A hierarchical evaluation method for evaluating active support control capability of a wind farm

Abstract

This invention discloses a graded evaluation method for the active support control capability of wind farms. The method includes determining primary indicators of the active support control capability of wind farms, constructing secondary indicators based on the primary indicators, determining the weights of the secondary indicators and obtaining the corresponding evaluation values ​​for each primary indicator, constructing a three-dimensional improved radar chart, and using the characteristics of the three-dimensional improved radar chart to introduce an evaluation function to calculate the evaluation value of the active support control capability of the wind farm, thus achieving a graded evaluation of the active support control capability of the wind farm. This invention can more comprehensively evaluate the active support capability of wind farms with better reliability, and can better facilitate grid dispatching and wind power companies' understanding of the performance of various active support control technologies.

Description

Technical Field

[0001] This invention relates to the field of control and operation and maintenance of grid-connected wind farms, and in particular to a graded evaluation method for the active support control capability of wind farms. Background Technology

[0002] Energy and environmental issues are increasingly attracting global attention, and low-carbon energy transition is one of the important means. With the continuous and rapid development of new energy sources in my country, the installed capacity of wind and solar power is constantly increasing, and some regions have formed a high-proportion new energy grid connection pattern. China currently has the largest cumulative installed capacity of wind power. However, wind power generation is characterized by randomness, volatility, and intermittency, and wind power grid-connected equipment also suffers from low immunity and weak support. The increasing proportion of wind power grid connection has brought great challenges to the operation and control of the power system. Therefore, in addition to continuously improving its traditional grid-connection performance such as power quality, grid adaptability, and grid protection, wind power should continuously improve its grid-friendly active support technology to provide a certain frequency and voltage support capability to ensure system frequency and voltage stability.

[0003] Meanwhile, current evaluation indicators for the active support and control capabilities of wind farms generally have certain limitations. Publicly available literature shows that some studies have proposed flexible evaluation index systems for wind farm integration into the power system that consider wind energy instability, and have studied the impact of load uncertainty and wind power uncertainty on flexibility indicators. Other studies have analyzed that large-scale wind turbine disconnection is due to the wind turbines' inability to meet low-voltage ride-through requirements. Still other studies have used wind farm SCADA monitoring systems as data sources to analyze the data structure and characteristics of wind farm operating parameters. In summary, current evaluation indicators lack a comprehensive and systematic evaluation index system.

[0004] Furthermore, the factors involved in active support control of wind farms are constantly being optimized or increased with research and development. Currently, some factors in the evaluation system are not yet fully considered. For example, some current primary frequency regulation standards and requirements require wind farms to reserve a certain proportion of their rated capacity for rapid frequency regulation response based on the predicted active power output, thus establishing a correlation between wind power prediction accuracy and active support control. However, existing evaluation indicators or systems rarely consider the accuracy of wind power prediction.

[0005] Current methods for calculating the weights and evaluation values ​​of active support control capabilities, based on publicly available literature and patents, include simple empirical scoring methods, stratified analysis, entropy methods, and network analysis. However, relying on a single method, or simply multiplying the weights after they have been determined, suffers from problems such as a lack of multi-dimensional evaluation levels, subjective weight determination, and insufficient reference value for the evaluation values. Summary of the Invention

[0006] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0007] In view of the problems existing in the above and / or existing graded assessment of the active support control capability of wind farms, this invention is proposed.

[0008] Therefore, the problem to be solved by the present invention is how to provide a graded evaluation method for the active support control capability of wind farms.

[0009] To address the aforementioned technical problems, this invention provides the following technical solution: a graded evaluation method for the active support control capability of wind farms, comprising: determining primary indicators of the active support control capability of wind farms; constructing secondary indicators based on the primary indicators; determining the weights of the secondary indicators and obtaining the corresponding evaluation values ​​for each primary indicator; constructing a three-dimensional improved radar chart; introducing an evaluation function based on the characteristics of the three-dimensional improved radar chart to calculate the evaluation value of the active support control capability of wind farms; and realizing a graded evaluation of the active support control capability of wind farms.

[0010] As a preferred embodiment of the wind farm active support control capability evaluation method of the graded assessment described in this invention, the primary indicators include wind power prediction accuracy, rapid voltage regulation indicator, primary frequency regulation indicator, and inertia response indicator.

[0011] As a preferred embodiment of the wind farm active support control capability evaluation method for graded assessment described in this invention, the primary frequency regulation index includes the root mean square value of frequency error, the performance index of primary frequency regulation effect, and the response time compliance rate.

[0012] As a preferred embodiment of the wind farm active support control capability evaluation method for graded assessment described in this invention, the formula for calculating the root mean square value of the frequency error is:

[0013]

[0014] Among them, f RMS f is the root mean square value of the frequency error. i f represents the frequency sample value of the i-th point; N represents the total number of sampling points; standard This is the standard value for the power grid frequency;

[0015] The formula for calculating the performance index of primary frequency modulation is as follows:

[0016] P pfr =ΔQsY / ΔQjY

[0017] Among them, P pfr ΔQsY is the performance index of primary frequency regulation, ΔQjY is the integral energy of the actual operation of primary frequency regulation, and ΔQjY is the theoretically calculated integral energy for the corresponding time.

