A method for evaluating active power complementary characteristics of wind and solar energy considering extreme fluctuations

Abstract

The application discloses a wind-solar-load active complementary characteristic evaluation method considering extreme fluctuation, comprising the following steps: obtaining wind power active output sequence, photovoltaic active output sequence and load active sequence of a target area in a preset evaluation period; constructing total wind-solar output sequence and net load sequence; calculating net load climbing sequence, load climbing sequence and total wind-solar output climbing sequence; constructing evaluation indexes; establishing an analytic hierarchy model composed of a target layer, a criterion layer and an index layer, wherein the target layer is a wind power, photovoltaic and load active complementary characteristic comprehensive evaluation target; the index layer comprises the above evaluation indexes; wherein the weight of the evaluation indexes is obtained through conventional sample set and extreme sample set data. The application solves the technical problems of insufficient wind-solar-load unified evaluation, insufficient extreme scene risk description and weak comprehensive evaluation capability in the prior art, and improves the identification capability and early warning pertinence of the extreme operation risk of the power grid.

Description

Technical Field

[0001] This application relates to the field of power grid assessment technology, and in particular to an evaluation method for the active power complementarity characteristics of wind-solar-load considering extreme fluctuations. Background Technology

[0002] After a high proportion of new energy sources are connected to the grid, the system regulation pressure no longer comes only from the average power imbalance, but is more reflected in the rapid ramp-up of net load, continuous high fluctuations and extreme tail mismatch events. The known technologies for the study of renewable energy complementarity mainly revolve around time complementarity, spatial complementarity and spatiotemporal complementarity. Commonly used methods include correlation coefficient method, volatility index method, probability statistics method and multi-index comprehensive evaluation method. Reference [1] (Jurasz J, Canales FA, Kies A, et al. A review on the complementarity of renewable energy sources: Concept, metrics, application and future research directions. Solar Energy, 2020, 195:703-724.) systematically reviews the concept of complementarity, evaluation indicators and application scenarios, and points out that existing studies mostly focus on the complementary relationship between wind power and photovoltaic and other renewable energy sources, but a single index is often difficult to fully characterize the multidimensional characteristics of complementarity.

[0003] In the existing technical solutions similar to this invention, reference [2] (Wan Jiahao, Su Hao, Feng Donghan, et al. Analysis and evaluation of wind-solar complementary characteristics considering source-load matching. Power Grid Technology, 2020, 44(9): 3219-3226.) proposed a method for analyzing and evaluating wind-solar complementary characteristics considering source-load matching. The basic idea is: first, obtain wind power, photovoltaic and load data; then establish the correlation and tail correlation between the total output of wind and solar power and the load; further calculate the load tracking degree and other indicators; finally realize the comprehensive evaluation of wind-solar complementarity and source-load matching. This method has expanded the research focus from simple wind-solar complementarity to source-load matching, and is one of the existing methods that is relatively close to this invention. Patent reference [3] CN111401756B proposed a wind-solar absorption evaluation method involving grid, source and load analysis. The main steps include: establishing a two-level evaluation index system, obtaining grid and wind-solar load data and calculating various indicators, normalizing the indicators, and finally outputting the wind-solar absorption evaluation results. This scheme can reflect the interaction between the grid, source, and load, but its focus is on the comprehensive evaluation of wind and solar power integration levels, rather than specifically characterizing the active power complementarity between wind power, photovoltaics, and load. Patent document CN114997753B proposes a method and system for assessing the complementary power generation potential considering wind and solar resource attributes. Its main steps include: collecting wind and solar resource information within a designated area, evaluating the effectiveness of wind and solar resources, combining wind and solar resources in different proportions, and if the combination stability is better than that of a single resource, then determining the optimal complementarity ratio and maximum power generation potential based on stability. This type of method mainly focuses on wind and solar resource allocation and power generation potential assessment, with less consideration for load-side characteristics.

[0004] In addition, the literature [5] (Al-Rasheedi M, Al-Khayat M. Assessing wind and solar energy complementarity using novel metrics based on residual load profiles. Energy, 2025, 335: 138107.) proposed a wind-solar complementarity evaluation method based on residual load characteristics, which incorporates the combined output of wind and solar power with the system load into the analysis, and focuses on examining the impact of wind and solar combination on indicators such as residual load ramping power, ramping cycle, full-load hours of conventional units, capacity reliability, peak load and base load. This type of method is closer to system operation than the traditional smoothness evaluation, but it still mainly characterizes complementarity from the perspective of residual load results.

[0005] In summary, most existing wind-solar complementarity or source-load matching evaluation methods primarily analyze local relationships between wind power and solar power, or between renewable energy and load. They lack technical means to evaluate the active power complementarity characteristics of wind power, solar power, and load as a unified whole, making it difficult to comprehensively reflect the actual characteristics of source-load coupled operation in high-proportion renewable energy power systems. Furthermore, existing evaluation methods often focus on average time-series correlation, fluctuation smoothing effects, or statistical correlation, failing to adequately characterize mismatch risks under extreme operating scenarios such as extreme low wind, low solar, high load, and significant ramp-up conditions. This makes it difficult to accurately characterize the impact of wind-solar-load combinations on system regulation pressure and operational safety under extreme fluctuation conditions. In addition, existing technologies typically perform time complementarity analysis, probability distribution matching analysis, and extreme risk analysis separately, lacking an evaluation method that can comprehensively characterize the time-series matching relationship, probability distribution matching relationship, and extreme fluctuation risk characteristics between wind power, solar power, and load within a unified framework. This makes it difficult to provide more comprehensive and accurate technical support for renewable energy absorption analysis, power system operation assessment, and dispatch decisions. Summary of the Invention

[0006] This application provides a method for evaluating the active power complementarity characteristics of wind-solar-load considering extreme fluctuations, in order to solve the technical problems of insufficient unified evaluation of wind-solar-load, inadequate characterization of risks in extreme scenarios, and weak comprehensive evaluation capabilities in the existing technology, thereby improving the ability to identify extreme operation risks of the power grid and the pertinence of early warning.

