A wind turbine primary frequency adaptive control method considering active power regulation capability

Abstract

The present invention belongs to the technical field of stable operation of electric power system, and discloses a method for adaptive control of primary frequency regulation of wind turbines considering active power regulation capability. First, a wind speed model is established, and an active load reduction control scheme for wind turbines is designed according to different wind speed intervals. An evaluation index for the potential of frequency regulation provided by active load reduction of wind turbines is proposed. Combined with subjective and objective weighting methods, a normal cloud combination is used for weight calculation to comprehensively evaluate the power regulation capability of wind turbines. When the electric power system encounters an active power shortage, the wind farm control center calculates the power regulation amount of the wind turbines based on the comprehensive evaluation results, issues adaptive power regulation instructions, and cooperates with conventional units to suppress the drop of grid frequency. The present invention adopts the comprehensive evaluation of the normal cloud combination as the basis for formulating the power regulation instructions of the wind turbines, effectively utilizes the reserve power reserved by the active load reduction of the units, optimizes the active power output of the wind farm grid connection point under disturbance, and improves the stability of the grid.

Description

Technical Field

[0001] The invention belongs to the technical field of stable operation of electric power systems, and in particular relates to a primary frequency adaptive control method for a wind turbine generator set taking active power regulation capability into consideration. Background Art

[0002] With the rapid growth of the scale of new energy installed capacity and the deepening of the degree of power electronics in the power grid, the problem of insufficient frequency regulation resources has become increasingly prominent, and frequency accidents have affected the safe and stable operation, bringing new challenges to the active power and frequency regulation of the power system. Therefore, in order to meet the demand for frequency support capabilities of power systems with a high proportion of renewable energy, it has become an international consensus to tap the frequency support capabilities of new energy stations. In fact, wind turbines can reduce their active output under normal operating conditions by adopting means such as rotor overspeed control and pitch angle control, and can reserve a certain amount of spare power, which can provide fast and effective frequency support for the power grid frequency. At present, some studies have focused on the application of the rapid frequency support capabilities of new energy stations. In the existing technology, the active control method in the frequency feedback inertia control is used to enable the reserved load shedding standby wind turbines to participate in the system frequency regulation; however, there is still little research on how each new energy unit in the station responds to frequency requirements, and how to combine the real-time active regulation capabilities of the units to issue appropriate power regulation instructions to participate in primary frequency regulation.

[0003] On the other hand, considering the influence of wake effect, the inflow wind speed of each unit in a large-scale wind farm is affected by the upstream unit, which is difficult to ignore, and puts forward higher requirements for the evaluation of the power regulation capability of the unit. At present, the national standard GB∕T 43904-2024 "Evaluation Index System of Wind Turbine Generators in Wind Energy Generation System" defines a variety of indicators that can be used to evaluate the operation of units from multiple perspectives. It describes multiple unit operation evaluation indicators such as power generation performance and availability. However, when the wind power generation system performs active frequency regulation, it is more dependent on the spare power of the unit. Therefore, the effectiveness index of the utilization of spare power needs to be further quantified and defined. On this basis, it is also necessary to reasonably allocate the regulation demand of active power, and finally achieve effective response to the frequency drop after the disturbance. Therefore, how to effectively evaluate the ability of wind turbines to participate in frequency regulation by load reduction and standby, and design a reasonable wind power active frequency support plan based on this, so as to improve the system's ability to respond to frequency drop accidents, and thus improve the stability of power grid operation, has become one of the issues that need to be considered at present. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a primary frequency regulation adaptive control method for wind turbines taking into account the active power regulation capability. By fully tapping the active power regulation capability of the wind farm, effectively utilizing the standby power of the wind turbines, and designing a reasonable wind power active frequency support scheme, the frequency regulation pressure of traditional units can be reduced to a certain extent, providing a feasible technical solution for the stable operation of the power grid under large-scale access of clean energy.

[0005] The present invention provides a method for adaptively controlling primary frequency regulation of a wind turbine generator system taking active power regulation capability into consideration, comprising the following steps:

[0006] Step 1: Considering the influence of wake effect, establish different wind turbine inflow wind speed models for a single wind farm, and design active load reduction control schemes for wind turbines according to different wind speed ranges;

[0007] Step 2: From the perspectives of wind turbine reserve availability, wind power operation characteristics and regulation reliability, propose evaluation indicators for wind turbine active load reduction to provide reserve and frequency regulation potential, including average reserve power, power generation availability, power characteristic offset, frequency regulation capability coefficient, average fault duration and regulation rate. The values ​​of various indicators are calculated based on real-time data collected from wind farms or historical experience.