[0018] The formula for calculating the response time compliance rate is as follows:

[0019] λ standard =N good / N all *100%

[0020] Where, λ standard To achieve the response time target, N good N represents the number of times within the statistical time span that the frequency modulation power increase reaches 90% and the time taken to reach the target power increase is less than 9 seconds. all This indicates the total number of frequency modulations within the same time span.

[0021] As a preferred embodiment of the wind farm active support control capability evaluation method for graded assessment described in this invention, the rapid voltage regulation index includes the voltage qualification rate U. passRate Automatic voltage control system uptime λ AVC and maximum reactive power support capacity Inertial response metrics include frequency drop lower limit metrics, power grid frequency damping effect metrics, and power grid frequency secondary drop evaluation metrics.

[0022] As a preferred embodiment of the wind farm active support control capability evaluation method for graded assessment described in this invention, the wind power prediction accuracy rate adopts the ultra-short-term wind power accuracy rate, and its calculation formula is as follows:

[0023]

[0024] Where, λ CDQ For ultra-short-term power prediction accuracy; P N This represents the daily operating capacity; P pi P represents the actual power at time i; mi Let be the predicted value at time i, and n be the number of samples;

[0025] P mi The expression is:

[0026]

[0027] Among them, P mi Let be the predicted value at time i.

[0028] As a preferred embodiment of the wind farm active support control capability evaluation method for graded assessment described in this invention, the method for determining the weights of secondary indicators is as follows: based on historical data samples, the weights of secondary indicators are determined using a combination of information entropy and information gain, specifically including the following steps:

[0029] For the rapid pressure regulation index, the results are divided into four levels: A, B, C, and D.

[0030] Based on historical data, historical sample data S is calculated. Each set of historical data includes three characteristics: voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity, and corresponds to one of the result levels A, B, C, and D.

[0031] According to the information entropy formula:

[0032]

[0033] Where Ent(S) represents the information entropy of sample data S, p k It represents the proportion of samples of class k in the sample data S;

[0034] Calculate the information entropy of historical samples:

[0035] Ent(S)=-(p A log2p A +p B log2p B +p C log2p C +p D log2p D )

[0036] Where, p A p B p C p D These correspond to the proportions of samples at levels A, B, C, and D in the sample data S, respectively.

[0037] The voltage qualification rate is divided into three levels, C1, C2, and C3, based on its numerical characteristics.

[0038] The operational rate and maximum reactive power support capacity of the automatic voltage control system are also classified according to the characteristics of the period values;

[0039] According to voltage qualification rate U passRate The sample dataset S is divided into three subsets, denoted as S1, S2, and S3, based on the grades C1, C2, and C3.

[0040] Based on the information entropy formula, the information entropies Ent(S1), Ent(S2), and Ent(S3) of the three subsets are calculated.

[0041] Voltage qualification rate U passRate The information gain for sample S is:

[0042]

[0043] Among them, Gain(S,U) passRate ) represents the voltage qualification rate U passRate For the information gain of sample S, Ent(S) represents the information entropy of sample S. v This indicates that the v-th subset contains the number of samples in S that take the value corresponding to the level of that subset, where S represents the total number of samples in the sample data S. Ent(S) v ) represents the information entropy of the v-th subset;

[0044] Similarly, the information gain Gain(S, λ) of the automatic voltage control system's operational rate is obtained. AVC Information gain of maximum reactive power support capability.

[0045] According to the formula:

[0046]

[0047] Among them, w j Let a represent the j-th weight value. j Let Gain(S, a) represent the j-th feature of the sample set S. j ) represents the information gain of the feature in the sample dataset S;

[0048] The weights of the secondary indicators corresponding to the rapid voltage regulation index, namely voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity, are obtained as W1, W2, and W3.

[0049] Similarly, the weight values ​​of the primary frequency modulation index and the inertia response index are obtained using the above method.

[0050] As a preferred embodiment of the wind farm active support control capability evaluation method for graded assessment described in this invention, obtaining the evaluation values ​​of each primary indicator includes the following steps:

[0051] The evaluation value of the rapid voltage regulation index is calculated using the following formula:

[0052]

[0053] Among them, W1, W2, and W3 are the weights of the secondary indicators voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity, respectively. passRate ,λ AVC , The results are the calculated voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity.

[0054] Similarly, the evaluation values ​​of the primary frequency regulation index and the inertial response index are calculated.

[0055] As a preferred embodiment of the wind farm active support control capability evaluation method for graded assessment described in this invention, the method includes the following steps: Constructing a three-dimensional improved radar map.

[0056] The weight values ​​of the rapid voltage regulation index, primary frequency regulation index, and inertial response index are determined using the advantage plot method, and are used as the angles between the axes of each index in the radar chart.

[0057] Using the values ​​of each indicator as the radius of the sector, a sector is drawn to replace the original triangular radar chart, thus forming an improved radar chart.

[0058] Using the wind power prediction accuracy as the height of the radar chart reference plane, the third-dimensional parameter of the radar chart is formed, resulting in a three-dimensional improved radar chart.

[0059] The parameters in the 3D improved radar chart—angle, radius, and height—correspond to the weight, the evaluation value of the control-related primary index, and the accuracy of wind power prediction, respectively.