[0007] The above-mentioned objective of this application is achieved through the following technical solution: a method for evaluating the active power complementarity characteristics of wind and solar loads taking into account extreme fluctuations, comprising the following steps:

[0008] Step 1: Obtain the wind power active power output sequence of the target area within the preset evaluation period. Photovoltaic active power output sequence and load active power sequence ;

[0009] Step 2: Construct the total wind and solar power output sequence based on the wind power active power output sequence, the photovoltaic active power output sequence, and the load active power output sequence. and net load sequence ;

[0010] Step 3: Based on the total wind and solar power output sequence and net load sequence, calculate the net load ramp-up sequence, load ramp-up sequence, and total wind and solar power output ramp-up sequence;

[0011] Step 4: Based on the wind power active power output sequence, photovoltaic active power output sequence, load active power output sequence, total wind and solar power output sequence, net load sequence and their corresponding ramp-up sequences, construct evaluation indicators;

[0012] Step 5: Establish a hierarchical analysis model consisting of a target layer, a criterion layer, and an indicator layer. The target layer is a comprehensive evaluation target of the complementary characteristics of wind power, photovoltaic power, and load active power. The indicator layer includes the above-mentioned evaluation indicators. The weights of the evaluation indicators are obtained through data from a regular sample set and an extreme sample set. The extreme sample set includes evaluation indicator data of the power grid under extreme fluctuation conditions.

[0013] Step 6: Output the comprehensive evaluation target of wind power, photovoltaic power and load active power complementarity characteristics through the hierarchical analysis model.

[0014] Furthermore, step 1 also includes the steps of: performing unified time scale alignment, missing value repair, and outlier processing on the wind power active power output sequence, photovoltaic active power output sequence, and load active power sequence.

[0015] Furthermore, in step 1, the unified time scale alignment includes unifying the wind power active power output sequence, photovoltaic active power output sequence, and load active power sequence to any time resolution of 5 minutes, 15 minutes, 30 minutes, or 1 hour; the missing value repair adopts one or more combinations of linear interpolation, neighboring time period mean method, or sliding window mean method; the outlier handling adopts one or more combinations of 3σ criterion, box plot method, or quantile truncation method.

[0016] Furthermore, in step 2, the total power output sequence of wind and solar power... and net load sequence The expressions are as follows:

[0017] (1)

[0018] (2)

[0019] The total power output sequence of wind and solar power The net load sequence is used to reflect the total active power supply capacity of wind and solar power at the same time; This is used to reflect the regulation targets that the system still needs to handle after deducting wind and solar power output, which are handled by conventional units, energy storage, or demand response.

[0020] Furthermore, in step 3, the net load ramp-up sequence, the load ramp-up sequence, and the total wind and solar power output ramp-up sequence are respectively expressed as follows:

[0021] (3)

[0022] (4)

[0023] (5)

[0024] The net load ramp-up sequence is used to characterize the degree of change of the system regulation object between adjacent time points under the combined action of wind power, photovoltaic and load; the load ramp-up sequence is used to reflect the rate of change of load-side demand; and the total wind and solar power output ramp-up sequence is used to reflect the rate of fluctuation of total power output on the renewable energy side.

[0025] Furthermore, in step 4, the evaluation indicators include:

[0026] (1) Net load variation coefficient

[0027] The net load variation coefficient is used to characterize the overall fluctuation intensity of net load, and its expression is:

[0028] (6)

[0029] Represents the net load sequence standard deviation Represents the net load sequence The mean, To prevent positive numbers with a denominator of zero;

[0030] (2) Average net load ramp rate

[0031] The average net load ramp-up rate is used to characterize the average regulating pressure of the system during the evaluation period, and its expression is:

[0032] (7)

[0033] in, This represents the time interval between adjacent sampling moments, consistent with the time resolution. This represents the unit change in net load between adjacent time points. This represents the average ramp rate over the entire sampling period T.

[0034] (3) Maximum net load ramp rate

[0035] The maximum net load ramp rate is used to characterize the extreme ramp level of net load during the most unfavorable period within the evaluation cycle, and its expression is:

[0036] (8)

[0037] (4) Time complementarity coefficient

[0038] The time complementarity coefficient is used to characterize the degree of matching between the total output of wind and solar power and the load over time. Its expression is:

[0039] (9)

[0040] in, Total power output sequence for wind and solar power With load sequence The correlation coefficient;

[0041] (10)

[0042] (11)

[0043] (12)

[0044] (5) Probability distribution overlap and complementarity coefficient

[0045] The probability distribution overlap and complementarity coefficient is used to characterize the degree of matching between the total wind and solar power output and the load at the probability distribution level. Its expression is:

[0046] (13)

[0047] in, and Let represent the probability density functions of total wind and solar power output and load, respectively;

[0048] (6) Extreme climbing coordination index

[0049] The extreme ramp synergy index is used to characterize the frequency of extreme events where a rapid increase in load and a rapid decrease in total wind and solar power output occur simultaneously. Its expression is:

[0050] (15)

[0051] in, The upper quantile threshold for the load ramping sequence. The threshold value for the lower quantile of the total power output climbing sequence of wind and solar power is used. It is an indicator function;

[0052] (7) Tail mismatch probability index

[0053] The tail mismatch probability index is used to characterize the tail risk that occurs simultaneously with low wind and solar power output and high load demand. Its expression is:

[0054] (16)

[0055] in, For the total power output sequence of wind and solar power quantiles, For load sequence quantiles, The value range is 0.05 to 0.10;

[0056] (8) Multi-timescale complementary stability

[0057] Multi-timescale complementary stability is used to characterize the complementary consistency between total wind and solar power output and load at different time scales. Its expression is:

[0058] (17)

[0059] in, For time scale sets, Time scale The weights, and satisfying ; and Representing time scales respectively The moving average of the load sequence and the total wind and solar power output sequence;

[0060] (9) Probability-weighted extreme climbing pressure index

[0061] The probability-weighted extreme ramp pressure index is used to characterize the extreme net load ramp pressure after taking into account the probability of scenario occurrence. Its expression is:

[0062] (18)

[0063] in, The total number of scenes, For the scene The probability of occurrence, For the scene Next moment Net load ramp rate This represents the extreme climbing threshold.

