Wind power fuzzy variable droop control strategy considering source load random fluctuation characteristics

Abstract

The invention discloses a wind power fuzzy variable droop control strategy considering the random fluctuation characteristic of a source load, and the control strategy comprises the steps: taking the fluctuation of two ends of the source load into consideration while calculating the input quantity of fuzzy control, and putting forward the comprehensive variable quantity of the source load as the input variable of the fuzzy control; and the droop coefficient parameter of the wind turbine generator frequency modulation is adaptively adjusted along with the fluctuation of the source load. According to the strategy, on the basis of a wind turbine generator mathematical model, a low-order system frequency response model, a droop control strategy model and a fuzzy control theory, source end power fluctuation is obtained by obtaining the mathematical relation of wind speed indexes such as turbulence characteristics and average wind speed, and a calculation method of source-load comprehensive variation is provided in combination with load end power fluctuation. Fuzzy control is adopted to carry out real-time setting on the droop coefficient, so that the frequency modulation effect under source-load random fluctuation is improved. The wind power fuzzy variable droop control strategy has good effectiveness and superiority in the aspects of frequency modulation effect, frequency stability and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbines participating in grid frequency regulation, and in particular to the technical field of fuzzy variable droop control strategies in primary frequency regulation of power systems. Specifically, the present invention relates to a wind power fuzzy variable droop control strategy that takes into account the random fluctuation characteristics of sources and loads. Background Art

[0002] The continuous increase in installed wind power capacity has brought a series of new challenges to the power grid. Most wind turbines today are connected to the grid through advanced power electronic converters, resulting in a decoupling of the turbine rotor speed and grid frequency. This means that the turbine cannot effectively and promptly respond to system-side frequency changes. The reduced inertia of the grid will affect the system's frequency stability.

[0003] Based on the current demand for wind turbines to respond to grid primary frequency regulation, methods for wind power participation in primary frequency regulation can be categorized into three types based on energy source: rotor kinetic energy control, power reserve control, and combined wind and energy storage control. Rotor kinetic energy control, including virtual inertia control and droop control, can maintain frequency stability by overgenerating power for a short period of time. However, the use of fixed-parameter droop control methods is not conducive to the dynamic response of wind turbines. Therefore, researchers have proposed a variable droop coefficient control strategy based on fuzzy control.

[0004] Existing fuzzy variable droop control strategies for wind power are mostly based on research scenarios involving constant wind speeds. In such scenarios, using frequency deviation and frequency change rate as deviation inputs for fuzzy control can effectively improve frequency regulation. However, in reality, wind power faces random fluctuations at both the source and load ends. Since actual wind speeds in nature are primarily turbulent wind speeds with high randomness and volatility, wind power control can become difficult or even fail. Furthermore, varying load-side fluctuations can also affect wind turbines operating in turbulent wind speeds, thereby impacting the frequency regulation of wind turbines participating in the grid.

[0005] When the source end has a randomly fluctuating wind speed, the frequency deviation indicator will have mixed fluctuations at both the source and load ends. The fuzzy control will not be able to correctly identify the frequency characteristics of wind power participating in frequency regulation and the system, resulting in the failure of the strategy. Specifically, when the source end wind speed increases, the frequency will fluctuate upward. At this time, load fluctuations will occur, and the frequency will tend to decrease. After the two are combined, the change in frequency decrease may be offset by the rising wind speed or even return to zero, resulting in a smaller input fuzzy control deviation and a smaller output droop coefficient, which will weaken the frequency regulation effect. When the source end wind speed decreases, the frequency will fluctuate downward. At this time, load fluctuations will occur, and the frequency will also tend to decrease. After the two are combined, the change in frequency decrease will increase. In addition, the rotor kinetic energy of wind power is low when the wind speed is decreasing, and the input fuzzy control deviation increases, that is, the additional output power increases, which will lead to an overdraft of rotor kinetic energy, which will also weaken the frequency regulation effect.

[0006] In summary, existing technologies have shortcomings in considering the random fluctuation characteristics of source loads. They all ignore the turbulent characteristics of source-end wind speeds and are relatively general in their consideration of wind speed characteristics, failing to examine important wind speed indicators such as average wind speed and turbulence intensity. Therefore, this paper proposes a wind power fuzzy variable droop control strategy that considers the random fluctuation characteristics of source loads, thereby improving the stability of wind turbines and the quality of grid frequency under different random fluctuations of source loads. Summary of the Invention

[0007] The purpose of the present invention is to provide a wind power fuzzy variable droop control strategy that takes into account the random fluctuation characteristics of source and load to address the problem that the traditional wind power fuzzy variable droop control faces the problem of incorrect fuzzy control input under random fluctuation of source and load, resulting in insufficient frequency modulation output.