[0060] As a preferred embodiment of the wind farm active support control capability evaluation method for graded assessment described in this invention, the method involves: calculating the wind farm active support control capability evaluation value based on the characteristics of a three-dimensional improved radar chart using an evaluation function, including the following steps:

[0061] The evaluation features are constructed, and the calculation formula for feature one is as follows:

[0062]

[0063] Among them, V i D represents the volumetric accumulation of the cone in the plotted 3D improved radar image. iU D is the evaluation value of the fast voltage regulation index of the i-th object. if D is the evaluation value of the frequency modulation index of the i-th object. iJ λ is the inertia response index evaluation value of the i-th object. CDQ It is the accuracy of ultra-short-term power prediction. W1, W2, and W3 represent the weights of the fast voltage regulation index, primary frequency regulation index, and inertia response index obtained by the dominance graph method. C1 represents a constant.

[0064] The formula for calculating feature two is:

[0065] A i =2π(D) iU ×w1×C1+D if ×w2×λCDQ +D iJ ×w3×C1)

[0066] Among them, A i Accumulate added value for the conical surface in the generated 3D improved radar map;

[0067] The formula for constructing the evaluation function using the geometric mean method is as follows:

[0068] EV i =γ i1 *γ i2 i = 1, 2, 3, ..., m

[0069] Among them, EV i γ is the evaluation value for the active support and control capability of a wind farm. i1 Gamma represents the strength of active support capabilities. i2 Represents the degree of balance in abilities;

[0070] γ i1 =V i / V max V max =max{V i |1,2,3,...,m}

[0071] Among them, V i V represents the feature-calculation result of the i-th object out of m objects to be evaluated. max This represents the maximum value of the feature-1 calculation result among m objects to be evaluated;

[0072] γ i2 =A min / A i A min =min{A i |1,2,3,...,m}

[0073] Among them, A i Let Amin represent the result of the second feature calculation for the i-th object among m objects to be evaluated, and let Amin represent the minimum value of the second feature calculation result among the m objects to be evaluated.

[0074] The beneficial effects of this invention are as follows: This invention establishes a set of objective and scientific evaluation index systems with more complete elements, realizes the calculation of evaluation values ​​more suitable for wind farms, and finally determines the corresponding evaluation function to calculate the evaluation value of the active support control capability of wind farms. It obtains a set of graded quantitative index evaluation values ​​including comprehensive evaluation level, comprehensive evaluation value, support capability strength, balance degree, and sub-item first-level index evaluation values, realizing the graded evaluation of the active support control capability of wind farms. Compared with existing evaluation methods, this invention can more comprehensively evaluate the active support capability of wind farms, improve the credibility, and better promote the understanding of the performance of various active support control technologies by grid dispatching and wind power companies. Attached Figure Description

[0075] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0076] Figure 1 This is the overall flowchart of the wind farm active support control capability evaluation method in Example 1, which involves graded assessment.

[0077] Figure 2 This is a flowchart and step diagram of the wind farm active support control capability evaluation method for graded assessment in Example 1. Detailed Implementation

[0078] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0079] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0080] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0081] Example 1

[0082] Reference Figure 1 and Figure 2This is the first embodiment of the present invention, which provides a graded evaluation method for assessing the active support control capability of wind farms, including:

[0083] S1: The primary indicators include wind power prediction accuracy, rapid voltage regulation, primary frequency regulation, and inertia response.

[0084] S2: The accuracy of wind power prediction is determined by using the accuracy of ultra-short-term wind power prediction.

[0085] Furthermore, the formula for the accuracy of ultra-short-term power prediction is:

[0086]

[0087] Where, λ CDQ For ultra-short-term power prediction accuracy; P N This represents the daily operating capacity; P pi P represents the actual power at time i; mi Let be the predicted value at time i, and n be the number of samples.

[0088] P mi The expression is:

[0089]

[0090] Among them, P mi Let be the predicted value at time i.

[0091] S3: The secondary indicators for determining the rapid voltage regulation index include voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity; the secondary indicators for the primary frequency regulation index include the root mean square value of frequency error, primary frequency regulation performance index, and response time compliance rate; the secondary indicators for the inertia response index include the lower limit of frequency drop index, grid frequency damping effect index, and grid frequency secondary drop evaluation index.

[0092] Furthermore, the formula for calculating the voltage qualification rate is:

[0093]

[0094] Among them, U passRate For voltage qualification rate, U abnormal For voltage exceeding the power supply / consumption limit, U all This represents a percentage of the total electricity supplied / consumed.

[0095] λ AVC =T run / (T total -T exit )

[0096] Where, λ AVCT represents the monthly operational rate of the automatic voltage control system, where Trun is the cumulative operational time per month. total T is the cumulative operating time of the main power generation unit. exit This refers to the time when the automatic voltage control system shuts down due to power grid issues.

[0097] The formula for calculating the maximum reactive power support capacity is:

[0098]

[0099] in, Q represents the maximum reactive power support capacity. real Q represents the actual output reactive power when the voltage is at the steady edge. theory To reserve reactive power for theoretical purposes.

