreme weather conditions. Addressing the issues of sudden changes in renewable energy power and increased system regulation pressure caused by extreme weather, this method establishes a unified technical route from extreme scenario generation and wind power prediction to multi-resource control capability assessment. This aims to improve the ability to identify source-load imbalance risks at the research section and the analytical level of multi-resource collaborative support capabilities, and to provide a foundation for the subsequent formulation of proactive emergency control strategies.

[0007] To achieve the above objectives, this invention provides a method for assessing the risk of source-load imbalance under extreme weather conditions, comprising the following steps: (1) Obtain historical operational data, meteorological observation data and numerical weather prediction data of the study area under extreme weather impact, and construct a multi-source spatiotemporal feature sample set; establish the MPG-Diffusion extreme scene generation model based on the multi-source spatiotemporal feature sample set to generate extreme weather scenes, and construct a source-load imbalance scene set for short-term high-risk common cause failure based on this. (2) Based on the multi-source spatiotemporal feature sample set and source-load imbalance scenario set, establish the MCNN-Mamba wind power prediction model to predict wind power output under extreme weather impact and obtain the predicted wind power sequence in the target period. (3) Based on the predicted wind power sequence, load power sequence, photovoltaic available output sequence, conventional unit benchmark output sequence and interconnection exchange plan sequence, construct the cross-sectional net loss measurement, combine the response boundaries of multiple types of adjustable resources to form an available control resource pool, and output the multi-resource control capability assessment results for extreme weather impacts.

[0008] Preferably, in step (1), the multi-source input sample set is constructed using a joint modeling method of meteorological features and operational features. The meteorological features are used to characterize the intensity and trend of external disturbances during extreme weather processes, while the operational features are used to characterize the internal evolution of wind power output and load response, thereby ensuring that the subsequent extreme scenario generation and wind power prediction processes have both meteorological driving information and operational status information.

[0009] Preferably, in step (1), the MPG-Diffusion extreme scene generation model includes a forward diffusion process and a reverse denoising process; wherein, the forward diffusion process gradually injects Gaussian noise into the original sample to establish a perturbation path that transforms the real sample distribution into a noise distribution, the expression of which is: in, and These represent the sample states at step k-1 and step k, respectively. For noise dispatch coefficient, It is an identity matrix.

[0010] Preferably, in step (1), the reverse denoising process is used to gradually recover extreme scene features from the noise samples, and its expression is: in, To reconstruct the sample, For noise estimation results, The parameters are accumulated during the diffusion process. The training loss function of the scene generation model is: in, For noise reconstruction loss, For trend consistency loss, For physical constraint loss, , and These are the weighting coefficients.

[0011] Preferably, in step (2), the MCNN-Mamba wind power prediction model includes a multi-scale convolutional feature extraction module and a Mamba time series modeling module; the multi-scale convolutional feature extraction module extracts the local disturbance features of the wind power sequence under extreme weather conditions in parallel using convolutional kernels of different scales, and its expression is: in, Given the input sequence, The output of the i-th convolutional branch is... As a feature of fusion, and These are the convolution weights and biases, respectively.

[0012] Preferably, the Mamba time series modeling module uses discrete state-space form to describe the dynamic evolution of the input sequence in the time dimension, and its expression is: in, For input features, It is in a hidden state. For output features, , and These are the state matrix, input matrix, and output matrix, respectively. The distance parameter is the distance from the step.

[0013] Preferably, in step (3), a cross-sectional net loss metric is constructed based on the power balance relationship between predicted wind power, load power, available photovoltaic output, conventional unit baseline output, and interconnection and switching plan. Its expression is: in, The net loss of the cross section at time t is used as a measure. For load power, For communication and switching program power, To predict wind power output, For photovoltaic power output, This is the benchmark output for conventional generating units.

[0014] Preferably, in step (3), the available controllable resource pool consists of energy storage resources, photovoltaic resources, conventional units, interruptible loads, demand response resources, pumped storage support resources, and flexible DC support resources; based on the adjustable power limit, sustainable adjustment time, ramp-up capability, and operating status of each type of resource within the target time period, its available adjustment capability is calculated, and the total available adjustment capability is obtained by summing them up, the expression of which is: in, Let be the available power of the i-th type of resource at time t. For total available adjustment capacity, The real-time availability coefficient of resources is determined based on the operational status, fault blocking status, state of charge or energy margin, ramp margin and cross-sectional transmission reachability of the i-th type of resource at time t, and the value range is from 0 to 1. This represents the total number of resource categories.