[0008] The purpose of the present invention is to be solved by the following technical solutions:

[0009] A fuzzy variable droop control strategy for wind power generation that considers the random fluctuation characteristics of source and load is proposed. This control strategy is constructed based on a mathematical model of the wind turbine, a low-order system frequency response model, a droop control strategy model, and a fuzzy control model. By simultaneously considering fluctuations at both the source and load ends when calculating the fuzzy control input, the strategy proposes the combined source and load variation as the input variable for fuzzy control. This allows the droop coefficient parameter for wind turbine frequency regulation to be adaptively adjusted with source and load fluctuations, reducing the mutual influence between the two and accurately reflecting the deviation of the input variable. During the wind power frequency regulation process, an improved fuzzy control method that considers source and load fluctuations is used to adjust the droop coefficient in real time. This enables the wind turbine to effectively respond to different wind conditions and load fluctuations, improving the frequency regulation effect under random source and load fluctuations.

[0010] The mathematical model of the wind turbine generator set adopted by the wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load provided by the present invention is:

[0011]

[0012] In formula (1)-formula (3), P m is the wind energy capture power of the wind turbine; ρ is the air density; R is the radius of the wind wheel; v is the real-time wind speed; C p is the wind energy utilization coefficient, β is the pitch angle, λ is the tip speed ratio, which is defined as λ = ωR / v; λ1 is the coefficient for calculating the tip speed ratio and pitch angle; ω is the wind turbine speed; H w is the inertia coefficient of the transmission system; s is the Laplace operator; F is the friction coefficient of the transmission system; T m is the mechanical torque of the wind turbine; T e is the electromagnetic torque of the wind turbine; P max K is the maximum active reference value under MPPT control;opt is the optimal wind capture coefficient; ω is the angular velocity of the wind turbine; is the maximum wind energy utilization coefficient; opt is the optimal tip speed ratio.

[0013] The low-order system frequency response (SFR) model adopted by the wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load provided by the present invention is:

[0014]

[0015]

[0016] In equations (4) and (5), Δf is the grid frequency deviation; H d is the inertia time constant of the power grid; s is the Laplace operator; D s is the damping coefficient of the power grid; P L is the real-time power at the load end; P e is the electromagnetic power output by the wind turbine; P M is the output power of the traditional synchronous generator set; K m is the mechanical power gain coefficient; F H is the power generated by the high-pressure steam turbine unit; T Re is the reheating time constant; R G is the droop coefficient of the synchronous generator.

[0017] The droop control strategy model adopted by the wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load provided by the present invention is:

[0018]

[0019] In formula (6), ΔP droop K is the droop additional power; d is the droop coefficient; f N is the rated frequency of the system; f is the actual operating frequency of the system; P e is the electromagnetic power output by the wind turbine; P max It is the maximum active reference value under MPPT control.

[0020] In power systems, the droop coefficient is a crucial parameter for wind turbines' participation in grid frequency regulation. It reflects the proportional relationship between the turbine's active power output and the grid frequency deviation. Its core significance lies in mimicking the frequency regulation characteristics of traditional synchronous generators, enabling wind turbines to proactively respond to frequency fluctuations and enhance the grid's dynamic stability. Specifically, when the grid frequency decreases, the wind turbine increases its active power output based on the droop coefficient; conversely, it reduces its output, thereby balancing power supply and demand and suppressing frequency deviation.

[0021] An excessively large droop factor can make wind turbines overly sensitive to frequency variations, potentially causing dramatic fluctuations in power output. For example, a slight frequency disturbance can trigger a significant output adjustment, exacerbating grid oscillations and even triggering a cascading failure. Frequent power adjustments can accelerate mechanical and electrical stresses on key components like converters and gearboxes, shortening equipment life. In one case, a wind farm experienced a 30% increase in converter failure rates due to an excessively high droop factor. Excessive frequency modulation can force turbines to operate off their optimal power curve, reducing wind energy capture efficiency.

[0022] A droop coefficient that is too small will weaken the unit's ability to respond to frequency changes, causing the grid to rely on traditional power sources for frequency regulation. In systems with high wind power penetration (e.g., wind power accounts for more than 40%), this may cause frequency deviations to exceed the national standard limit of ±0.5Hz. It may even fail to meet the requirements of new energy grid connection guidelines (such as GB / T 19963-2021), potentially forcing the wind farm to withdraw from the frequency regulation market, impacting operator profits. An excessively small coefficient is essentially a short-sighted act of sacrificing system stability for short-term power generation revenue. In today's new power system dominated by renewable energy, its harmful effects will be further amplified as penetration increases, and it must be avoided through refined control strategies.

[0023] In summary, the reasonable configuration of the droop coefficient is the key to wind power's participation in the active support of the power grid. It is necessary to strike a balance between system stability, equipment tolerance and economy, which is of great significance to the construction of a new power system.

[0024] Fuzzy control is a control method based on fuzzy mathematics. It mimics the way humans process uncertain information and describes the logical relationship between system inputs and outputs using fuzzy rules. This method first compiles operator or expert experience into fuzzy rules. It then fuzzifies real-time signals from sensors, uses the fuzzified signals as input to the fuzzy rules, performs fuzzy inference, and applies the resulting output to the actuators.