[0064] Furthermore, in the hierarchical analysis model, the criterion layer includes four criterion dimensions: volatility intensity, extreme ramp, temporal complementarity, and probabilistic complementarity; the volatility intensity dimension corresponds to the net load variation coefficient. and average net load ramp rate Extreme ramp dimension corresponds to the maximum net load ramp rate. Extreme climbing synergy index Probability-weighted extreme climbing pressure index The time complementarity dimension corresponds to the time complementarity coefficient. Complementary stability across multiple time scales Complementary probabilities correspond to overlapping probability distributions and complementary coefficients. Tail mismatch probability index .

[0065] Furthermore, in the hierarchical analysis model, an adaptive combined weighting method for suppressing nonlinear redundancy in two scenarios oriented towards extreme fluctuations is used to determine the basic global weights of the evaluation indicators, including the following steps:

[0066] Based on the hierarchical structure model of target layer—criteria layer—indicator layer, the analytic hierarchy process (AHP) is used to determine the first... Subjective weights of each evaluation indicator ;

[0067] Based on the extreme fluctuation trigger coefficient or historical statistical threshold, the sample set is divided into a regular sample set. and extreme sample sets And calculate the objective weight of each evaluation index in normal and extreme scenarios respectively; let the first... Each evaluation metric in the scenario The standard deviation below is , No. The evaluation index and the first The nonlinear correlation coefficient between the evaluation indicators is ,in Then the first Each evaluation metric in the scenario The information content below is represented as:

[0068] (19)

[0069] No. Each evaluation metric in the scenario The objective weights are expressed as follows:

[0070] (20)

[0071] in, It is a non-linear correlation metric used to characterize the degree of information redundancy between different evaluation indicators;

[0072] Based on the extreme fluctuation trigger coefficient of the current evaluation object Construct scene switching coefficients:

[0073] (twenty one)

[0074] And based on this, the first Adaptive objective weights for each evaluation indicator:

[0075] (twenty two)

[0076] In the formula and These represent the objective weights corresponding to the normal sample set and the extreme sample set within scenario r, respectively; then the subjective weights are... With adaptive objective weights By combining, we obtain the first... The basic global weight of each evaluation indicator Its expression is:

[0077] (twenty three)

[0078] in, The subjective preference coefficient is used; the basic global weight takes into account the results of hierarchical analysis, the distribution characteristics of indicator data, and the differences between normal and extreme scenarios, and is used for the subsequent calculation of the comprehensive evaluation value of normal scenarios and the comprehensive evaluation value of extreme scenarios.

[0079] Furthermore, in the analytic hierarchy process (AHP) model, the evaluation indicators are subjected to positive transformation and dimensionless transformation:

[0080] For positive indicators where a larger value indicates better complementarity, the following standardization method is adopted:

[0081] (twenty four)

[0082] For inverse indicators where smaller values ​​indicate better complementarity, the following standardization method is adopted:

[0083] (25)

[0084] in, and They represent the first The maximum and minimum values ​​of each evaluation indicator in the sample set. For the standardized first Each evaluation indicator value;

[0085] The standardized evaluation index values ​​are weighted according to the global weights to obtain the comprehensive evaluation value, which is expressed as follows:

[0086] (26)

[0087] in, The larger the value, the better the active power complementarity between wind power, photovoltaic power, and load.

[0088] Furthermore, the end fluctuation triggering coefficient is based on the extreme slope coordination index. Probability-weighted extreme climbing pressure index Construct, represented as:

[0089] (27)

[0090] in, The threshold for the extreme climbing coordination index. The threshold for the probability-weighted extreme climbing pressure index;

[0091] when At that time, for extremely correlated indicators , , and Basic global weights Perform two-layer adaptive correction;

[0092] Let the first The threshold for each extreme correlation indicator is ,in Then the first The individual deviation coefficient of each extreme correlation indicator is defined as:

[0093] (28)

[0094] No. The adaptive amplification factor corresponding to each extreme correlation index is expressed as follows:

[0095] (29)

[0096] in, For the first The base amplification factor of each extreme correlation indicator For the first Deviation response coefficient of each extreme correlation indicator;

[0097] The intermediate weight adjustment value for extremely correlated indicators is expressed as:

[0098] (30)

[0099] Then, the intermediate weights Normalization is performed to obtain the corrected global weights. Its expression is:

[0100] (31)

[0101] And based on the corrected global weights Calculate the comprehensive evaluation value of extreme scenarios Its expression is:

[0102] (32)

[0103] in, For the first The standardized index values ​​of the evaluation indicators after positiveization and dimensionless processing.

[0104] In summary, the beneficial effects of this application are as follows:

[0105] 1. This invention takes wind power, photovoltaics, and load as a unified research object. By constructing the total output sequence of wind and solar power and the net load sequence, it comprehensively describes the active power complementarity relationship between the source and the load. This overcomes the problem that existing technologies often focus on the pairwise relationship analysis of wind power and photovoltaics, wind power and load, or photovoltaics and load, and are difficult to fully reflect the actual power grid operation characteristics. This makes the evaluation results closer to the real operation scenario of power systems with high penetration of new energy.