[0008] Step 3: For each evaluation index calculation of each wind turbine, the subjective and objective weighting methods are used to calculate the weight separately, and then the normal cloud combination method is used to calculate the comprehensive weight, and then the similarity value between each wind turbine and the ideal cloud combination model is calculated, and the power regulation capability of each wind turbine is comprehensively evaluated;

[0009] Step 4: When the power system encounters an active power shortage, the wind farm control center calculates the active power adjustment amount in the adjustment command based on the estimated active power demand and the comprehensive evaluation results of each wind turbine unit and the adjustable power amount, and issues adaptive power adjustment commands to each unit to coordinate with conventional units to jointly suppress the drop in grid frequency; the power allocation coefficient is used to integrate the power control capability of the unit and the comprehensive evaluation results, and then the power increase adjustment amount of each unit is obtained, and finally the active power output of the entire wind farm grid connection point is optimized.

[0010] Furthermore, in step 1, a model of wind speed inflow of different wind turbines is established for a single wind farm, namely, the Jensen wake model. According to the law of conservation of mass and momentum theory, the inflow wind speed of the downstream wind turbine i under the influence of the wake effect in the Jensen wake model is obtained. for:

[0011] (1)

[0012] In the formula, is the incoming wind speed; b is the attenuation coefficient, which is used to describe the change of the wake radius of the wind turbine; r is the wind turbine rotor radius; x is used to quantify the spatial distance between the downstream unit and the upstream unit; C T is the thrust coefficient of the wind turbine, when C T When it increases, the influence of wake effect on the wind speed of downstream units increases;

[0013] In a large wind farm, the downstream wind turbines are affected by the superposition of the wake effects of multiple upstream turbines. At this time, the inflow wind speed of the downstream turbine i under the superposition of the upstream turbine i-1 is for:

[0014] (2)

[0015] In the formula, v j and x j are the inflow wind speed of the jth unit and the distance between units i and j respectively.

[0016] Furthermore, for a wind farm with n units, its output power is expressed as:

[0017] (3)

[0018] In the formula, is the total output active power of the wind farm; is the air density; P i and C Pi are the output power and wind energy utilization coefficient of the i-th unit under the influence of the wake;

[0019] Assuming the load reduction level is d%, the load reduction reserve capacity of the entire wind farm is ; Considering different wind speed ranges, different load reduction schemes are adopted for the units as follows:

[0020] when When the unit does not participate in frequency regulation, if the tip speed ratio of the unit after load reduction is , then the minimum operating wind speed From the minimum operating speed Relationship to obtain;

[0021] when When the inflow wind speed is less than the rated wind speed When the unit adopts overspeed load reduction control;

[0022] when When the inflow wind speed is less than the cut-off wind speed When the unit is at load reduction control, the pitch angle is used.

[0023] Furthermore, in step 2, evaluation indicators are proposed from multiple perspectives, including reserve availability, wind power operation characteristics, and regulation reliability, to evaluate the active regulation capability of wind turbines using active load reduction control, specifically:

[0024] 1) Reserve availability:

[0025] The response of each wind turbine to the load reduction command is evaluated, and the average reserve power index I1 is defined to reflect the average power reserve of the wind turbine to meet the system frequency regulation demand during the test and recording period. The expression is as follows:

[0026] (4)

[0027] Where T is the test duration, t is time, and I(t) is a 0-1 integer variable. When the current speed of the wind turbine is greater than the minimum operating speed hour, , otherwise it is 0; The load-shedding standby power for unit i;

[0028] The indicator I2 is defined as the available power generation rate of the unit under load reduction, and the actual power generation of the wind turbine under load reduction operation is calculated as a percentage of the potential power generation; the larger I2 is, the smaller the potential of the wind turbine to participate in frequency regulation is;

[0029] (5)

[0030] Where P ih is the actual power generation of unit i, P iLW P is the power loss caused by wind turbines; iLNW The power loss is caused by reasons other than wind turbines, including power loss due to load shedding, power limiting, shutdown, etc.

[0031] 2) Wind power operation characteristics:

[0032] In order to measure the control level of wind turbine power generation performance, the index I3 is set as the power characteristic offset; wherein, the standardized deviation of the normalized power data distribution is calculated by the following formula (6):

[0033] (6)

[0034] In the formula, is the power characteristic deviation of wind turbine i; N represents the number of wind turbine active output data collected within a period of time, represents the normalized average power during this period, Represents the normalized output power data of wind turbine i;

[0035] The frequency regulation capability coefficient index I4 is defined to reflect the frequency regulation capability of the wind turbine. When the system frequency decreases, the ability of the wind turbine to participate in frequency regulation is determined by the actual rotor kinetic energy that can be released by the wind turbine. Therefore, the available rotor kinetic energy factor k of the wind turbine is defined. C for:

[0036] (7)

[0037] Where: , and are the wind turbine speed, the minimum operating speed of the wind turbine and the maximum operating speed of the wind turbine respectively, k 、E kmax and E kmin are the rotor kinetic energy of the wind turbine at the current speed, maximum speed and minimum speed respectively;

[0038] In order to reflect the frequency regulation potential of wind turbines, the available capacity factor k is defined as J for:

[0039] (8)

[0040] Where: P N_i and P imin is the rated output power of the wind turbine and the output power when it is at the lower speed limit;

[0041] Taking the rotor kinetic energy and available capacity into consideration to reflect the frequency regulation capability of the unit, the final indicator I4 is expressed as:

[0042] (9)

[0043] In the formula, The speed of the wind turbine when it is running at reduced load;

[0044] 3) Regulation reliability:

[0045] There are problems such as transmission delay and data packet loss in the information transmission and command issuance stages of each wind turbine in the wind farm, which makes the unit unable to respond to the load reduction command and participate in the system frequency regulation; the indicator I5 is defined as the average fault time, which is expressed as:

[0046] (10)

[0047] Where: T is the test duration, D(t) is a 0-1 variable reflecting whether the wind turbine is faulty at time t. When the operating point at this time is shut down or under maintenance, , otherwise it is 0;

[0048] Define the adjustment rate index I6, which reflects the speed at which the wind turbine responds to the power change command, and is expressed as:

[0049] (11)

[0050] In the formula, is the power change value of wind turbine i in the jth test; is the duration of the jth test; is the rated output power of wind turbine i; l is the total number of tests.

[0051] Furthermore, in step 3, the weights are calculated separately using subjective and objective weighting methods, specifically:

[0052] The hierarchical analysis method in the subjective weighting method is used to construct a judgment matrix for indicators I1~I6 , where element a in A uv For the evaluation index I u Compared with index I v The scale value of the degree of importance, , using a 1 to 9 level scale;

[0053] The consistency test of the judgment matrix is ​​performed, and the consistency index C I and consistency ratio C R The calculation formula is:

[0054] (12)

[0055] In the formula, a is the order of the judgment matrix; R I It is a random consistency indicator for judging the matrix order as a; after passing the consistency test, the maximum eigenvalue of matrix A is calculated The corresponding eigenvector Z is normalized to obtain the weight coefficient w of each evaluation index 1k :

[0056] (13);

[0057] The entropy weight method in the objective weighting method is used to obtain the information utility value of the selected indicator, thereby quantifying the impact of the indicator change and determining the weight of each indicator; assuming that the active control capability of n wind turbines in the wind farm is measured by 6 indicators, the initial matrix of the sample is for:

[0058] (14)

[0059] In the formula, x ik Refers to the value of the kth indicator corresponding to the i-th wind turbine, ; Each index is normalized as follows to obtain index x ik The standardized form of for:

[0060] (15)

[0061] Determine the information entropy value e of the indicator k and the information utility value g of each indicator k for:

[0062] (16)

[0063] Get the weight w of each indicator 2k for:

[0064] (17);

[0065] The entropy weight method is very sensitive to abnormal indicators, so the combined coefficient of variation method is considered to weaken the impact of abnormal indicators and increase the rationality of the weighting results. The coefficient of variation method (CVM) is another objective weighting method. Its basic idea is to directly use the information contained in the indicator to calculate the weight; on the basis of completing the normalization of the sample data, the average value of different sample data is obtained. and standard deviation :

[0066] (18)

[0067] Then the coefficient of variation E is calculated k and the weight of each indicator w 3k :

[0068] (19).

[0069] Furthermore, in step 3, the normal cloud combined weight method is used for comprehensive evaluation, specifically:

[0070] Construct a weighted sample set W={W1,W2,W3}, where W i ={w i1 ,w i2 ,…,w i6}, i∈{1,2,3}, forming a weight sample matrix as shown in formula (20):

[0071] (20)

[0072] Based on the sample matrix, the sample mean of the indicator weight is calculated using statistical methods , the first-order absolute center distance of weighted samples and the sample variance On this basis, we get the normal cloud combination weight N W The expected Ex, entropy En and super entropy He of (Ex, En, He) are:

[0073] (twenty one),

[0074] For two normal clouds N1 (Ex1, En1, He1) and N2 (Ex2, En2, He2), using the concept of fuzzy proximity, the cloud model similarity calculation method based on fuzzy proximity is given as follows:

[0075] (twenty two),

[0076] In the formula, is the fuzzy proximity calculation parameter, is the standard normal distribution function, is the model similarity of normal clouds N1 and N2.