[0015] Preferably, in step (3), the multi-resource control capacity margin and coverage rate are constructed based on the total available adjustment capacity and the net loss of cross-section, and their expressions are as follows: Based on the changes in control capacity margin and coverage in different time periods, the multi-resource control capacity assessment results for extreme weather impacts are output.

[0016] Through the above-described scheme, this invention addresses the problems of increased uncertainty in renewable energy output under extreme weather conditions, the inability of traditional scenarios to reflect high-impact processes, and the difficulty in uniformly representing different regulatory resources. It constructs a complete technical chain from extreme scenario generation to wind power prediction, and then to multi-resource control capability assessment. First, by introducing the MPG-Diffusion extreme scenario generation model, under the condition of limited extreme samples, it can effectively learn the coupling distribution characteristics between meteorological and operational variables, generating representative extreme scenario samples that meet physical constraints, thereby improving the completeness and rationality of source-load imbalance scenario construction. Second, by constructing the MCNN-Mamba wind power prediction model, it comprehensively utilizes multi-scale local feature extraction capabilities and long-sequence dynamic modeling capabilities, enabling more accurate characterization of the drastic fluctuations in wind power output under extreme meteorological impacts, improving wind power prediction accuracy and model adaptability. Finally, by constructing cross-sectional net loss measurement, available control resource pool, control capability margin, and coverage indicators, it achieves collaborative assessment of multiple types of regulatory resources under a unified cross-sectional power balance framework, which can more intuitively reflect the resource support capability for cross-sectional source-load imbalance at different time scales. Detailed Implementation

[0017] The following provides a detailed description of specific embodiments of the present invention. It should be understood that the specific embodiments described herein are for illustrative and explanatory purposes only and are not intended to limit the scope of the invention.

[0018] Example The method for assessing the risk of source-load imbalance under extreme weather conditions as described in this invention includes: (1) Obtain historical operational data, meteorological observation data and numerical weather prediction data of the study area under extreme weather impact, and construct a multi-source spatiotemporal feature sample set; establish the MPG-Diffusion extreme scene generation model based on the multi-source spatiotemporal feature sample set to generate extreme weather scenes, and construct a source-load imbalance scene set for short-term high-risk common cause failure based on this. (2) Based on the multi-source spatiotemporal feature sample set and source-load imbalance scenario set, establish the MCNN-Mamba wind power prediction model to predict wind power output under extreme weather impact and obtain the predicted wind power sequence in the target period. (3) Based on the predicted wind power sequence, load power sequence, photovoltaic available output sequence, conventional unit benchmark output sequence and interconnection exchange plan sequence, construct the cross-sectional net loss measurement, combine the response boundaries of multiple types of adjustable resources to form an available control resource pool, and output the multi-resource control capability assessment results for extreme weather impacts.

[0019] In step (1), the historical operating data includes at least wind power output time series, photovoltaic power output time series, load power time series, conventional unit benchmark output series, energy storage status series, and flexible load adjustable capacity series; the meteorological data includes at least temperature, wind speed, wind direction, humidity, air pressure, precipitation, irradiance, and other meteorological variables reflecting the characteristics of extreme weather processes. The historical operating data and meteorological data are time-aligned according to a unified sampling period to ensure temporal consistency between data from different sources. For missing samples, duplicate records, and abnormal fluctuations that may exist in the original data, interpolation, adjacent time period compensation, or sliding window mean substitution methods are used to repair missing values, and threshold discrimination or statistical outlier identification methods are combined to remove outliers; finally, the processed input features are normalized to reduce the impact of different dimensions on model training. The normalization process can be expressed as: in, The original variable value, and These are the minimum and maximum values ​​of the variable in the sample set, respectively. These are the normalized variable values.

[0020] In step (1), to enhance the expressive power of extreme weather process characteristics, meteorological variables and operational quantities are jointly modeled to construct the input sample matrix: in, Let be the joint input vector at time t. Characteristics of meteorological variables, Characteristics of wind power output Characteristics of photovoltaic power output For load power characteristics, This represents the baseline output characteristic of a conventional unit. As energy storage state characteristics, This represents the characteristics of flexible loads or demand response resources. By constructing the aforementioned multi-source input sample set, a unified data foundation is provided for subsequent extreme scenario generation and wind power prediction.