[0025] The fuzzy control model used in the wind power fuzzy droop control strategy provided by the present invention, which takes into account the random fluctuation characteristics of source and load, is: constructing a droop fuzzy control rule surface based on a two-dimensional fuzzy controller. The specific process of constructing the droop fuzzy control rule surface is as follows:

[0026] First, according to the experiment, the value range of error e is (-3, 3) and the value range of error rate ec is (-3, 3). Then the domain corresponding to error e and error rate ec on the fuzzy set is defined as: e, ec = {-3, -2, -1, 0, 1, 2, 3}, and the fuzzy subset corresponding to the domain of error e and error rate ec is: {NL, NM, NS, ZO, PS, PM, PL}. Using the triangular membership function, the relationship between the domain of error e and error rate ec and the fuzzy subset is obtained as follows: Figure 4 As shown, the fuzzy subsets of error e and error rate ec correspond one-to-one to language quantities, namely negative large, negative medium, negative small, zero, positive small, positive medium, and positive large. Then the corresponding relationships among the domain, fuzzy subset, and language quantity of error e and error rate ec are (-3, NL, negative large), (-2, NM, negative medium), (-1, NS, negative small), (0, ZO, zero), (1, PS, positive small), (2, PM, positive medium), and (3, PL, positive large).

[0027] Next, set the droop coefficient K d The value range is (0,0.6), then the droop coefficient K d The corresponding domain on the fuzzy set is defined as: K d ={0,0.1,0.2,0.3,0.4,0.5,0.6}, droop coefficient K d The fuzzy subset corresponding to the domain of discourse is also: {NL, NM, NS, ZO, PS, PM, PL}, and the droop coefficient K is obtained using the triangular membership function d The relationship between the domain of discourse and fuzzy subsets is as follows Figure 5 As shown, the droop coefficient K d The fuzzy subsets of the language quantities are negative large, negative medium, negative small, zero, positive small, positive medium, and positive large, respectively. The droop coefficient K d The corresponding relationships among the domain, fuzzy subset, and language quantity are (0, NL, negative large), (0.1, NM, negative medium), (0.2, NS, negative small), (0.3, ZO, zero), (0.4, PS, positive small), (0.5, PM, positive medium), and (0.6, PL, positive large).

[0028] Again, the droop coefficient K is set for the two-dimensional fuzzy controller as shown in Table 1 d The fuzzy rule table is used to obtain the corresponding fuzzy subset values ​​when the error e and error rate ec are -3, -2, -1, 0, 1, 2, and 3 respectively. The corresponding droop coefficient K is found based on the fuzzy subset values ​​of the error e and error rate ec. d The fuzzy subset values ​​are NL, NM, NS, ZO, PS, PM, PL, respectively, and the droop coefficient K d The fuzzy subset values ​​of the droop coefficient K are obtained in one-to-one correspondence dThe domain of the error e and error rate ec is 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6; that is, the domain of the error e and error rate ec and the corresponding fuzzy subset, droop coefficient K d Fuzzy rule table, droop coefficient K d The fuzzy subset and the corresponding domain can obtain the corresponding droop coefficient K when the error e and error rate ec take different domain values. d The domain value of ;

[0029] Finally, the error e is the X-axis, the error rate ec is the Y-axis, and the droop coefficient K d Construct a three-dimensional coordinate system for the Z axis, with the domain of the error e being the value of the X axis, the domain of the error rate ec being the value of the Y axis, and the corresponding droop coefficient K d The domain of discourse is the value of the Z axis. The coordinates of the corresponding points are determined in the three-dimensional coordinate system. The coordinates of adjacent points are connected to obtain a surface graph constructed by several small squares, such as Figure 6 The surface diagram of the variable droop fuzzy control rule is shown.

[0030] Figure 6 The surface diagram of the variable droop fuzzy control rule shown provides a regular surface observer. By observing the surface of the fuzzy control rule, the designer can easily check whether the designed fuzzy control rule table covers all possible situations. Only when the rule table is fully covered can it be guaranteed that the designed control rule can effectively play a control role.

[0031] The two-dimensional fuzzy controller used in this invention adopts Figure 6 The surface diagram of the variable droop fuzzy control rule is shown. When the specific values ​​of the error e and the error rate ec are input into the two-dimensional fuzzy controller, the droop coefficient K can be directly obtained. d The specific value of , thus showing the two-dimensional fuzzy controller for the droop coefficient K d display effect.