[0077] Assume any n normal clouds N1(Ex1, En1, He1), N2(Ex2, En2, He2), ..., N n (Ex n ,En n , He n ) and constant , where Ex is the expected value of the cloud, En is the entropy value of the cloud, and He is the super entropy value of the cloud. The comprehensive cloud model N(Ex) after the weighted integration of these n normal clouds is N ,En N , He N )ExpectedEx N 、Entropy N and super entropy He N It is expressed as:

[0078] (twenty three),

[0079] Based on equations (20) to (23), considering indicators I1 to I6, a comprehensive evaluation is conducted on n wind turbines to calculate the power regulation capability score of each wind turbine. The specific steps are as follows:

[0080] 1) Construct a sample matrix and obtain X max =[max(x i1 ),max(x i2 ),…,max(x i6 )] is the ideal unit state, where i={1,2,…,n}, max(x i1 ) is the ideal unit state of index I1, and so on;

[0081] 2) Calculate the weights of each indicator under the analytic hierarchy process, entropy weight method and coefficient of variation method, construct the weight matrix and calculate the normal cloud combination weight;

[0082] 3) According to formula (23), the comprehensive cloud model after weighted integration of normal clouds under the state of n units and ideal units is calculated, and the cloud similarity is calculated according to formula (22) to obtain the similarity value S between the state of n units and the ideal unit. The larger the value, the better the state of the unit and the greater the frequency regulation potential.

[0083] 4) Score the units according to their similarity value S, and set the unit m with the highest value among the n units as the full score SC m , the similarity values ​​of unit i and unit m are S i and S m , then we get the score SC of unit i i for .

[0084] Furthermore, in step 4, when a disturbance occurs in the system causing an active power shortage, which in turn leads to a frequency drop fault, an adaptive instruction is issued to call the standby power of the wind turbine to provide support and coordinate with the conventional unit to perform a frequency adjustment according to the evaluation result of the active power regulation capability of the wind turbine.

[0085] When the frequency is disturbed, the power shortage observation value Proportional to the rate of change of frequency, determined from the equation of motion for the system's center of inertia:

[0086] (twenty four),

[0087] Where H is the system equivalent inertia time constant, f n is the rated frequency of the system; based on formula (24), the power forecast value required for system compensation is obtained as the total power demand instruction of the system, and the power demand of the wind farm is obtained based on historical operation experience. Compared with the actual power output of the entire wind farm, the active power deviation is obtained. ;This deviation is the total power that the wind farm needs to adjust to meet the real-time needs of the grid;

[0088] On the basis of considering the adjustable power of wind turbines, the wind farm active power distributor further divides the wind farm active power deviation ΔP according to the comprehensive score of the regulation potential of the i-th wind turbine. farm Assigned to each unit to achieve power regulation according to capacity; Specifically for wind turbine i, the power regulation amount received in the instruction is expressed as:

[0089] (25),

[0090] In the formula, k i represents the power increase weight coefficient of unit i, and are the power allocation coefficient according to the adjustable power of the unit and the power allocation coefficient according to the evaluation score of the unit, respectively, and , P fi represents the adjustable power of unit i, then the active power P of the wind farm grid connection point PCC It is expressed as:

[0091] (26).

[0092] The beneficial effects described in the present invention are as follows: the method described in the present invention takes into account the comprehensive situation of the internal units of the wind farm, selects indicators that consider factors such as load reduction characteristics, communication capabilities and unit response capabilities, and uses the normal cloud combination method to comprehensively score the wind turbines. This is used as the basis for calculating the unit power adjustment instructions. The spare power reserved for active load reduction of the unit can be effectively utilized, the active power output of the wind farm grid-connected point under disturbance can be optimized, the reliability and rapidity of the adjustment can be ensured, and the stability of the power grid can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] Figure 1 is a flow chart of the method of the present invention;

[0094] Figure 2 It is the wind farm grid connection topology diagram;

[0095] Figure 3 It is a schematic diagram of the weights of various indicators of the unit;

[0096] Figure 4 It is a schematic diagram of frequency adjustment in different scenarios. DETAILED DESCRIPTION

[0097] In order to make the contents of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments in conjunction with the accompanying drawings.

[0098] like Figure 1 As shown, the present invention provides a method for adaptively controlling primary frequency regulation of a wind turbine generator system taking active regulation capability into consideration, comprising the following steps:

[0099] Step 1: Considering the influence of wake effect, establish different wind turbine inflow wind speed models for a single wind farm, and design active load reduction control schemes for wind turbines according to different wind speed ranges;

[0100] Step 2: From the perspectives of wind turbine reserve availability, wind power operation characteristics and regulation reliability, propose evaluation indicators for wind turbine active load reduction to provide reserve and frequency regulation potential, including average reserve power, power generation availability, power characteristic offset, frequency regulation capability coefficient, average fault duration and regulation rate. The values ​​of various indicators are calculated based on real-time data collected from wind farms or historical experience.