[0106] 2. This invention constructs a nine-index evaluation system from four dimensions: fluctuation intensity, extreme ramp, temporal complementarity, and probabilistic complementarity. It not only uses the temporal complementarity coefficient and the probability distribution overlap complementarity coefficient to characterize the matching relationship between the total wind and solar power output and the load at the temporal variation and probability distribution levels, but also further characterizes the tail risk, large ramp risk, and multi-scale stability characteristics under extreme scenarios through the extreme ramp synergy index, tail mismatch probability index, multi-timescale complementary stability, and probability-weighted extreme ramp pressure index. This improves the comprehensiveness, hierarchy, and sensitivity to extreme risks of the evaluation results.

[0107] 3. This invention constructs a dual-scenario adaptive combined weighting method for extreme fluctuations. By dividing the samples into normal scenarios and extreme scenarios, the objective weights of the evaluation indicators under different operating states are determined respectively, and the basic global weights are formed by combining the subjective weights. This overcomes the problem that traditional static weighting methods cannot reflect the dynamic adjustment of the importance of indicators as the operating state changes, thereby improving the rationality, adaptability and objectivity of the weight allocation.

[0108] 4. In the objective weighting process, this invention introduces a mechanism to suppress the redundancy of the index dispersion and nonlinear correlation between indices. This not only reflects the ability of each evaluation index to distinguish sample differences, but also weakens the impact of information duplication between indices on the comprehensive evaluation results, thereby improving the comprehensive evaluation model's ability to extract independent and effective information and its ability to represent complex nonlinear source-load relationships.

[0109] 5. This invention constructs an extreme fluctuation trigger coefficient using an extreme slope synergy index and a probability-weighted extreme slope pressure index, and performs a two-layer adaptive correction on the basic global weights of extreme-related indicators after triggering: on the one hand, the weights of extreme-related indicators are uniformly amplified according to the overall degree of extreme fluctuation; on the other hand, the weights of each extreme-related indicator are differentiated and enhanced according to the degree of deviation of each extreme-related indicator from its respective threshold. This allows the evaluation system to remain stable in normal scenarios and to highlight the importance of extreme events such as large slopes, tail mismatch, and source-load synergy in extreme scenarios, thereby improving the model's ability to identify extreme operational risks and the pertinence of early warnings. Attached Figure Description

[0110] Figure 1 This is a flowchart of the framework of the wind-solar-load active power complementarity characteristic evaluation method considering extreme fluctuations in this invention.

[0111] Figure 2 This is a schematic diagram of the hierarchical analysis model structure of the present invention;

[0112] Figure 3 The schematic diagram of the extreme fluctuation triggering and weight correction process of this invention is shown in the figure;

[0113] Figure 4 This is a sequence diagram of wind and solar load output in a scenario of Embodiment 1 of the present invention. Detailed Implementation

[0114] The specific embodiments of this application are described in detail below with reference to the accompanying drawings.

[0115] Example: An evaluation method for the active power complementarity characteristics of wind and solar loads considering extreme fluctuations, referenced Figure 1 This includes the following steps:

[0116] Step 1: Obtain the wind power active power output sequence of the target area within the preset evaluation period. Photovoltaic active power output sequence and load active power sequence Furthermore, the wind power active power output sequence, photovoltaic active power output sequence, and load active power sequence are aligned to a unified time scale, missing value repaired, and outlier processed.

[0117] The unified time scale alignment includes standardizing the active power output series of wind power, photovoltaic active power output, and load active power to any time resolution of 5 minutes, 15 minutes, 30 minutes, or 1 hour, to ensure comparability of data from different sources at the same time scale. Missing value imputation employs at least one of the following methods: linear interpolation, neighboring time period mean method, or sliding window mean method. Outlier handling employs at least one of the following methods: 3σ criterion, box plot method, or quantile truncation method.

[0118] in, This represents the discrete sampling time within the evaluation period. This indicates the total number of sampling points.

[0119] Step 2: Based on the preprocessed wind power active power output sequence, photovoltaic active power output sequence, and load active power sequence, construct the total wind and solar power output sequence. and net load sequence Their expressions are as follows:

[0120] (1)

[0121] (2)

[0122] The total power output sequence of wind and solar power The net load sequence is used to reflect the total active power supply capacity of wind and solar power at the same time; This is used to reflect the regulation targets that still need to be handled by conventional units, energy storage, or demand response after deducting wind and solar power output. By constructing a net load sequence, the actual operating pressure of the system under scenarios with a high proportion of renewable energy can be more accurately characterized.

[0123] Step 3: Based on the total wind and solar power output sequence and net load sequence, further calculate the net load ramp-up sequence, load ramp-up sequence, and total wind and solar power output ramp-up sequence, which are expressed as follows:

[0124] (3)

[0125] (4)

[0126] (5)

[0127] The net load ramp-up sequence is used to characterize the degree of change of the system's regulating objects between adjacent time points under the combined effect of wind power, photovoltaic power, and load; the load ramp-up sequence is used to reflect the rate of change in load-side demand; and the total wind and solar power output ramp-up sequence is used to reflect the rate of fluctuation of total power output on the renewable energy side. By simultaneously characterizing the source-side ramp-up and the load-side ramp-up, a foundation can be provided for subsequent identification of extreme fluctuation scenarios where "rapid load increase and rapid wind and solar power decrease are superimposed".

[0128] Step 4: Based on the wind power active power output sequence, photovoltaic active power output sequence, load active power sequence, total wind and solar power output sequence, net load sequence and their corresponding ramp sequences, construct and calculate nine evaluation indicators.

[0129] The nine evaluation indicators include five basic indicators and four enhanced indicators. The five basic indicators are used to characterize the overall fluctuation of net load, average ramp, maximum ramp, temporal complementarity, and probability distribution matching characteristics; the four enhanced indicators are used to characterize extreme ramp synergy, tail mismatch, multi-timescale complementary stability, and extreme ramp pressure taking into account scenario probabilities.