[0032] The specific steps of the wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load provided by the present invention are as follows:

[0033] S1. Initialization, setting the control period T and the number of iterations k = 1;

[0034] S2. using wind speed prediction to obtain the average wind speed and turbulence intensity of the current period;

[0035] S3. collecting the real-time wind speed, obtaining the real-time acceleration of the real-time wind speed, and correcting the real-time wind speed to obtain the corrected acceleration;

[0036] S4, obtaining the real-time source end power fluctuation amount based on step S2 and step S3;

[0037] S5. Obtaining real-time load-end power fluctuation according to the real-time operation status of the power grid or load forecast changes;

[0038] S6. Obtaining a source-load integrated change based on the source-end power fluctuation in step S4 and the load-end power fluctuation in step S5;

[0039] S7, the source load comprehensive change ΔP S As the error e, the error change rate ec is obtained by taking the derivative of the error e;

[0040] S8. Input the error e and error change rate ec as input variables directly into the second-order fuzzy controller to obtain the output variable droop coefficient update K d ';

[0041] S9, update the droop coefficient K d ′ Input the droop controller to obtain the droop additional power update amount ΔP droop ′, and then obtain the control instruction P of the wind turbine output electromagnetic power eref , and adopt the control instruction P of wind turbine output electromagnetic power eref Update the electromagnetic power P output by the wind turbine e , real-time control of wind turbines;

[0042] S10. Determine whether the control cycle is over. If the control cycle is over, set k=k+1 and return to step S2. If the control cycle is not over, return to step S3.

[0043] The present invention adopts the average wind speed as the main index to measure turbulent wind speed and turbulence intensity T i (k) to participate in the wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load. In this wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load, the average wind speed and turbulence intensity T i (k) Update according to the control cycle.

[0044] The wind speed forecast is used to obtain the average wind speed of the current period and turbulence intensity T i (k), the average wind speed in step S2 and turbulence intensity T i The formula for calculating (k) is:

[0045]

[0046] In formula (7)-formula (8), v(k) is the wind speed of the current cycle obtained by wind speed prediction; T is the duration of the control cycle; is the average wind speed of the current period; T i (k) is the turbulence intensity of the wind speed in the current period.

[0047] In order to simultaneously consider the dynamic characteristics and periodic characteristics of turbulent wind speed, the real-time acceleration of the real-time wind speed is selected to reflect the changing trend of the wind turbine power. When the real-time acceleration is negative, in order to prevent the wind turbine from overdrawing the rotor kinetic energy when the real-time wind speed decreases, the real-time acceleration needs to be corrected. The formula used for the real-time acceleration a of the real-time wind speed and the corrected acceleration a′ of the real-time wind speed in step S3 is:

[0048]

[0049] In formula (9) and formula (10), v is the real-time wind speed; a is the real-time acceleration of the real-time wind speed; and a′ is the corrected acceleration of the real-time wind speed.

[0050] The power fluctuation amount ΔP at the source end in step S4 wind The steps to obtain are:

[0051] S41. Calculate the average power fluctuation ΔP at the source end avg :

[0052]

[0053] In formula (11)-formula (12), P m is the wind energy capture power of the wind turbine; ρ is the air density; R is the radius of the wind rotor; v is the real-time wind speed; Δv is the average deviation of the real-time wind speed around the average wind speed; C p is the wind energy utilization coefficient, β is the pitch angle, λ is the tip speed ratio, which is defined as λ = ωR / v, ω is the wind turbine speed;

[0054] S42, calculate the average deviation value Δv of the real-time wind speed fluctuation around the average wind speed, and approximate the average deviation value Δv of the real-time wind speed fluctuation around the average wind speed to the standard deviation σ of the turbulent wind speed in the current cycle, that is, Δv≈σ, and since So we get:

[0055]

[0056] In formula (13), is the average wind speed of the current period; T i (k) is the turbulence intensity of the wind speed in the current period;

[0057] Substitute equations (12) and (13) into equation (11) and replace the average wind speed of the current period with Substituting the real-time wind speed v, we get:

[0058]

[0059] Simplifying formula (14) we can get:

[0060]

[0061] In formula (15), K1 is the gain coefficient of the average value of the power fluctuation at the source end;

[0062] S43, calculate the power fluctuation ΔP at the source end wind :

[0063] ΔP wind =ΔP avg ·a′ (16)

[0064] In formula (16), a′ is the corrected acceleration of the real-time wind speed;

[0065] The average power fluctuation ΔP at the source end in formula (15) avg Substituting into formula (16), we can get the power fluctuation ΔP at the source end: wind for:

[0066]

[0067] The load end power fluctuation ΔP in step S5 L It is directly obtained based on the real-time grid operation status or load forecast changes, and the formula for obtaining it is:

[0068] ΔP L =P L1 -P L0 (18)

[0069] In formula (18), ΔP L P is the load power fluctuation, which means the difference between two adjacent collected grid load power values; L0 is the power load value of the power grid collected last time; P L1 It is the power value of the power grid load collected last time.

[0070] The source-load comprehensive change ΔP in step S6 s The formula for obtaining is:

[0071] ΔP s =λ L ΔP L -(1+a)·ΔP wind (19)

[0072] In formula (19), λ L is the weight coefficient of load end power fluctuation, which is 2; ΔP L is the load end power fluctuation; ΔP windis the power fluctuation at the source end; a is the real-time acceleration of the real-time wind speed, ranging from [-1,1], in m / s. ( 1+a ) is the weight coefficient of the power fluctuation at the source end.