[0021] In step (1), based on the obtained multi-source spatiotemporal feature sample set, an MPG-Diffusion extreme weather scenario generation model is established to learn the joint distribution characteristics among meteorological variables, wind power output, photovoltaic power output, load power, and other resource state variables under extreme weather impacts; wherein, MPG-Diffusion is a physically guided multi-granularity diffusion scenario generation model. The model includes a forward diffusion process and a reverse denoising process. The forward diffusion process gradually introduces Gaussian noise into the real samples, so that the samples gradually transition from the real distribution to the known noise distribution, and its expression is: Furthermore, from the forward diffusion process, the expression of the sample state at any given time relative to the initial sample can be obtained as follows: in, and Let represent the sample states at step k-1 and step k, respectively. Represents the noise scheduling coefficient at step k. This is the cumulative noise attenuation coefficient. Standard Gaussian noise, It is an identity matrix.

[0022] In step (1), the inverse denoising process is used to recover scene samples that conform to the characteristics of extreme weather impacts from the noise samples, and its reconstruction expression is: in, The samples obtained from model reconstruction This is the noise estimation result output by the parameterized denoising network.

[0023] To improve the reasonableness of the generated scenes in terms of trend evolution and operational boundaries, the following training loss function is constructed: in, For noise reconstruction loss term, To maintain the loss term for the trend, For physical constraint loss terms, , and These are the weighting coefficients.

[0024] Furthermore, the multi-granularity trend preservation loss and physical constraint loss are respectively expressed as: in, For multi-grained decomposition layers, For the l-th layer time downsampling operator, For the corresponding scale weights, Let be the generated wind power component at time t. The physical power limit is defined as the upper bound of wind power; the physical constraint loss penalizes both cases where the upper bound of wind power exceeds the limit and cases where the wind power is less than zero, to ensure that the generated samples satisfy the basic physical boundary.

[0025] The upper bound of the physical power of wind power can be expressed as: in, Let t be the wind speed at the hub height. and These are the cut-in wind speed and the cut-out wind speed, respectively. The wind energy utilization coefficient, air density, For the area swept by the wind turbine, This refers to the rated installed capacity.

[0026] Multiple extreme weather scenario samples were generated using the trained MPG-Diffusion model, and wind power output, photovoltaic power output, load power, and interconnection exchange feature sequences were extracted according to a given time period. Based on the changes in source-side output and load-side demand under different extreme scenarios, a source-load imbalance scenario set was formed to characterize various short-term high-risk common-cause failure states that may occur at the research section during extreme weather impacts.

[0027] In step (2), based on the multi-source spatiotemporal feature sample set and source-load imbalance scenario set obtained in step (1), an MCNN-Mamba wind power prediction model is established; wherein, MCNN-Mamba is a prediction model formed by coupling a multi-scale convolutional neural network and a Mamba state-space temporal modeling network. The model includes a multi-scale convolutional feature extraction module and a Mamba temporal modeling module, which are used to extract local disturbance information and characterize long-term temporal dependencies, respectively. The multi-scale convolutional feature extraction module extracts local features in wind power output and meteorological variables in parallel through multiple convolutional branches of different scales, and its expression is: in, Given the input sequence, Output features for the i-th convolutional branch. For the fused feature representation, and These are the kernel parameters and the bias term, respectively. For batch normalization operators, This is a residual mapping.

[0028] In step (2), the Mamba time series modeling module uses a discrete state space to model the dynamic features of the input sequence over a long time span. The state update process is as follows: in, Let be the input features at time t. These are hidden state variables. For output features, , and These are the state matrix, input matrix, and output matrix, respectively. The distance parameter is denoted by _____. The above model can output the predicted wind power sequence for the target time period. .

[0029] In step (3), based on the predicted wind power sequence Load power sequence for the corresponding time period Available photovoltaic power output sequence Standard unit benchmark output sequence and contact exchange program sequence The net imbalance level of the cross-section is calculated at each time point; wherein, the studied cross-section is a target interconnection section or key transmission section in the power grid used to monitor the power supply and receiving balance and the risk of power flow exceeding limits. The formula for calculating the net imbalance level of the cross-section is: in, This formula represents the net power deficit at time t between load demand, interconnection exchange pressure, and available output on the source side. Its engineering significance lies in transforming the impact of changes in renewable energy output, load growth, and cross-sectional exchange demand under extreme weather conditions on the supply-demand balance into a quantifiable power deficit indicator, providing a unified benchmark for adjusting resource allocation.