[0130] (1) Net load variation coefficient

[0131] The net load variation coefficient is used to characterize the overall fluctuation intensity of net load, and its expression is:

[0132] (6)

[0133] Represents the net load sequence standard deviation Represents the net load sequence The mean, To prevent positive numbers with a denominator of zero.

[0134] (2) Average net load ramp rate

[0135] The average net load ramp-up rate is used to characterize the average regulating pressure of the system during the evaluation period, and its expression is:

[0136] (7)

[0137] in, This represents the time interval between adjacent sampling moments, consistent with the time resolution. This represents the unit change in net load between adjacent time points. This represents the average ramp rate over the entire sampling period T.

[0138] (3) Maximum net load ramp rate

[0139] The maximum net load ramp rate is used to characterize the extreme ramp level of net load during the most unfavorable period within the evaluation cycle, and its expression is:

[0140] (8)

[0141] (4) Time complementarity coefficient

[0142] The time complementarity coefficient is used to characterize the degree of matching between the total output of wind and solar power and the load over time. Its expression is:

[0143] (9)

[0144] in, Total power output sequence for wind and solar power With load sequence The correlation coefficient;

[0145] (10)

[0146] (11)

[0147] (12)

[0148] When the total output of wind and solar power and the load are highly synchronized in time, the time complementarity coefficient takes a larger value; when the time change trends of the two are inconsistent, the time complementarity coefficient takes a smaller value.

[0149] (5) Probability distribution overlap and complementarity coefficient

[0150] The probability distribution overlap and complementarity coefficient is used to characterize the degree of matching between the total wind and solar power output and the load at the probability distribution level. Its expression is:

[0151] (13)

[0152] in, and Let represent the probability density functions of total wind and solar power output and load, respectively.

[0153] The probability distribution overlap and complementarity coefficients can also be expressed in discrete form as follows:

[0154] (14)

[0155] in, This represents the total number of discrete intervals. and They represent the first The probability of total wind and solar power output and load within a discrete interval.

[0156] (6) Extreme climbing coordination index

[0157] The extreme ramp synergy index is used to characterize the frequency of extreme events where a rapid increase in load and a rapid decrease in total wind and solar power output occur simultaneously. Its expression is:

[0158] (15)

[0159] in, The upper quantile threshold for the load ramping sequence. The threshold value for the lower quantile of the total power output climbing sequence of wind and solar power is used. This is an indicator function.

[0160] The more frequently "rapid load increase" and "rapid wind and solar power decrease" occur together, the higher the risk of source-load mismatch in the system under extreme fluctuation scenarios.

[0161] (7) Tail mismatch probability index

[0162] The tail mismatch probability index is used to characterize the tail risk that occurs simultaneously with low wind and solar power output and high load demand. Its expression is:

[0163] (16)

[0164] in, For the total power output sequence of wind and solar power quantiles, For load sequence quantiles, The value range is 0.05 to 0.10.

[0165] The higher the tail mismatch probability index, the higher the risk of insufficient total wind and solar power output overlapping with peak load demand in low-probability, high-impact scenarios.

[0166] (8) Multi-timescale complementary stability

[0167] Multi-timescale complementary stability is used to characterize the complementary consistency between total wind and solar power output and load at different time scales. Its expression is:

[0168] (17)

[0169] in, For time scale sets, Time scale The weights, and satisfying ; and Representing time scales respectively The moving average of the load sequence and the total wind and solar power output sequence.

[0170] The greater the multi-timescale complementary stability, the better the complementary consistency between wind and solar power and load is in both short-term and long-term timescales.

[0171] (9) Probability-weighted extreme climbing pressure index

[0172] The probability-weighted extreme ramp pressure index is used to characterize the extreme net load ramp pressure after taking into account the probability of scenario occurrence. Its expression is:

[0173] (18)

[0174] in, The total number of scenes, For the scene The probability of occurrence, For the scene Next moment Net load ramp rate This represents the extreme climbing threshold.

[0175] By applying probability weighting to the net load ramp portion exceeding the extreme ramp threshold under different scenarios, the overall extreme ramp risk level faced by the system under random fluctuation conditions can be more comprehensively reflected.

[0176] Step 5: Establish a hierarchical analysis model consisting of a target layer, a criterion layer, and an indicator layer.

[0177] refer to Figure 2 The target layer is a comprehensive evaluation target of the complementary characteristics of wind power, photovoltaic power, and load active power; the criterion layer includes four criterion dimensions: fluctuation intensity, extreme ramping, time complementarity, and probabilistic complementarity; the indicator layer includes the aforementioned nine evaluation indicators.

[0178] Where: the fluctuation intensity dimension corresponds to the net load variation coefficient. and average net load ramp rate Extreme ramp dimension corresponds to the maximum net load ramp rate. Extreme climbing synergy index Probability-weighted extreme climbing pressure index The time complementarity dimension corresponds to the time complementarity coefficient. Complementary stability across multiple time scales Complementary probabilities correspond to overlapping probability distributions and complementary coefficients. Tail mismatch probability index .

[0179] By setting up the above hierarchical structure, it is possible to achieve layer-by-layer decomposition and evaluation from the criterion layer to the indicator layer.