[0101] Step 3: For each evaluation index calculation of each wind turbine, the subjective and objective weighting methods are used to calculate the weight separately, and then the normal cloud combination method is used to calculate the comprehensive weight, and then the similarity value between each wind turbine and the ideal cloud combination model is calculated, and the power regulation capability of each wind turbine is comprehensively evaluated;

[0102] Step 4: When the power system encounters an active power shortage, the wind farm control center calculates the active power adjustment amount in the adjustment command based on the estimated active power demand and the comprehensive evaluation results of each wind turbine unit and the adjustable power amount, and issues adaptive power adjustment commands to each unit to coordinate with conventional units to jointly suppress the drop in grid frequency; the power allocation coefficient is used to integrate the power control capability of the unit and the comprehensive evaluation results, and then the power increase adjustment amount of each unit is obtained, and finally the active power output of the entire wind farm grid connection point is optimized.

[0103] In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.

[0104] Building a wind farm model Figure 2 As shown, the wind farm contains 15 direct-drive wind turbines with a rated capacity of 2MW, which are collected at site 1 with a voltage level of 35kV after passing through a 0.69 / 35kV transformer and cables, and then sent to the 230kV distribution network after being stepped up.

[0105] Among them, the real-time status acquisition device for the calculation source of the effectiveness index of the wind turbine in the present invention is the wind farm SCADA system (Supervisory Control and Data Acquisition), that is, the monitoring and data acquisition system, which is the core technology platform for wind farm operation and management. The system integrates advanced sensor technology, communication protocols, data processing and monitoring software to realize real-time data acquisition, transmission, processing and analysis of key equipment such as wind turbines, substations, and meteorological stations in the wind farm. The main functions of the SCADA system cover data acquisition, data transmission, data processing, monitoring and control, and safety management, ensuring the safety, stability and efficiency of the operation of the wind farm. The original data collected by the wind farm SCADA system is processed and mined to obtain the initial values ​​of the indicators of each unit under load reduction operation. After standardization, the values ​​of the six indicators of each unit are obtained as shown in Table 1 below.

[0106] Table 1 Wind turbine frequency regulation potential index values

[0107]

[0108] Since the meanings of various indicators are different, such as indicators I2, I3, and I5 are negative indicators, the larger their values ​​are, the smaller the unit's regulation potential is. Therefore, before evaluation, all indicators must be converted into positive indicators, that is, the larger the value is, the greater the unit's regulation potential is. After that, the hierarchical analysis method is used to subjectively assign weights to each indicator, and the judgment matrix is ​​constructed as follows:

[0109] ,

[0110] Query the standard table of average random consistency indicators and get R when the matrix dimension is 6 I =1.24, and finally the weighting results of each indicator under the hierarchical analysis method are shown in Table 2.

[0111] Table 2 Subjective weights of each indicator in the analytic hierarchy process (AHP)

[0112]

[0113] Then, the indicator weight value w under the entropy weight method and the coefficient of variation method is calculated 2k and w 3k As shown in Table 3.

[0114] Table 3 Objective weights of each indicator using entropy weight method and coefficient of variation method

[0115]

[0116] The weights of the three indicators are as follows: Figure 3 shown.

[0117] Three weighting methods were used to evaluate the above 15 units separately, and the scores of each unit are shown in Table 4.

[0118] Table 4 Evaluation results of different weighting methods

[0119]

[0120] Based on the above subjective and objective weight calculation results, the normal cloud combination method is used for comprehensive scoring and the unit ranking results are shown in Table 5.

[0121] Table 5 Comprehensive evaluation results of wind turbines

[0122]

[0123] The weights and scoring results obtained by the normal cloud combination method can comprehensively reflect the adjustable capacity of the unit, increase the effectiveness and rationality of the results, and sort the states of wind turbines in this way, providing a basis for the formulation of active power support schemes for the scenarios involved in the present invention.