[0073] The value range of the error e in step S7 is (-3, 3), and the value range of the error rate ec is (-3, 3).

[0074] The droop coefficient update amount K in step S8 d The value range of ′ is (0,0.6).

[0075] The droop additional power update amount ΔP in step S9 droop The formula for obtaining ′ is:

[0076] ΔP droop ′=K d ′(f N -f) (20)

[0077] In formula (20), ΔP droop ′ is the droop additional power update amount; K d ′ is the updated value of droop coefficient; f N is the rated frequency of the system; f is the actual operating frequency of the system;

[0078] The control instruction P of the electromagnetic power output by the wind turbine in step S9 is eref The formula for obtaining is:

[0079] P eref =ΔP droop ′+P max (twenty one)

[0080] In formula (21), P eref P is the control instruction of electromagnetic power output by wind turbine; max It is the maximum active reference value under MPPT control.

[0081] The present invention has the following advantages over the prior art:

[0082] The wind power fuzzy variable droop control strategy method considering the random fluctuation characteristics of source and load of the present invention proposes a strategy for dynamically optimizing the droop coefficient of the wind turbine according to the source and load fluctuation situation. The strategy proposes the comprehensive change amount of source and load and provides a calculation method. Based on the comprehensive change amount of source and load as the input of fuzzy control, the droop coefficient parameter of the wind turbine frequency regulation is adaptively adjusted with the source and load fluctuation. The proposed strategy has good effectiveness and superiority in terms of frequency regulation effect and frequency stability.

[0083] The wind power fuzzy variable droop control strategy provided by the present invention takes into account the random fluctuation characteristics of source and load. Compared with the existing traditional droop control strategy and fuzzy variable droop control strategy, it not only takes into account the impact of turbulent wind speed fluctuations on wind turbines at non-frequency modulation moments, but also takes into account the dual pressure brought to wind turbines by the dual random fluctuations of source and load at frequency modulation moments; when calculating the fuzzy control input, it takes into account the fluctuations at both ends of the source and load, reduces the mutual influence between the two, and thus correctly reflects the deviation of the input variable; it can reasonably adjust the droop coefficient of the wind turbine under the dual random fluctuations of source and load, give full play to the frequency modulation potential of the wind turbine, and provide support for the frequency modulation of power grids with a high proportion of wind power. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Attachment Figure 1 This is a control block diagram of a wind power fuzzy variable droop control strategy that takes into account the random fluctuation characteristics of source and load provided by the present invention;

[0085] Attachment Figure 2 This is a control block diagram of a low-order system frequency response (SFR) model used in a wind power fuzzy variable droop control strategy that considers random source and load fluctuation characteristics provided by the present invention;

[0086] Attachment Figure 3 A flow chart of a wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load provided by the present invention;

[0087] Attachment Figure 4 A relationship diagram between the domain of error and error rate obtained based on the triangle membership function provided by the present invention and the fuzzy subset;

[0088] Attachment Figure 5 A relationship diagram between the domain of the droop coefficient obtained based on the triangle membership function provided by the present invention and the fuzzy subset;

[0089] Attachment Figure 6 The surface diagram of the variable droop fuzzy control rule provided by the present invention;

[0090] Attachment Figure 7 A comparison chart of simulation verification results of the frequency deviation index of the improved strategy of the present invention, the fixed coefficient droop control strategy, and the fuzzy variable droop control shown in the embodiment;

[0091] Attachment Figure 8 A comparison chart of simulation verification results of output power indicators of the improved strategy of the present invention, the fixed coefficient droop control strategy, and the fuzzy variable droop control shown in the embodiment;

[0092] Attachment Figure 9The figure shows the comparison of simulation verification results of the fan speed index of the improved strategy of the present invention, the fixed coefficient droop control strategy, and the fuzzy variable droop control shown in the embodiment. DETAILED DESCRIPTION

[0093] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0094] like Figure 1-3 As shown in the figure, a wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load is proposed. The control strategy is constructed based on the mathematical model of the wind turbine, the low-order system frequency response model, the droop control strategy model and the fuzzy control model. The control strategy considers the fluctuations of both ends of the source and load when calculating the fuzzy control input, and proposes the comprehensive source and load variation as the input variable of the fuzzy control, so that the droop coefficient parameters of the wind turbine frequency regulation are adaptively adjusted with the source and load fluctuations, which can enable the wind turbine to effectively cope with different wind conditions and load fluctuations, and improve the frequency regulation effect under random source and load fluctuations.