[0100] The formula for calculating the root mean square value of frequency error is:

[0101]

[0102] Among them, f RMS f is the root mean square value of the frequency error. i f represents the frequency sample value of the i-th point; N represents the total number of sampling points; standard The standard value for the power grid frequency is 50Hz, which is taken here.

[0103] The formula for calculating the performance index of primary frequency modulation is:

[0104] P pfr =ΔQsY / ΔQjY(if P pfr <0, P pfr =0)

[0105] Among them, P pfr ΔQsY represents the performance index of primary frequency regulation, where ΔQsY is the integral charge of the actual operation of primary frequency regulation, and ΔQjY is the integral charge of the theoretical calculation for the corresponding time.

[0106] The formula for calculating the response time compliance rate is:

[0107] λ standard =N good / N all *100%

[0108] Where, λ standard To achieve the response time target, N good This indicates the number of times within the statistical time span that the time taken for the frequency modulation power increase to reach 90% of the target power increase is less than 9 seconds; N all This indicates the total number of frequency modulations within the same time span.

[0109] The formula for calculating the lower limit of frequency drop is:

[0110]

[0111] Among them, S sys This represents the system's baseline capacity; r p M represents the power growth rate; sys The system's total inertia level; f B The reference frequency for the power grid is 50Hz, which is taken here; Δf DB The dead zone frequency; ΔP L =P m0 -P e0 That is, unbalanced power, P m0 For mechanical power, P e0 This refers to electromagnetic power.

[0112] Formula for calculating the power grid frequency damping effect index:

[0113]

[0114] Where, Δf max f is the maximum value of the change in grid frequency during the dynamic process; B The reference frequency for the power grid is 50Hz, which is selected here.

[0115] Formula for calculating the second frequency sag index of power grid:

[0116]

[0117] Among them, Δf2_ max f is the maximum deviation value of the second frequency drop in the power grid. B The reference frequency for the power grid is 50Hz, which is selected here.

[0118] S4: The weights of secondary indicators can be obtained by using historical data samples, information entropy, and information gain; the values ​​of secondary indicators can be obtained by using various calculation formulas.

[0119] Furthermore, by using historical data samples, information entropy, and information gain, the specific methods for determining the weights of secondary indicators and calculating the evaluation values ​​of primary indicators are as follows:

[0120] For the rapid voltage regulation index, the results are divided into four levels: A, B, C, and D.

[0121] Based on historical data, historical sample data S is calculated. Each set of historical data includes three characteristics: voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity, and corresponds to one of the result levels A, B, C, and D.

[0122] According to the information entropy formula:

[0123]

[0124] Where Ent(S) represents the information entropy of sample data S, p k It represents the proportion of the k-th class of samples in the sample data S.

[0125] Calculate the information entropy of historical samples:

[0126] Ent(S)=-(p A log2p A +p B log2p B +p C log2p C +p D log2p D )

[0127] Where, p A p B p C p D These correspond to the proportions of samples at levels A, B, C, and D in the sample data S, respectively.

[0128] The voltage qualification rate is divided into three levels, C1, C2, and C3, based on its numerical characteristics.

[0129] The operational rate and maximum reactive power support capacity of the automatic voltage control system are also classified according to the characteristics of the periodic values.

[0130] According to voltage qualification rate U passRate The sample dataset S is divided into three subsets, denoted as S1, S2, and S3, with C1, C2, and C3 as the levels.

[0131] Based on the information entropy formula, the information entropies Ent(S1), Ent(S2), and Ent(S3) of the three subsets are calculated.

[0132] Furthermore, the voltage qualification rate U passRate The information gain for sample S is:

[0133]

[0134] Among them, Gain(S,U) passRate ) represents the voltage qualification rate U passRate For the information gain of sample S, Ent(S) represents the information entropy of sample S. v This indicates that the v-th subset contains the number of samples in S that take the value corresponding to the level of that subset, where S represents the total number of samples in the sample data S. Ent(S) v ) represents the information entropy of the v-th subset.

[0135] Similarly, the information gain Gain(S, λ) of the automatic voltage control system's operational rate is obtained. AVC Information gain of maximum reactive power support capability.

[0136] According to the formula:

[0137]

[0138] Among them, w j Let a represent the j-th weight value. j Let Gain(S, a) represent the j-th feature of the sample set S. j ) represents the information gain of the feature in the sample dataset S.

[0139] The weights of the secondary indicators corresponding to the rapid voltage regulation index—voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity—are W1, W2, and W3, respectively.

[0140] Similarly, the weight values ​​of the primary frequency modulation index and the inertia response index are obtained using the above method.

[0141] The evaluation value D of the rapid voltage regulation index U It can be calculated that:

[0142]

[0143] Among them, W1, W2, and W3 are the weights of the secondary indicators voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity, respectively. passRate ,λ AVC , The results are the calculated voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity.

[0144] Similarly, the weight values ​​of the primary frequency modulation index and the inertial response index can be obtained, and the evaluation value D of the primary frequency modulation index can be calculated. f The evaluation value D of the inertial response index J .

[0145] S5: The evaluation values ​​of three primary indicators—rapid voltage regulation, primary frequency regulation, and inertia response—can be calculated from the secondary weights and secondary indicator values; the evaluation value of the ultra-short-term forecast accuracy, i.e., the primary indicator wind power forecast accuracy, can be calculated using the formula.