[0124] The rated power generation of the system is 100MW, the rated power generation of the wind farm is 30MW, and the wind power penetration rate is 30%. When the power grid system experiences a total active disturbance of 0.04pu in 2s, the frequency regulation results under different scenarios are compared. Figure 4 As shown in Figure 2, three scenarios are compared: ① Wind turbines do not participate in frequency regulation, only conventional turbines do frequency regulation; ② Wind turbines and conventional turbines participate in frequency regulation at the same time, where wind turbines use the traditional frequency regulation scheme and the power demand estimate ΔP is farm The average power regulation amount received by each unit is the same; ③ The wind turbine and conventional units participate in frequency regulation at the same time, among which the wind turbine adopts the scheme of the present invention to adjust the frequency, and the power demand estimate ΔP is calculated based on the comprehensive evaluation results. farm Send it down and take .pass Figure 4 It can be seen that the participation of wind turbines in frequency modulation can effectively suppress frequency drops, and the method of the present invention has better frequency regulation benefits, faster frequency response, and smaller maximum frequency drop compared to the traditional primary frequency modulation method.

[0125] The method described in the present invention takes into account the influence of the wake effect on the operation and regulation capability of the unit, describes the regulation of the wind turbine in the actual operation process from multiple angles such as the reserve availability, wind power operation characteristics and regulation reliability, and combines a variety of subjective and objective weighting methods. The normal cloud combination method is used to evaluate the active regulation capability of the unit, effectively dealing with the uncertainty of weight assignment and information fusion problems, improving the effectiveness and rationality of the combined weight results, and providing a basis for the setting of the scheme for wind farms to participate in frequency regulation under active disturbances. When the power grid is disturbed by frequency, the frequency regulation method that issues instructions based on the evaluation results of the unit's regulation potential can effectively suppress the frequency drop, and the frequency response is faster than the traditional method, and the maximum value of the frequency drop is smaller.

[0126] The above description is only a preferred embodiment of the present invention and is not intended to be a further limitation of the present invention. All equivalent changes made using the contents of the present specification and drawings are within the protection scope of the present invention.

Claims

1. A method for adaptive control of primary frequency regulation of a wind turbine generator system considering active power regulation capability, characterized in that: The following steps are involved: Step 1: Considering the influence of wake effect, establish different wind turbine inflow wind speed models for a single wind farm, and design active load reduction control schemes for wind turbines according to different wind speed ranges; Step 2: From the perspectives of wind turbine reserve availability, wind power operation characteristics and regulation reliability, propose evaluation indicators for wind turbine active load reduction to provide reserve and frequency regulation potential ， Includes average standby power indicator I 1. Power generation availability index I 2. Power characteristic offset index I 3. Frequency modulation capability coefficient index I 4. Average failure time indicator I 5 and adjustment rate indicators I 6. Calculate the values ​​of various indicators of the computer group based on the real-time data or historical experience collected from the wind farm; Step 3: For each evaluation index calculation of each wind turbine, the subjective and objective weighting methods are used to calculate the weight separately, and then the normal cloud combination method is used to calculate the comprehensive weight, and then the similarity value between each wind turbine and the ideal cloud combination model is calculated, and the power regulation capability of each wind turbine is comprehensively evaluated; Among them, the weights are calculated separately using subjective and objective weighting methods, specifically: Using the hierarchical analysis method, a judgment matrix is ​​constructed for indicators I1~I6 , where element a in A uv For the evaluation index I u Compared with index I v The scale value of the degree of importance, , using a 1 to 9 level scale; The consistency test of the judgment matrix is ​​performed, and the consistency index C I and consistency ratio C R The calculation formula is: (12)， In the formula, a is the order of the judgment matrix; R I It is a random consistency indicator for judging the matrix order as a; after passing the consistency test, the maximum eigenvalue of matrix A is calculated The corresponding eigenvector Z is normalized to obtain the weight coefficient w of each evaluation index 1k : (13)； The entropy weight method is used to obtain the information utility value of the selected indicator, thereby quantifying the impact of the indicator change and determining the weight of each indicator; assuming that the active control capability of n wind turbines in the wind farm is measured by 6 indicators, the initial matrix of the sample is for: (14)， In the formula, x ik Refers to the value of the kth indicator corresponding to the i-th wind turbine, ; Each index is normalized as follows to obtain index x ik The standardized form of for: (15)， Determine the information entropy value e of the indicator k and the information utility value g of each indicator k for: (16)， Get the weight w of each indicator 2k for: (17)； Considering the influence of abnormal indicators, the coefficient of variation method is used to obtain the average value of different sample data on the basis of completing the normalization of sample data. and standard deviation : (18)， Then the coefficient of variation E is calculated k and the weight of each indicator w 3k : (19)； The normal cloud combination method is used to calculate the comprehensive weight, specifically: Construct a weighted sample set W={W1,W2,W3}, where W i ={w i1 ,w i2 ,…,w i6 }, i∈{1,2,3}, forming a weight sample matrix as shown in formula (20): (20)， Based on the sample matrix, the sample mean of the indicator weight is calculated using statistical methods , the first-order absolute center distance of weighted samples and the sample variance On this basis, we get the normal cloud combination weight N W The expected Ex, entropy En and super entropy He of (Ex, En, He) are: (21)， For two normal clouds N1 (Ex1, En1, He1) and N2 (Ex2, En2, He2), using the concept of fuzzy proximity, the cloud model similarity calculation method based on fuzzy proximity is given as follows: (22)， In the formula, is the fuzzy proximity calculation parameter, is the standard normal distribution function, is the model similarity of normal clouds N1 and N2; Assume any n normal clouds N1(Ex1, En1, He1), N2(Ex2, En2, He2), ..., N n (Ex n ,En n , He n ) and constant , where Ex is the expected value of the cloud, En is the entropy value of the cloud, and He is the super entropy value of the cloud. The comprehensive cloud model N(Ex) after the weighted integration of these n normal clouds is N ,En N , He N )ExpectedEx N 、Entropy N and super entropy He N It is expressed as: (23)； Step 4: When the power system encounters an active power shortage, the wind farm control center calculates the active power adjustment amount in the adjustment command based on the estimated active power demand and the comprehensive evaluation results of each wind turbine unit and the adjustable power amount, and issues adaptive power adjustment commands to each unit to coordinate with conventional units to jointly suppress the drop in grid frequency; the power allocation coefficient is used to integrate the power control capability of the unit and the comprehensive evaluation results, and then the power increase adjustment amount of each unit is obtained, and finally the active power output of the entire wind farm grid connection point is optimized.