[0095] The mathematical model of the wind turbine generator set adopted by the wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load provided by the present invention is:

[0096]

[0097]

[0098] In formula (1)-formula (3), P m is the wind energy capture power of the wind turbine; ρ is the air density; R is the radius of the wind wheel; v is the real-time wind speed; C p is the wind energy utilization coefficient, β is the pitch angle, λ is the tip speed ratio, which is defined as λ = ωR / v; λ1 is the coefficient for calculating the tip speed ratio and pitch angle; ω is the wind turbine speed; H w is the inertia coefficient of the transmission system; s is the Laplace operator; F is the friction coefficient of the transmission system; T m is the mechanical torque of the wind turbine; T e is the electromagnetic torque of the wind turbine; P max K is the maximum active reference value under MPPT control; opt is the optimal wind capture coefficient; ω is the angular velocity of the wind turbine; is the maximum wind energy utilization coefficient; opt is the optimal tip speed ratio.

[0099] The low-order system frequency response (SFR) model adopted by the wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load provided by the present invention is:

[0100]

[0101] In equations (4) and (5), Δf is the grid frequency deviation; H d is the inertia time constant of the power grid; s is the Laplace operator; D s is the damping coefficient of the power grid; P L is the real-time power at the load end; P e is the electromagnetic power output by the wind turbine; P M is the output power of the traditional synchronous generator set; K m is the mechanical power gain coefficient; F H is the power generated by the high-pressure steam turbine unit; T Re is the reheating time constant; R G is the droop coefficient of the synchronous generator.

[0102] The droop control strategy model adopted by the wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load provided by the present invention is:

[0103]

[0104] In formula (6), ΔP droop K is the droop additional power; d is the droop coefficient; f N is the rated frequency of the system; f is the actual operating frequency of the system; P e is the electromagnetic power output by the wind turbine; P max It is the maximum active reference value under MPPT control.

[0105] like Figure 3 As shown, the specific steps of the wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load provided by the present invention are as follows:

[0106] S1. Initialization, setting the control period T and the number of iterations k = 1;

[0107] S2. Use wind speed prediction to obtain the average wind speed of the current period and turbulence intensity T i (k):

[0108]

[0109] In formula (7)-formula (8), v(k) is the wind speed of the current cycle obtained by wind speed prediction; T is the duration of the control cycle; is the average wind speed of the current period; T i (k) is the turbulence intensity of the wind speed in the current period;

[0110] S3. Collect the real-time wind speed v, obtain the real-time acceleration a of the real-time wind speed, and correct it to obtain the corrected acceleration a′ of the real-time wind speed:

[0111]

[0112] In formula (9)-formula (10), v is the real-time wind speed; a is the real-time acceleration of the real-time wind speed; a′ is the corrected acceleration of the real-time wind speed;

[0113] S4, based on step S2 and step S3, obtain the real-time source end power fluctuation ΔP wind , the specific steps are:

[0114] S41. Calculate the average power fluctuation ΔP at the source end avg :

[0115]

[0116] In formula (11)-formula (12), P m is the wind energy capture power of the wind turbine; ρ is the air density; R is the radius of the wind rotor; v is the real-time wind speed; Δv is the average deviation of the real-time wind speed around the average wind speed; C p is the wind energy utilization coefficient, β is the pitch angle, λ is the tip speed ratio, which is defined as λ = ωR / v, ω is the wind turbine speed;

[0117] S42, calculate the average deviation value Δv of the real-time wind speed fluctuation around the average wind speed, and approximate the average deviation value Δv of the real-time wind speed fluctuation around the average wind speed to the standard deviation σ of the turbulent wind speed in the current cycle, that is, Δv≈σ, and since So we get:

[0118]

[0119] In formula (13), is the average wind speed of the current period; T i (k) is the turbulence intensity of the wind speed in the current period;

[0120] Substitute equations (12) and (13) into equation (11) and replace the average wind speed of the current period with Substituting the real-time wind speed v, we get:

[0121]

[0122] Simplifying formula (14) we can get:

[0123]

[0124] In formula (15), K1 is the gain coefficient of the average value of the power fluctuation at the source end;

[0125] S43, calculate the power fluctuation ΔP at the source end wind :

[0126] ΔP wind =ΔP avg ·a′ (16)

[0127] In formula (16), a′ is the corrected acceleration of the real-time wind speed;

[0128] The average power fluctuation ΔP at the source end in formula (15) avg Substituting into formula (16), we can get the power fluctuation ΔP at the source end: wind for:

[0129]

[0130] S5. Obtain the real-time load-side power fluctuation ΔP according to the real-time operation of the power grid or the load forecast change L :

[0131] ΔP L =P L1 -P L0 (18)

[0132] In formula (18), ΔP L P is the load power fluctuation, which means the difference between two adjacent collected grid load power values; L0 is the power load value of the power grid collected last time; P L1 is the power value of the power grid load collected last time;

[0133] S6. When the wind turbine adopts the wind power fuzzy variable droop control strategy method considering the random fluctuation characteristics of the source load, based on the source end power fluctuation ΔP in step S4, wind and the load end power fluctuation ΔP in step S5 L Obtain the comprehensive change of source and load ΔP s :

[0134] ΔP s =λ L ΔP L -(1+a)·ΔP wind (19)