[0146] S6: Construct a three-dimensional improved radar chart based on the evaluation values ​​of the primary indicators. The parameters (angle, radius, and height) of the three-dimensional improved radar chart are mapped to (weight, evaluation values ​​of the control-related primary indicators, and wind power prediction accuracy).

[0147] Among them, the evaluation values ​​of the control-related primary indicators are the rapid voltage regulation indicator, the primary frequency regulation indicator, and the inertia response indicator.

[0148] Furthermore, the specific method for constructing a three-dimensional improved radar chart is as follows:

[0149] First, the weight values ​​of the rapid voltage regulation index, primary frequency regulation index, and inertial response index are determined using the advantage plot method, which serve as the angles between the axes of each index in the radar chart.

[0150] Then, using the values ​​of each indicator as the radius of the sector, a radar chart is drawn that replaces the original triangle with a sector, forming an improved radar chart.

[0151] Then, using the wind power prediction accuracy as the height of the radar chart reference plane, the third-dimensional parameter of the radar chart is formed, resulting in a three-dimensional improved radar chart.

[0152] Ultimately, the parameters (angle, radius, and height) in the improved 3D radar chart correspond to (weight, evaluation value of control-related primary index, and wind power prediction accuracy), respectively.

[0153] Among them, the evaluation values ​​of the control-related primary indicators are the rapid voltage regulation indicator, the primary frequency regulation indicator, and the inertia response indicator.

[0154] S7: Construct evaluation features based on three-dimensional improved radar charts.

[0155] Furthermore, the volumetric accumulation value of the cone in the 3D improved radar image is calculated to characterize the strength of its active support capability; the surface accumulation value of the cone in the 3D improved radar image is calculated to characterize the balance of its performance.

[0156] S8: The evaluation function is constructed using the geometric mean method. The evaluation value of the active support control capability of the wind farm is calculated through the evaluation function, and the corresponding evaluation levels of excellent, good, medium and poor are obtained.

[0157] Based on steps S7 and S8, the specific method for introducing evaluation features and constructing the evaluation function using the geometric mean method is as follows:

[0158] Construct evaluation features. Feature one is the volume accumulation bonus of the cone in the 3D improved radar image. The larger the volume accumulation bonus, the stronger the active support capability.

[0159] Feature 2 is the surface accumulation value of the cone in the three-dimensional improved radar image. With a fixed volume, the smaller the surface area, the closer the shape is to a cylinder, which indicates more average performance.

[0160] Specifically, suppose there are m objects to be evaluated. Then, for the i-th object:

[0161] The evaluation features are constructed, and the calculation formula for feature one is as follows:

[0162]

[0163] Among them, V i D represents the volumetric accumulation of the cone in the plotted 3D improved radar image. iU D is the evaluation value of the fast voltage regulation index of the i-th object. if D is the evaluation value of the frequency modulation index of the i-th object. iJ λ is the inertia response index evaluation value of the i-th object. CDQ It is the accuracy of ultra-short-term power prediction. W1, W2, and W3 represent the weights of the fast voltage regulation index, primary frequency regulation index, and inertia response index obtained by the dominance graph method. C1 represents a constant, usually 1.

[0164] The formula for calculating feature two is:

[0165] A i =2π(D) iU ×w1×C1+D if ×w2×λ CDQ +D iJ ×w3×C1)

[0166] Among them, A i Accumulate added value for the conical surface in the generated 3D improved radar map.

[0167] The formula for constructing the evaluation function using the geometric mean method is as follows:

[0168] EV i =γ i1 *γ i2 i = 1, 2, 3, ..., m

[0169] Among them, EV i γ is the evaluation value for the active support and control capability of a wind farm. i1 Gamma represents the strength of active support capabilities. i2 It represents the degree of balance in abilities.

[0170] γ i1 =V i / V max V max =max{V i |1,2,3,...,m}

[0171] Among them, V i V represents the feature-calculation result of the i-th object out of m objects to be evaluated. max This represents the maximum value of the feature-1 calculation result among the m objects to be evaluated.

[0172] γ i2 =A min / A i A min =min{A i |1,2,3,...,m}

[0173] Among them, A i Let Amin represent the result of the second feature calculation for the i-th object among m objects to be evaluated, and let Amin represent the minimum value of the second feature calculation result among the m objects to be evaluated.

[0174] In summary, this invention establishes a more objective, scientific, and comprehensive evaluation index system. The primary evaluation indexes include wind power prediction accuracy, dynamic voltage index, primary frequency regulation index, and inertia response index. The weights of secondary indicators are determined through historical data samples, information entropy, and information gain, enabling evaluation value calculations more suitable for wind farms. Evaluation features are introduced based on the characteristics of a three-dimensional improved radar chart, and the corresponding evaluation function is ultimately determined to calculate the evaluation value of the wind farm's active support control capability. A set of graded and quantitative evaluation indexes is obtained, including comprehensive evaluation level, comprehensive evaluation value, support capability strength, balance, and sub-items, achieving a graded evaluation of the wind farm's active support control capability. Compared with existing evaluation methods, this invention can more comprehensively assess the active support capability of wind farms, with better reliability, and can better facilitate grid dispatching and wind power companies' understanding of the performance of various active support control technologies.