[0030] In step (3), for energy storage, photovoltaic, conventional units, interruptible loads, demand response resources, pumped storage support resources, and flexible DC support resources, the available power at each time point is calculated based on their rated capacity, availability status, ramping constraints, state of charge, interruptible upper limit, and cross-sectional transmission margin, and the total available regulation capacity is summarized. The available power of the i-th type of resource at time t and the total available regulation capacity of multiple resources can be expressed as: in, Let be the available power of the i-th type of resource at time t. To provide overall resource availability adjustment capability, It is used to characterize the real-time availability status of resources, and is determined comprehensively based on the operational status, fault blocking status, state of charge or energy margin, ramp margin and cross-sectional transmission reachability of the i-th type of resource at time t, with a value range of 0 to 1. Let be the maximum up-adjustment power of the i-th type of resource at time t. To increase the climbing speed, For the scheduling period, It is an energy state quantity. This is the lower limit of energy. This represents the total number of resource categories. The purpose of this formula is to uniformly convert different types of resources into the same power dimension of support capacity, in order to construct a pool of available control resources for proactive emergency control.

[0031] In step (3), the control margin and coverage are further calculated, and their expressions are as follows: in, To control capability margin, To control capacity coverage, To avoid positive constants with a denominator of zero. When ≥0 and When ≥1, it is determined that multiple resources have the ability to cover the load imbalance gap at time t; when <0 or When the value is less than 1, it is determined that multiple resources have insufficient control capabilities at time t. By statistically and comprehensively analyzing the continuous changes in control capability margin and coverage in each time period, the time-series assessment results of the control capability of multiple resources during extreme weather impacts, the weak control periods, and the corresponding resource support shortcomings can be obtained.

[0032] The method described in this embodiment achieves closed-loop analysis from external meteorological shock input to internal resource support capability output through a sequential processing flow of "extreme scenario generation—new energy grid power prediction—multi-resource control capability assessment." Specifically, the extreme scenario generation stage addresses the problem of insufficient extreme samples, the prediction stage improves the characterization of wind power output changes, and the assessment stage enables quantitative determination of the level of multi-resource collaborative support. This method can be applied to the operational analysis, pre-deployment of regulatory resources, and assessment of proactive emergency control capabilities in areas with high new energy penetration under extreme weather conditions.

[0033] The above are merely preferred embodiments of the present invention; however, the present invention is not limited thereto. Within the scope of the inventive concept, various simple modifications can be made to the technical solutions of the present invention, including combining the various technical features in any other suitable manner. Such modifications or substitutions should all fall within the protection scope of the present invention.

Claims

1. A method for assessing the risk of source-load imbalance under extreme weather conditions, characterized in that, The method includes: (1) Based on historical operational data, meteorological observation data and numerical weather prediction data in the study area, a multi-source spatiotemporal feature sample set is constructed. The MPG-Diffusion scene generation model is used to learn the coupling evolution law of meteorological variables and source load state variables under extreme meteorological processes. The MPG-Diffusion is a physical-guided multi-granularity diffusion scene generation model. Under the combined effect of noise reconstruction loss, multi-granularity temporal trend consistency loss and physical boundary constraint loss, an extreme meteorological scene set that meets the spatiotemporal propagation characteristics and physical operational boundaries is generated. Based on this, a source load imbalance scene set for short-term high-risk common cause failure is constructed. (2) Based on the multi-source spatiotemporal feature sample set and the source-load imbalance scenario set, an MCNN-Mamba wind power prediction model is constructed, wherein the MCNN-Mamba is a prediction model formed by coupling a multi-scale convolutional neural network with a Mamba state-space temporal modeling network; local mutation features are extracted by multi-scale convolutional branches, and long-term dynamic dependencies are characterized by Mamba modules to predict wind power output under extreme weather impacts, thereby obtaining the predicted wind power sequence within the target time period; (3) Based on the predicted wind power sequence, load power sequence, available photovoltaic output sequence, conventional unit benchmark output sequence and interconnection and exchange plan sequence, construct the cross-sectional net loss measurement, and combine the response boundaries of energy storage resources, photovoltaic regulation resources, conventional units, interruptible loads, demand response resources, pumped storage support resources and flexible DC support resources to form an available control resource pool, evaluate the collaborative support capability of multiple resources for cross-sectional source-load imbalance under extreme weather impact, and output the multi-resource control capability assessment results.