[0180] In the hierarchical analysis model, an adaptive combined weighting method for suppressing nonlinear redundancy in two scenarios oriented towards extreme fluctuations determines the basic global weights of nine evaluation indicators, specifically including:

[0181] First, based on the hierarchical structure model of target layer—criteria layer—indicator layer, the analytic hierarchy process (AHP) is used to determine the first... Subjective weights of each evaluation indicator ;

[0182] Secondly, based on the extreme fluctuation trigger coefficient or historical statistical threshold, the sample set is divided into a regular sample set. and extreme sample sets The sample set consists of index data collected from the power grid under multiple scenarios, and the objective weights of each evaluation index are calculated in both normal and extreme scenarios; let the first... Each evaluation metric in the scenario The standard deviation below is , No. The evaluation index and the first The nonlinear correlation coefficient between the evaluation indicators is ,in Then the first Each evaluation metric in the scenario The information content below is represented as:

[0183] (19)

[0184] No. Each evaluation metric in the scenario The objective weights are expressed as follows:

[0185] (20)

[0186] in, It is a non-linear correlation metric used to characterize the degree of information redundancy between different evaluation indicators;

[0187] Based on the extreme fluctuation trigger coefficient of the current evaluation object Construct scene switching coefficients:

[0188] (twenty one)

[0189] And based on this, the first Adaptive objective weights for each evaluation indicator:

[0190] (twenty two)

[0191] In the formula and These represent the objective weights corresponding to the normal sample set and the extreme sample set within scenario r, respectively; then the subjective weights are... With adaptive objective weights By combining, we obtain the first... The basic global weight of each evaluation indicator Its expression is:

[0192] (twenty three)

[0193] in, The subjective preference coefficient is used; the basic global weight takes into account the results of hierarchical analysis, the distribution characteristics of indicator data, and the differences between normal and extreme scenarios, and is used for the subsequent calculation of the comprehensive evaluation value of normal scenarios and the comprehensive evaluation value of extreme scenarios.

[0194] The nine evaluation indicators are positiveized and dimensionless to obtain standardized evaluation indicator values.

[0195] For positive indicators where a larger value indicates better complementarity, the following standardization method is adopted:

[0196] (twenty four) For inverse indicators where smaller values ​​indicate better complementarity, the following standardization method is adopted:

[0197] (25)

[0198] in, and They represent the first The maximum and minimum values ​​of each evaluation indicator in the sample set. For the standardized first Each evaluation indicator value.

[0199] The standardized evaluation index values ​​are weighted according to the global weights to obtain the comprehensive evaluation value, which is expressed as follows:

[0200] (26)

[0201] in, The larger the value, the better the active power complementarity between wind power, photovoltaics, and load, the higher the source-load matching degree of the system within the evaluation period, and the lower the regulation pressure and mismatch risk faced.

[0202] refer to Figure 3 An extreme fluctuation trigger coefficient is constructed. When the extreme fluctuation trigger coefficient is met, the global weights of extreme-related indicators are adaptively adjusted, and the comprehensive evaluation value of the extreme scenario is calculated based on the adjusted weights.

[0203] The extreme fluctuation triggering coefficient is based on the extreme climb coordination index. Probability-weighted extreme climbing pressure index Construct, represented as:

[0204] (27)

[0205] in, The threshold for the extreme climbing coordination index. The threshold for the probability-weighted extreme climbing pressure index.

[0206] when At that time, for extremely correlated indicators , , and Basic global weights A two-layer adaptive correction is performed; the first layer adaptive correction is used to characterize the uniform amplification effect of the overall extreme volatility on the weights, and the second layer adaptive correction is used to characterize the differential amplification effect of the deviation of each extreme correlation index from its respective threshold on the weights.

[0207] Let the first The threshold for each extreme correlation indicator is ,in Then the first The individual deviation coefficient of each extreme correlation indicator is defined as:

[0208] (28)

[0209] No. The adaptive amplification factor corresponding to each extreme correlation index is expressed as follows:

[0210] (29)

[0211] in, For the first The base amplification factor of each extreme correlation indicator For the first Deviation response coefficient of each extreme correlation indicator.

[0212] The intermediate weight adjustment value for extremely correlated indicators is expressed as:

[0213] (30)

[0214] Then, the intermediate weights Normalization is performed to obtain the corrected global weights. Its expression is:

[0215] (31)

[0216] And based on the corrected global weights Calculate the comprehensive evaluation value of extreme scenarios Its expression is:

[0217] (32)

[0218] in, For the first The standardized index values ​​of the evaluation indicators after positiveization and dimensionless processing; the two-layer adaptive correction method makes the weight of extreme correlation indicators increase as the overall extreme volatility increases, and when the deviation of a certain extreme correlation indicator from its threshold is higher, the weight of that indicator in the comprehensive evaluation is further increased, thereby enhancing the sensitivity of the evaluation results to the risks of large uphill climbs, tail mismatch risks and extreme events of source-load synergy.

[0219] In summary, when applying this method to comprehensively evaluate the active power complementarity of wind, solar, and load in a power grid, the active power output data of wind power, solar power, and load are first acquired. After time alignment, missing data repair, and anomaly handling, the total wind and solar power output, net load, and ramp-up sequence are constructed, and nine evaluation indicators are calculated. Subsequently, the comprehensive evaluation value for normal scenarios is determined through hierarchical analysis and combined weighting. Finally, the existence of extreme fluctuations is identified. If triggered, the weights of extreme-related indicators are adaptively adjusted, the comprehensive evaluation value for extreme scenarios is calculated, and the final result is output.

[0220] Taking a typical summer day in a certain region as an example, 24-hour operating data of wind power, photovoltaic power, and load were selected as samples. After time alignment, missing value imputation, and outlier processing, a total wind and solar power output sequence, a net load sequence, and a ramp-up sequence were constructed. Based on this, extreme environmental disturbances such as a sudden drop in afternoon photovoltaic power, a sudden increase in evening load, and a decrease in wind power were introduced to create extreme scenarios. The wind and solar power output sequence diagram is shown below. Figure 4 .

[0221] Nine evaluation indicators were calculated, including the probability distribution overlap and complementarity coefficients calculated using either kernel density estimation or histogram methods. An extreme slope identification method was used to calculate the extreme slope synergy index. The probability-weighted extreme climbing pressure index is calculated using a scenario probability-weighted method. Then, the basic weights for typical scenarios are determined and the comprehensive evaluation value is calculated; when and When the preset threshold is exceeded, the weight of the extreme relevant indicators is adaptively adjusted to obtain the comprehensive evaluation result of the extreme scenario.