[0135] In formula (19), λ L is the weight coefficient of load end power fluctuation, which is 2; ΔP L is the load end power fluctuation; ΔP wind is the power fluctuation at the source end; a is the real-time acceleration of the real-time wind speed, ranging from [-1, 1], in m / s; (1+a) is the weight coefficient of the power fluctuation at the source end;

[0136] S7, the source load comprehensive change ΔP S As the error e, the error change rate ec is obtained by differentiating the error e. The value range of the error e is (-3, 3) and the value range of the error rate ec is (-3, 3);

[0137] S8. Input the error e and error change rate ec as input variables directly into the second-order fuzzy controller to obtain the output variable droop coefficient update K d ′, droop coefficient update amount K d The value range of ′ is (0,0.6);

[0138] S9, update the droop coefficient K d ′ Input the droop controller to obtain the droop additional power update amount ΔP droop ′:

[0139] ΔP droop ′=K d ′(f N -f) (20)

[0140] Then the control instruction P of the electromagnetic power output by the wind turbine is obtained eref :

[0141] P eref =ΔP droop ′+P max (twenty one)

[0142] S10. Determine whether the control cycle is over. If the control cycle is over, set k=k+1 and return to step S2. If the control cycle is not over, return to step S3.

[0143] Example

[0144] A specific embodiment is provided below to further illustrate the control effect of a wind power fuzzy variable droop control strategy taking into account the random fluctuation characteristics of source and load provided by the present invention.

[0145] 1. MATLAB simulation analysis

[0146] In order to verify the superiority and effectiveness of the improved strategy of the present invention, wind power participates in frequency modulation through droop control, and a comparative analysis is conducted on the fixed coefficient droop control, fuzzy variable droop control and the improved strategy of the present invention (source-load fuzzy variable droop control).

[0147] (1) Simulation parameters are shown in Table 2.

[0148] Table 2 Simulation parameters

[0149]

[0150] (2) Analysis of simulation results

[0151] The wind power ratio of the example scenario is set to 15%, the simulation time is 200s, the average wind speed is 8m / s, the turbulence intensity is level A, and the wind power participates in the frequency regulation through droop control. The simulation results are shown in Figure 7-9 The statistical results are shown in Table 3.

[0152] Table 3 Strategy comparison table

[0153]

[0154] It can be seen from the simulation data that the improved strategy of the present invention can effectively suppress the frequency fluctuations caused by turbulent wind speed disturbances at the source end during non-frequency modulation moments. When the wind speed enters an upward cycle, the improved strategy of the present invention correctly identifies the wind speed changes through the comprehensive change of the source and load, and adaptively adjusts the droop coefficient update amount. When entering a downward wind speed cycle, the improved strategy of the present invention can enable the wind turbine to effectively utilize the droop coefficient update amount dynamically adjusted by fuzzy control, avoid excessive overdraft of rotor kinetic energy, and thus improve the frequency stability when only considering source-end fluctuations.

[0155] Table 3 compares and summarizes the simulation data of different control strategies. It can be seen that the source-load fuzzy variable droop control strategy provided by the present invention has an improvement of 3.14% in average frequency deviation compared to the fixed coefficient droop control strategy. It performs well at the lowest point of the frequency modulation moment, an important frequency modulation indicator, with an improvement of 12.05% compared to the fixed coefficient droop control strategy. Compared with the fuzzy variable droop control strategy, the source-load fuzzy variable droop control strategy provided by the present invention still has an improvement of 1.87% at the lowest point of the frequency modulation moment while ensuring the average frequency deviation indicator, reflecting the superiority of the source-load fuzzy variable droop control strategy.

[0156] from Figure 7 The frequency deviation comparison chart shows that during frequency modulation, the source-load fuzzy variable droop control strategy provided by the present invention effectively responds to load-side fluctuations, compensating for the power shortage caused by load fluctuations and improving the lowest frequency point. During non-frequency modulation, the source-load fuzzy variable droop control strategy provided by the present invention also smooths frequency fluctuations caused by source-side fluctuations, effectively reducing the amplitude of the fluctuations and improving the frequency operation stability of the power grid.

[0157] from Figure 8 As can be seen, the source-load fuzzy variable droop control strategy provided by this invention can effectively identify power shortages caused by load-side fluctuations during frequency modulation, dynamically and adaptively increase the droop control coefficient, thereby allowing wind power to output more power and improve frequency modulation performance. Even during non-frequency modulation periods, source-side fluctuations can be identified and the droop coefficient can be adaptively modified to smooth source-side power fluctuations, thereby improving wind power's ability to regulate the grid.

[0158] The above embodiments are intended only to illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. Such modifications or replacements do not deviate from the essence of the corresponding technical solutions of the embodiments of the present invention. Technologies not covered by the present invention can be implemented using existing technologies.