[0175] Example 2

[0176] Referring to Table 1, which is the second embodiment of the present invention, and based on the first embodiment, the embodiment data description is provided to verify its beneficial effects.

[0177] The specific method for calculating the weights is as follows: for the rapid voltage regulation index, the results are divided into four levels: A, B, C, and D.

[0178] Based on historical data, historical sample data S is calculated. Each set of historical data includes three characteristics: voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity, and corresponds to one of the result levels A, B, C, and D.

[0179] According to the information entropy formula:

[0180]

[0181] Where Ent(S) represents the information entropy of sample data S, p k It represents the proportion of the k-th class of samples in the sample data S.

[0182] Calculate the information entropy of historical samples:

[0183] Ent(S)=-(p Alog2p A +p B log2p B +p C log2p C +p D log2p D )

[0184] Where, p A p B p C p D These correspond to the proportions of samples at levels A, B, C, and D in the sample data S, respectively.

[0185] The voltage qualification rate is divided into three levels, C1, C2, and C3, based on its numerical characteristics.

[0186] The commissioning rate and maximum reactive power support capacity of the automatic voltage control system are also classified according to the characteristics of the periodic values.

[0187] According to voltage qualification rate U passRate The levels C1, C2, and C3 can be used to divide the sample dataset S into three subsets, denoted as S1, S2, and S3, which correspond to the number of samples at levels C1, C2, and C3, respectively.

[0188] Furthermore, within the S1 subset, there are also samples of four levels: A, B, C, and D. The proportion of each level in the S1 subset is denoted as p. A,s1 p B,s1 p C,s1 p D,s1 .

[0189] Similarly, we can also obtain p in the S2 subset. A,s2 p B,s2 p C,s2 p D,s2 p in the S3 subset A,s3 p B,s3 p C,s3 p D,s 3 .

[0190] The information entropy of the three subsets was calculated as follows:

[0191] Ent(S1)=-(p A,s1 log2p A,s1 +p B,s1 log2p B,s1 +p C,s1 log2p C,s1 +p D,s1 log2p D,s1 )

[0192] Wherein, Ent(S1) is the information entropy of the subset S1.

[0193] Ent(S2)=-(p A,s2 log2p A,s2 +p B,s2 log2p B,s2 +p C,s2 log2p C,s2 +p D,s2 log2p D,s2 )

[0194] Wherein, Ent(S2) is the information entropy of the subset S2.

[0195] Ent(S3)=-(p A,s3 log2p A,s3 +p B,s3 log2p B,s3 +p C,s3 log2p C,s3 +p D,s3 log2p D,s3 )

[0196] Wherein, Ent(S3) is the information entropy of the S3 subset.

[0197] The voltage qualification rate U was obtained by further calculation. passRate The information gain for sample S is:

[0198]

[0199] Using the same method described above, calculate the information gain Gain(S, λ) of the automatic voltage control system's operational rate. AVC Information gain of maximum reactive power support capability.

[0200] Based on the three information gain values ​​obtained above, the weights of the secondary indicators—voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity—can be calculated:

[0201]

[0202] Among them, w1 is the weight of the secondary indicator voltage qualification rate.

[0203]

[0204] Where w2 is the weight of the automatic voltage control system's operational rate.

[0205]

[0206] Among them, w3 is the weight of the maximum reactive power support capacity.

[0207] Calculate the evaluation value D of the rapid voltage regulation index. U :

[0208]

[0209] Among them, D U This is the evaluation value for the rapid voltage regulation index.

[0210] Continuing with the above method for obtaining weights, we obtain the weight values ​​for the primary frequency modulation index and the inertia response index, and calculate the evaluation value D of the primary frequency modulation index. f The evaluation value D of the inertial response index J .

[0211] Table 1. Example diagrams of the advantage map method for calculating weights.

[0212]

[0213] Furthermore, a three-dimensional improved radar map is constructed.

[0214] The parameters (angle, radius, height) of the 3D improved radar chart are divided into corresponding (weight, evaluation value of control-related primary index, and wind power prediction accuracy), thus yielding three sets of data for the 3D improved radar chart:

[0215] (Weight W1, evaluation value of rapid voltage regulation index D) U (Constant C1), (Weight W2, Primary frequency regulation index evaluation value D) f Wind power prediction accuracy λ CDQ ) and (weight W3, inertia response index evaluation value D) J (Constant C1).

[0216] Since the frequency regulation index and wind power prediction value are only related in the previous text, the value of the third dimension of the first and third groups is constant C1 = 1.

[0217] The calculation methods for weights W1, W2, and W3 are determined using the example graph of the dominance graph method in Table 1.

[0218] Based on the three sets of data above, a three-dimensional improved radar chart can be constructed. The so-called improvement mainly refers to drawing a radar chart on the base plane of the radar chart, using the values ​​of each index as the radius of the sector, and replacing the original triangle with a sector.

[0219] Calculate the volumetric accumulation of cones in a 3D improved radar image.

[0220] The volume value reflects the calculation results of the weight and corresponding evaluation value under the condition of wind power. Therefore, the larger the volume accumulation value, the stronger the active support capability.