2. A method for adaptive control of primary frequency regulation of a wind turbine generator system considering active power regulation capability according to claim 1, characterized in that: In step 1, a model of wind speed inflow for different wind turbines is established for a single wind farm, namely the Jensen wake model. According to the law of conservation of mass and momentum theory, the inflow wind speed of the downstream wind turbine i under the influence of the wake effect in the Jensen wake model is obtained: for: (1)， In the formula, is the incoming wind speed; b is the attenuation coefficient, which is used to describe the change of the wake radius of the wind turbine; r is the wind turbine rotor radius; x is used to quantify the spatial distance between the downstream unit and the upstream unit; C T is the thrust coefficient of the wind turbine, when C T When it increases, the influence of wake effect on the wind speed of downstream units increases; In a large wind farm, the downstream wind turbines are affected by the superposition of the wake effects of multiple upstream turbines. At this time, the inflow wind speed of the downstream turbine i under the superposition of the upstream turbine i-1 is for: (2)， In the formula, v j and x j are the inflow wind speed of the jth unit and the distances between units i and j respectively.

3. A method for adaptive control of primary frequency regulation of wind turbine generator system considering active power regulation capability according to claim 2, characterized in that: For a wind farm with n units, its output power is expressed as: (3)， In the formula, is the total output active power of the wind farm; is the air density; P i and C Pi are the output power and wind energy utilization coefficient of the i-th unit under the influence of the wake; Assuming the load reduction level is d%, the load reduction reserve capacity of the entire wind farm is ; Considering different wind speed ranges, different load reduction schemes are adopted for the units as follows: when When the unit does not participate in frequency modulation, if the tip speed ratio of the unit after load reduction is , then the minimum operating wind speed From the minimum operating speed Relationship to seek; when When the inflow wind speed is less than the rated wind speed When the unit adopts overspeed load reduction control; when When the inflow wind speed is less than the cut-off wind speed When the unit is at load reduction control, the pitch angle is used.