2. The method for assessing the risk of source-load imbalance under extreme weather conditions according to claim 1, characterized in that, The MPG-Diffusion extreme weather scene generation model described in step (1) includes a forward diffusion process and a reverse denoising generation process; wherein, the forward diffusion process obtains the k-th perturbation sample by gradually adding Gaussian noise to the real sample, and its expression is: in, and Let represent the sample states at steps k-1 and k, respectively. Represents the noise scheduling coefficient at step k. It is an identity matrix.

3. The method for assessing the risk of source-load imbalance under extreme weather conditions according to claim 2, characterized in that, In step (1), the sample state at any step size corresponding to the forward diffusion process satisfies: The reverse denoising process uses a denoising network to predict noise terms and recover the estimated value of the clean sample. Its expression is as follows: in, This is the cumulative noise attenuation coefficient. Standard Gaussian noise, This is the noise estimate output by the denoising network at step k.

4. The method for assessing the risk of source-load imbalance under extreme weather conditions according to claim 3, characterized in that, In step (1), to ensure that the generated scenario simultaneously satisfies the temporal evolution law of extreme weather and the physical constraints of power operation, a composite loss function is constructed for the MPG-Diffusion scenario generation model, the expression of which is: in, For noise reconstruction loss, For multi-granularity time-series trend consistency loss, For physical boundary constraint loss, , and These are the weighting coefficients.

5. The method for assessing the risk of source-load imbalance under extreme weather conditions according to claim 4, characterized in that, The multi-granularity temporal trend consistency loss and physical boundary constraint loss are respectively expressed as: in, For multi-grained decomposition layers, For the l-th layer time downsampling operator, These are the weighting coefficients for the corresponding scale. For time t, generate the wind power component in the sample. This represents the upper limit of the physical power of wind power.

6. The method for assessing the risk of source-load imbalance under extreme weather conditions according to claim 5, characterized in that, The upper limit of wind power physical power The expression is: in, Let t be the wind speed at the hub height. To cut into wind speed, To cut off the wind speed, The wind energy utilization coefficient, air density, For the area swept by the wind turbine, This refers to the rated installed capacity.

7. The method for assessing the risk of source-load imbalance under extreme weather conditions according to claim 1, characterized in that, In step (2), the MCNN-Mamba prediction model includes a multi-scale convolutional feature extraction module and a Mamba temporal modeling module; wherein, the multi-scale convolutional feature extraction module performs parallel convolution on the input sequence X to obtain the i-th branch output, and the branch outputs are concatenated and convolved pointwise to obtain the fused feature, the expression of which is: in, and Let be the weights and biases of the i-th convolutional branch, respectively. For batch normalization operators, This is a residual mapping.

8. The method for assessing the risk of source-load imbalance under extreme weather conditions according to claim 7, characterized in that, The Mamba time series modeling module adopts an input-dependent state-space modeling approach, and its discretization parameters and state update process satisfy the following: in, Let be the input features at time t. These are hidden state variables. For output features, , and These are the state matrix, input matrix, and output matrix, respectively. The distance parameter is the distance from the step.

9. The method for assessing the risk of source-load imbalance under extreme weather conditions according to claim 1, characterized in that, In step (3), the cross-sectional net loss measurement is constructed as follows: in, The net loss of the cross section at time t is used as a measure. For load power, For communication and switching program power, To predict wind power output, For photovoltaic power output, This is the benchmark output for conventional generating units; The available power of resource type i at time t and the total available adjustability of multiple resources are expressed as follows: in, This represents the real-time availability coefficient of resources. Let be the maximum up-adjustment power of the i-th type of resource at time t. To increase the climbing speed, For the scheduling period, It is an energy state quantity. This is the lower limit of energy. This represents the total number of resource categories.

10. The method for assessing the risk of source-load imbalance under extreme weather conditions according to claim 9, characterized in that, In step (3), the multi-resource control capacity margin and coverage rate are constructed based on the cross-sectional net loss measurement and total available adjustment capacity, and their expressions are as follows: in, The control margin at time t, Let be the control capability coverage at time t. To avoid positive constants with a denominator of zero; when ≥0 and When ≥1, it is determined that multiple resources have the ability to cover the net imbalance gap at time t; when <0 or When <1, it is determined that the control capability of multiple resources is insufficient at time t; based on the temporal distribution of control capability margin and coverage in each time period, the assessment results of the control capability of multiple resources under extreme weather impact are output.

Images