[0221] Specifically, suppose there are m objects to be evaluated. Calculate the cumulative volume value for each object. Then, for the i-th object:

[0222]

[0223] Among them, D iU D is the evaluation value of the fast voltage regulation index of the i-th object. if It is the evaluation value of the frequency modulation index of the i-th object. λ is the inertia response index evaluation value of the i-th object. CDQ It is the accuracy of ultra-short-term power prediction. w1, w2, and w3 represent the weights of the fast voltage regulation index, primary frequency regulation index, and inertia response index obtained above. C1 represents a constant with a value of 1.

[0224] Calculate the surface accumulation value of a cone in a 3D improved radar image.

[0225] Given a fixed volume, the smaller the surface area, the closer the shape is to a cylinder, which indicates a more balanced performance.

[0226] A i =2π(D) iU ×w1×C1+D if ×w2×λ CDQ +D iJ ×w3×C1)

[0227] Among them, D iU D is the evaluation value of the fast voltage regulation index of the i-th object. if D is the evaluation value of the frequency modulation index of the i-th object. iJ λ is the inertia response index evaluation value of the i-th object. CDQ It is the accuracy of ultra-short-term power prediction. w1, w2, and w3 represent the weights of the fast voltage regulation index, primary frequency regulation index, and inertia response index obtained above. C1 represents a constant with a value of 1.

[0228] The formula for constructing the evaluation function using the geometric mean method is as follows:

[0229] EV i =γ i1 *γ i2 i = 1, 2, 3, ..., m

[0230] Among them, EV i γ is the evaluation value for the active support and control capability of a wind farm. i1 Gamma represents the strength of active support capabilities. i2 It represents the degree of balance in abilities.

[0231] γ i1 =Vi / V max V max =max{V i |1,2,3,...,m}

[0232] Among them, V i V represents the feature-calculation result of the i-th object out of m objects to be evaluated. max This represents the maximum value of the feature-1 calculation result among the m objects to be evaluated.

[0233] γ i2 =A min / A i A min =min{A i |1,2,3,...,m}

[0234] Among them, A i A represents the feature calculation result of the i-th object out of m objects to be evaluated. min This represents the minimum value of the calculation result of feature 2 among m objects to be evaluated.

[0235] Furthermore, the evaluation score thresholds E1, E2, and E3 are obtained by dividing the evaluation function, and E1 > E2 > E3.

[0236] When the evaluation value of the active support control capability of a wind farm is greater than or equal to E1, the active support control capability is considered excellent.

[0237] When the evaluation value of the active support control capability of a wind farm is greater than or equal to E2 and less than E1, the active support control capability is considered good.

[0238] When the evaluation value of the active support control capability of a wind farm is greater than or equal to E3 and less than E2, the active support control capability is considered to be of medium quality.

[0239] When the evaluation value of the active support control capability of a wind farm is less than E3, the active support control capability is poor.

[0240] Thus, the evaluation index values ​​for the graded evaluation are derived: the comprehensive evaluation level and the comprehensive evaluation value (EV). i γ i1 γ i2 The evaluation value D of the first-level indicator of each item iU D if D iJ , λ CDQ .

[0241] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A graded evaluation method for the active support control capability of wind farms, characterized in that: include: Determine the primary indicators of the active support and control capabilities of wind farms, and construct secondary indicators based on the primary indicators; Determine the weights of the secondary indicators and obtain the corresponding evaluation values ​​for each primary indicator; A three-dimensional improved radar map is constructed, and an evaluation function is introduced based on the characteristics of the three-dimensional improved radar map to calculate the evaluation value of the active support control capability of the wind farm. To achieve a graded assessment of the active support and control capabilities of wind farms; The construction of the improved 3D radar map includes the following steps: The weight values ​​of the rapid voltage regulation index, primary frequency regulation index, and inertial response index are determined using the advantage plot method, and are used as the angles between the axes of each index in the radar chart. Using the values ​​of each indicator as the radius of the sector, a sector is drawn to replace the original triangular radar chart, thus forming an improved radar chart. Using the wind power prediction accuracy as the height of the radar chart reference plane, the third-dimensional parameter of the radar chart is formed, resulting in a three-dimensional improved radar chart. The parameters in the 3D improved radar chart—angle, radius, and height—correspond to: weight, evaluation value of control-related primary index, and wind power prediction accuracy, respectively. The method for calculating the evaluation value of the active support control capability of a wind farm based on the characteristics of the improved 3D radar image includes the following steps: The evaluation features are constructed, and the calculation formula for feature one is as follows: in, This represents the volumetric accumulation of the cone in the plotted 3D improved radar image. It is the evaluation value of the fast voltage regulation index of the i-th object. It is the evaluation value of the frequency modulation index of the i-th object. λ is the inertia response index evaluation value of the i-th object. CDQ It is the accuracy of ultra-short-term power prediction. w1, w2, and w3 represent the weights of the fast voltage regulation index, primary frequency regulation index, and inertial response index obtained by the dominance graph method. C1 represents a constant. The formula for calculating feature two is: Among them, A i Accumulate added value for the conical surface in the generated 3D improved radar map; The formula for constructing the evaluation function using the geometric mean method is as follows: in, This is the evaluation value for the active support and control capability of the wind farm. This represents the strength of their proactive support capabilities. Represents the degree of balance in abilities; in, This represents the feature-calculation result of the i-th object out of m objects to be evaluated. This represents the maximum value of the feature-1 calculation result among m objects to be evaluated; in, This represents the result of feature 2 calculation for the i-th object out of m objects to be evaluated. This represents the minimum value of the calculation result of feature 2 among m objects to be evaluated.