4. A method for adaptive control of primary frequency regulation of wind turbine generator system considering active power regulation capability according to claim 3, characterized in that: In step 2, evaluation indicators are proposed from multiple perspectives, including reserve availability, wind power operation characteristics, and regulation reliability, to evaluate the active regulation capability of wind turbines using active load reduction control. Specifically, they are: 1) Reserve availability: The response of each wind turbine to the load reduction command is evaluated, and the average reserve power index I1 is defined to reflect the average power reserve of the wind turbine to meet the system frequency regulation demand during the test and recording period. The expression is as follows: (4)， Where T is the test duration, t is time, and I(t) is a 0-1 integer variable. When the current speed of the wind turbine is greater than the minimum operating speed hour, , otherwise it is 0; The load-shedding standby power for unit i; The indicator I2 is defined as the available power generation rate of the unit under load reduction, and the actual power generation of the wind turbine under load reduction operation is calculated as a percentage of the potential power generation; the larger I2 is, the smaller the potential of the wind turbine to participate in frequency regulation is; (5)， Where P ih is the actual power generation of unit i, P iLW P is the power loss caused by wind turbines; iLNW The power is lost due to reasons other than wind turbines; 2) Wind power operation characteristics: In order to measure the control level of wind turbine power generation performance, the index I3 is set as the power characteristic offset; wherein, the standardized deviation of the normalized power data distribution is calculated by the following formula (6): (6)， In the formula, is the power characteristic deviation of wind turbine i; N represents the number of wind turbine active output data collected within a period of time, represents the normalized average power during this period, Represents the normalized output power data of wind turbine i; The frequency regulation capability coefficient index I4 is defined to reflect the frequency regulation capability of the wind turbine. When the system frequency decreases, the ability of the wind turbine to participate in frequency regulation is determined by the actual rotor kinetic energy that can be released by the wind turbine. Therefore, the available rotor kinetic energy factor k of the wind turbine is defined. C for: (7)， Where: , and are the wind turbine speed, the minimum operating speed of the wind turbine and the maximum operating speed of the wind turbine respectively, k 、E kmax and E kmin are the rotor kinetic energy of the wind turbine at the current speed, maximum speed and minimum speed respectively; In order to reflect the frequency regulation potential of wind turbines, the available capacity factor k is defined as J for: (8)， Where: P N_i and P imin is the rated output power of the wind turbine and the output power when it is at the lower speed limit; Taking the rotor kinetic energy and available capacity into consideration to reflect the frequency regulation capability of the unit, the final indicator I4 is expressed as: (9)， In the formula, The speed of the wind turbine when it is running at reduced load; 3) Regulation reliability: The indicator I5 is defined as the average failure time, expressed as: (10)， Where: T is the test duration, D(t) is a 0-1 variable reflecting whether the wind turbine is faulty at time t. When the operating point at this time fails to shut down or is in maintenance status, , otherwise it is 0; Define the adjustment rate index I6, which reflects the speed at which the wind turbine responds to the power change command, and is expressed as: (11)， In the formula, is the power change value of wind turbine i in the jth test; is the duration of the jth test; is the rated output power of wind turbine i; l is the total number of tests.

5. A method for adaptive control of primary frequency regulation of wind turbine generator system considering active power regulation capability according to claim 4, characterized in that: In step 3, based on equations (20) to (23), considering indicators I1 to I6, a comprehensive evaluation is performed on n wind turbines to calculate the power regulation capability score of each wind turbine. The specific steps are as follows: 1) Construct a sample matrix and obtain X max =[max(x i1 ),max(x i2 ),…,max(x i6 )] is the ideal unit state, where i={1,2,…,n}, max(x i1 ) is the ideal unit state of index I1, and so on; 2) Calculate the weights of each indicator under the analytic hierarchy process, entropy weight method and coefficient of variation method, construct the weight matrix and calculate the normal cloud combination weight; 3) According to formula (23), the comprehensive cloud model after weighted integration of normal clouds under the state of n units and ideal units is calculated, and the cloud similarity is calculated according to formula (22) to obtain the similarity value S between the state of n units and the ideal unit. The larger the value, the better the state of the unit and the greater the frequency regulation potential. 4) Score the units according to their similarity value S, and set the unit m with the highest value among the n units as the full score SC m , the similarity values ​​of unit i and unit m are S i and S m , then we get the score SC of unit i i for .

6. A method for adaptive control of primary frequency regulation of a wind turbine generator system taking active power regulation capability into consideration according to claim 5, characterized in that: In step 4, when a disturbance occurs in the system causing an active power shortage, which in turn leads to a frequency drop fault, an adaptive instruction is issued to call the standby power of the wind turbine to provide support and coordinate with the conventional unit to perform a frequency adjustment according to the evaluation result of the active power regulation capability of the wind turbine. When the frequency is disturbed, the power shortage observation value Proportional to the rate of change of frequency, determined from the equation of motion for the system's center of inertia: (24)， Where H is the system equivalent inertia time constant, f n is the rated frequency of the system; based on formula (24), the power forecast value required for system compensation is obtained as the total power demand instruction of the system, and the power demand of the wind farm is obtained based on historical operation experience. Compared with the actual power output of the entire wind farm, the active power deviation is obtained. ; This deviation is the total power that the wind farm needs to adjust to meet the real-time demand of the grid; On the basis of considering the adjustable power of wind turbines, the wind farm active power distributor further divides the wind farm active power deviation ΔP according to the comprehensive score of the regulation potential of the i-th wind turbine. farm Assigned to each unit to achieve power regulation according to capacity; Specifically for wind turbine i, the power regulation amount received in the instruction is expressed as: (25)， In the formula, k i represents the power increase weight coefficient of unit i, and are the power allocation coefficient according to the adjustable power of the unit and the power allocation coefficient according to the evaluation score of the unit, respectively, and , P fi represents the adjustable power of unit i, then the active power P of the wind farm grid connection point PCC It is expressed as: (26)。

Images