[0222] To demonstrate the effectiveness of the method of this invention, a traditional method based solely on temporal complementarity coefficients and probability distribution overlap complementarity coefficients was selected as a comparison. Further comparisons are shown in the table below:

[0223] Traditional methods 0.661 0.745 +12.7% Method of the present invention 0.670 0.290 -56.7%

[0224] The comparison results show that traditional methods, which only consider overall correlation and the degree of overlap in probability distributions, have limited reduction in comprehensive evaluation values ​​under extreme scenarios, making it difficult to fully reveal the operational risks when high load, low renewable energy, and large net load ramping occur simultaneously. In contrast, the method of this invention introduces an extreme ramping synergy index, a tail mismatch probability index, multi-timescale complementary stability, and a probability-weighted extreme ramping pressure index, and adaptively adjusts the weights of relevant indicators after detecting extreme fluctuations. Therefore, it can more sensitively reflect the degree of deterioration of the wind-solar-load complementarity relationship under the influence of extreme environmental factors.

[0225] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several modifications and improvements can be made without departing from the inventive concept of this application, and these all fall within the protection scope of this application.

Claims

1. A method for evaluating the active power complementarity characteristics of wind-solar load considering extreme fluctuations, characterized in that, Includes the following steps: Step 1: Obtain the wind power active power output sequence of the target area within the preset evaluation period. Photovoltaic active power output sequence and load active power sequence ; Step 2: Construct the total wind and solar power output sequence based on the wind power active power output sequence, the photovoltaic active power output sequence, and the load active power output sequence. and net load sequence ; Step 3: Based on the total wind and solar power output sequence and net load sequence, calculate the net load ramp-up sequence, load ramp-up sequence, and total wind and solar power output ramp-up sequence; Step 4: Based on the wind power active power output sequence, photovoltaic active power output sequence, load active power output sequence, total wind and solar power output sequence, net load sequence and their corresponding ramp-up sequences, construct evaluation indicators; Step 5: Establish a hierarchical analysis model consisting of a target layer, a criterion layer, and an indicator layer. The target layer is a comprehensive evaluation target of the complementary characteristics of wind power, photovoltaic power, and load active power. The indicator layer includes the above-mentioned evaluation indicators. The weights of the evaluation indicators are obtained through data from a regular sample set and an extreme sample set. The extreme sample set includes evaluation indicator data of the power grid under extreme fluctuation conditions. Step 6: Output the comprehensive evaluation target of wind power, photovoltaic power and load active power complementarity characteristics through the hierarchical analysis model.

2. The method for evaluating the active power complementarity characteristics of wind-solar load considering extreme fluctuations as described in claim 1, characterized in that, Step 1 also includes the following steps: performing unified time scale alignment, missing value repair, and outlier processing on the wind power active power output sequence, photovoltaic active power output sequence, and load active power output sequence.

3. The method for evaluating the active power complementarity characteristics of wind and solar loads considering extreme fluctuations according to claim 1, characterized in that, In step 1, the unified time scale alignment includes unifying the wind power active power output sequence, photovoltaic active power output sequence, and load active power sequence to any time resolution of 5 minutes, 15 minutes, 30 minutes, or 1 hour; the missing value repair adopts one or more combinations of linear interpolation, neighboring time period mean method, or sliding window mean method; the outlier handling adopts one or more combinations of 3σ criterion, box plot method, or quantile truncation method.

4. The method for evaluating the active power complementarity characteristics of wind and solar loads considering extreme fluctuations according to claim 1, characterized in that, In step 2, the total power output sequence of wind and solar power. and net load sequence The expressions are as follows: （1） （2） The total power output sequence of wind and solar power The net load sequence is used to reflect the total active power supply capacity of wind and solar power at the same time; This is used to reflect the regulation targets that the system still needs to handle after deducting wind and solar power output, which are handled by conventional units, energy storage, or demand response.

5. The method for evaluating the active power complementarity characteristics of wind and solar loads considering extreme fluctuations according to claim 4, characterized in that, In step 3, the net load ramp-up sequence, the load ramp-up sequence, and the total wind and solar power output ramp-up sequence are respectively represented as follows: （3） （4） （5） The net load ramp-up sequence is used to characterize the degree of change of the system regulation object between adjacent time points under the combined action of wind power, photovoltaic and load; the load ramp-up sequence is used to reflect the rate of change of load-side demand; and the total wind and solar power output ramp-up sequence is used to reflect the rate of fluctuation of total power output on the new energy side.

6. The method for evaluating the active power complementarity characteristics of wind and solar loads considering extreme fluctuations according to claim 5, characterized in that, In step 4, the evaluation indicators include: (1) Net load variation coefficient The net load variation coefficient is used to characterize the overall fluctuation intensity of net load, and its expression is: （6） Represents the net load sequence standard deviation Represents the net load sequence The mean, To prevent positive numbers with a denominator of zero; (2) Average net load ramp rate The average net load ramp-up rate is used to characterize the average regulating pressure of the system during the evaluation period, and its expression is: （7） in, This represents the time interval between adjacent sampling moments, consistent with the time resolution. This represents the unit change in net load between adjacent time points. This represents the average ramp rate over the entire sampling period T. (3) Maximum net load ramp rate The maximum net load ramp rate is used to characterize the extreme ramp level of net load during the most unfavorable period within the evaluation cycle, and its expression is: （8） (4) Time complementarity coefficient The time complementarity coefficient is used to characterize the degree of matching between the total output of wind and solar power and the load over time. Its expression is: （9） in, Total power output sequence for wind and solar power With load sequence The correlation coefficient; （10） （11） （12） (5) Probability distribution overlap and complementarity coefficient The probability distribution overlap and complementarity coefficient is used to characterize the degree of matching between the total wind and solar power output and the load at the probability distribution level. Its expression is: （13） in, and Let represent the probability density functions of total wind and solar power output and load, respectively; (6) Extreme climbing coordination index The extreme ramp synergy index is used to characterize the frequency of extreme events where a rapid increase in load and a rapid decrease in total wind and solar power output occur simultaneously. Its expression is: （15） in, The upper quantile threshold for the load ramping sequence. The threshold value for the lower quantile of the total power output climbing sequence of wind and solar power is used. It is an indicator function; (7) Tail mismatch probability index The tail mismatch probability index is used to characterize the tail risk that occurs simultaneously with low wind and solar power output and high load demand. Its expression is: （16） in, For the total power output sequence of wind and solar power quantiles, For load sequence quantiles, The value range is 0.05 to 0.10; (8) Multi-timescale complementary stability Multi-timescale complementary stability is used to characterize the complementary consistency between total wind and solar power output and load at different time scales. Its expression is: （17） in, For time scale sets, Time scale The weights, and satisfying ; and Representing time scales respectively The moving average of the load sequence and the total wind and solar power output sequence; (9) Probability-weighted extreme climbing pressure index The probability-weighted extreme ramp pressure index is used to characterize the extreme net load ramp pressure after taking into account the probability of scenario occurrence. Its expression is: （18） in, The total number of scenes, For the scene The probability of occurrence, For the scene Next moment Net load ramp rate This represents the extreme climbing threshold.