2. The method for evaluating the active support control capability of wind farms through graded assessment as described in claim 1, characterized in that: The primary indicators include wind power prediction accuracy, rapid voltage regulation, primary frequency regulation, and inertia response.

3. The method for evaluating the active support control capability of wind farms through graded assessment as described in claim 2, characterized in that: The primary frequency modulation (PMDM) indicators include the root mean square value of the frequency error, the performance indicators of the primary PMDM effect, and the response time compliance rate.

4. The method for evaluating the active support control capability of wind farms through graded assessment as described in claim 3, characterized in that: The formula for calculating the root mean square value of the frequency error is: in, f is the root mean square value of the frequency error. i f represents the frequency sample value of the i-th point; N represents the total number of sampling points; standard This is the standard value for the power grid frequency; The formula for calculating the performance index of the primary frequency modulation effect is as follows: in, This refers to the performance indicators of primary frequency modulation. The integral of the power consumption for a single frequency modulation operation. The integral charge is theoretically calculated for the corresponding time. The formula for calculating the response time compliance rate is as follows: Where, λ standard To achieve the response time target, N good N represents the number of times within the statistical time span that the frequency modulation power increase reaches 90% and the time taken to reach the target power increase is less than 9 seconds. all This indicates the total number of frequency modulations within the same time span.

5. The method for evaluating the active support control capability of wind farms through graded assessment as described in claim 2, characterized in that: The rapid voltage regulation index includes the voltage qualification rate. Automatic voltage control system operational rate and maximum reactive power support capacity ; The inertial response indicators include the frequency drop lower limit indicator, the power grid frequency damping effect indicator, and the power grid frequency secondary drop evaluation indicator.

6. The method for evaluating the active support control capability of wind farms through graded assessment as described in claim 2, characterized in that: The wind power prediction accuracy rate uses the ultra-short-term wind power accuracy rate, and its calculation formula is as follows: in, For ultra-short-term power prediction accuracy; P N This represents the daily operating capacity; P pi P represents the actual power at time i; mi Let be the predicted value at time i, and n be the number of samples; P mi The expression is: Among them, P mi Let be the predicted value at time i.

7. The method for evaluating the active support control capability of wind farms through graded assessment as described in claim 1, characterized in that: The method for determining the weights of the secondary indicators is as follows: based on historical data samples, the weights of the secondary indicators are determined using a combination of information entropy and information gain, specifically including the following steps: For the rapid pressure regulation index, the results are divided into four levels: A, B, C, and D. Based on historical data, historical sample data S is calculated. Each set of historical data includes three characteristics: voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity, and corresponds to one of the result levels A, B, C, and D. According to the information entropy formula: in, The information entropy of sample data S is represented by... It represents the proportion of samples of class k in the sample data S; Calculate the information entropy of historical samples: in, , , , These correspond to the proportions of samples at levels A, B, C, and D in the sample data S, respectively. The voltage qualification rate is divided into three levels, C1, C2, and C3, based on its numerical characteristics. The operational rate and maximum reactive power support capacity of the automatic voltage control system are also classified according to their numerical characteristics; According to voltage qualification rate U passRate The sample dataset S is divided into three subsets, denoted as S1, S2, and S3, based on the grades C1, C2, and C3. Based on the information entropy formula, the information entropies Ent(S1), Ent(S2), and Ent(S3) of the three subsets are calculated. Voltage qualification rate U passRate The information gain for sample S is: in, Indicates voltage qualification rate U passRate For the information gain of sample S The information entropy of sample S is represented by... Let v represent the number of samples in S that have all values ​​corresponding to the level of that subset, and let S represent the total number of samples in the sample data S. This represents the information entropy of the v-th subset; Similarly, the information gain Gain(S) of the automatic voltage control system's operational rate is obtained. The information gain Gain (S) of the maximum reactive power support capability. ); According to the formula: in, This represents the j-th weight value. Let the j-th feature of the sample set S be represented. This represents the information gain of the feature in the sample dataset S; The weights of the secondary indicators corresponding to the rapid voltage regulation index, namely voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity, are obtained. Similarly, the weight values ​​of the primary frequency modulation index and the inertia response index are obtained using the above method.

8. The method for evaluating the active support control capability of wind farms through graded assessment as described in claim 1 or 2, characterized in that: The process of obtaining the evaluation values ​​for each primary indicator includes the following steps: The evaluation value of the rapid voltage regulation index is calculated using the following formula: Among them, W1, W2, and W3 are the weights of the secondary indicators voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity. The results are the calculated voltage qualification rate, automatic voltage control system commissioning rate, and maximum reactive power support capacity. Similarly, the evaluation values ​​of the primary frequency regulation index and the inertial response index are calculated.

Images