7. The method for evaluating the active power complementarity characteristics of wind and solar loads considering extreme fluctuations according to claim 6, characterized in that, In the hierarchical analysis model, the criterion layer includes four criterion dimensions: volatility intensity, extreme ramp, temporal complementarity, and probabilistic complementarity; the volatility intensity dimension corresponds to the net load variation coefficient. and average net load ramp rate Extreme ramp dimension corresponds to the maximum net load ramp rate. Extreme climbing synergy index Probability-weighted extreme climbing pressure index The time complementarity dimension corresponds to the time complementarity coefficient. Complementary stability across multiple time scales Complementary probabilities correspond to overlapping probability distributions and complementary coefficients. Tail mismatch probability index .

8. The method for evaluating the active power complementarity characteristics of wind and solar loads taking into account extreme fluctuations as described in claim 7, characterized in that, In the hierarchical analysis model, the basic global weights of the evaluation indicators are determined using a dual-scenario nonlinear redundancy suppression adaptive combination weighting method oriented towards extreme fluctuations. This includes the following steps: Based on the hierarchical structure model of target layer—criteria layer—indicator layer, the analytic hierarchy process (AHP) is used to determine the first... Subjective weights of each evaluation indicator ; Based on the extreme fluctuation trigger coefficient or historical statistical threshold, the sample set is divided into a regular sample set. and extreme sample sets And calculate the objective weight of each evaluation index in normal and extreme scenarios respectively; let the first... Each evaluation metric in the scenario The standard deviation below is , No. The evaluation index and the first The nonlinear correlation coefficient between the evaluation indicators is ,in Then the first Each evaluation metric in the scenario The information content below is represented as: （19） No. Each evaluation metric in the scenario The objective weights are expressed as follows: （20） in, It is a non-linear correlation metric used to characterize the degree of information redundancy between different evaluation indicators; Based on the extreme fluctuation trigger coefficient of the current evaluation object Construct scene switching coefficients: （21） And based on this, the first Adaptive objective weights for each evaluation indicator: （22） In the formula and These represent the objective weights corresponding to the normal sample set and the extreme sample set within scenario r, respectively; then the subjective weights are... With adaptive objective weights By combining, we obtain the first... The basic global weight of each evaluation indicator Its expression is: （23） in, The subjective preference coefficient is used; the basic global weight takes into account the results of hierarchical analysis, the distribution characteristics of indicator data, and the differences between normal and extreme scenarios, and is used for the subsequent calculation of the comprehensive evaluation value of normal scenarios and the comprehensive evaluation value of extreme scenarios.

9. The method for evaluating the active power complementarity characteristics of wind and solar loads considering extreme fluctuations according to claim 8, characterized in that, In the analytic hierarchy process (AHP) model, the evaluation indicators are subjected to positive transformation and dimensionless transformation: For positive indicators where a larger value indicates better complementarity, the following standardization method is adopted: （24） For inverse indicators where smaller values ​​indicate better complementarity, the following standardization method is adopted: （25） in, and They represent the first The maximum and minimum values ​​of each evaluation indicator in the sample set. For the standardized first Each evaluation indicator value; The standardized evaluation index values ​​are weighted according to the global weights to obtain the comprehensive evaluation value, which is expressed as follows: （26） in, The larger the value, the better the active power complementarity between wind power, photovoltaic power, and load.

10. The method for evaluating the active power complementarity characteristics of wind and solar loads considering extreme fluctuations according to claim 9, characterized in that, The end fluctuation triggering coefficient is based on the extreme slope coordination index. Probability-weighted extreme climbing pressure index Construct, represented as: （27） in, The threshold for the extreme climbing coordination index. The threshold for the probability-weighted extreme climbing pressure index; when At that time, for extremely correlated indicators , , and Basic global weights Perform two-layer adaptive correction; Let the first The threshold for each extreme correlation indicator is ,in Then the first The individual deviation coefficient of each extreme correlation indicator is defined as: （28） No. The adaptive amplification factor corresponding to each extreme correlation index is expressed as follows: （29） in, For the first The base amplification factor of each extreme correlation indicator For the first Deviation response coefficient of each extreme correlation indicator; The intermediate weight adjustment value for extremely correlated indicators is expressed as: （30） Then, the intermediate weights Normalization is performed to obtain the corrected global weights. Its expression is: （31） And based on the corrected global weights Calculate the comprehensive evaluation value of extreme scenarios Its expression is: （32） in, For the first The standardized index values ​​of the evaluation indicators after positiveization and dimensionless processing.

Images