Claims

1. A wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load, characterized by: The control strategy is based on the mathematical model of the wind turbine, the low-order system frequency response model, the droop control strategy model and the fuzzy control model. The specific steps are as follows: S1. Initialization, setting the control period T and the number of iterations k = 1; S2. Use wind speed prediction to obtain the average wind speed of the current period and turbulence intensity T i (k); S3, collecting the real-time wind speed v, obtaining the real-time acceleration a of the real-time wind speed, and correcting it to obtain the corrected acceleration a′ of the real-time wind speed; S4, based on step S2 and step S3, obtain the real-time source end power fluctuation ΔP wind ; S5. Obtain the real-time load-side power fluctuation ΔP according to the real-time operation of the power grid or the load forecast change L ; S6, based on the source end power fluctuation ΔP in step S4 wind and the load end power fluctuation ΔP in step S5 L Obtain the comprehensive change of source and load ΔP s ; S7, the source load comprehensive change ΔP S As the error e, the error change rate ec is obtained by taking the derivative of the error e; S8. Input the error e and error change rate ec as input variables directly into the second-order fuzzy controller to obtain the output variable droop coefficient update K d '; S9, update the droop coefficient K d ′ Input the droop controller to obtain the droop additional power update amount ΔP droop ′, and then obtain the control instruction P of the wind turbine output electromagnetic power eref , and adopt the control instruction P of wind turbine output electromagnetic power eref Update the electromagnetic power P output by the wind turbine e , real-time control of wind turbines; S10. Determine whether the control cycle is over. If the control cycle is over, set k=k+1 and return to step S2. If the control cycle is not over, return to step S3.

2. The wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load according to claim 1 is characterized by: The mathematical model of the wind turbine generator set is: In formula (1)-formula (3), P m is the wind energy capture power of the wind turbine; ρ is the air density; R is the radius of the wind wheel; v is the real-time wind speed; C p is the wind energy utilization coefficient, β is the pitch angle, λ is the tip speed ratio, which is defined as λ = ωR / v; λ1 is the coefficient for calculating the tip speed ratio and pitch angle; ω is the wind turbine speed; H w is the inertia coefficient of the transmission system; s is the Laplace operator; F is the friction coefficient of the transmission system; T m is the mechanical torque of the wind turbine; T e is the electromagnetic torque of the wind turbine; P max K is the maximum active reference value under MPPT control; opt is the optimal wind capture coefficient; ω is the angular velocity of the wind turbine; is the maximum wind energy utilization coefficient; opt is the optimal tip speed ratio; The low-order system frequency response model is: In equations (4) and (5), Δf is the grid frequency deviation; H d is the inertia time constant of the power grid; s is the Laplace operator; D s is the damping coefficient of the power grid; P L is the real-time power at the load end; P e is the electromagnetic power output by the wind turbine; P M is the output power of the traditional synchronous generator set; K m is the mechanical power gain coefficient; F H is the power generated by the high-pressure steam turbine unit; T Re is the reheating time constant; R G is the droop coefficient of the synchronous generator; The droop control strategy model is: In formula (6), ΔP droop K is the droop additional power; d is the droop coefficient; f N is the rated frequency of the system; f is the actual operating frequency of the system; P e is the electromagnetic power output by the wind turbine; P max It is the maximum active reference value under MPPT control.

3. The wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load according to claim 1 or 2 is characterized by: The average wind speed in step S2 and turbulence intensity T i The formula for calculating (k) is: In formula (7)-formula (8), v(k) is the wind speed of the current cycle obtained by wind speed prediction; T is the duration of the control cycle; is the average wind speed of the current period; T i (k) is the turbulence intensity of the wind speed in the current period; σ is the standard deviation of the turbulent wind speed.

4. The wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load according to claim 1 or 2 is characterized by: The formulas used for the real-time acceleration a of the real-time wind speed and the corrected acceleration a′ of the real-time wind speed in step S3 are: In formula (9) and formula (10), v is the real-time wind speed; a is the real-time acceleration of the real-time wind speed; and a′ is the corrected acceleration of the real-time wind speed.

5. The wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load according to claim 1 or 2 is characterized by: The power fluctuation amount ΔP at the source end in step S4 wind The steps to obtain are: S41. Calculate the average power fluctuation ΔP avg : In formula (11)-formula (12), P m is the wind energy capture power of the wind turbine; ρ is the air density; R is the radius of the wind rotor; v is the real-time wind speed; Δv is the average deviation of the real-time wind speed around the average wind speed; C p is the wind energy utilization coefficient, β is the pitch angle, λ is the tip speed ratio, which is defined as λ = ωR / v, ω is the wind turbine speed; S42, calculate the average deviation value Δv of the real-time wind speed fluctuation around the average wind speed, and approximate the average deviation value Δv of the real-time wind speed fluctuation around the average wind speed to the standard deviation σ of the turbulent wind speed in the current cycle, that is, Δv≈σ, and since So we get: In formula (13), is the average wind speed of the current period; T i (k) is the turbulence intensity of the wind speed in the current period; Substitute equations (12) and (13) into equation (11) and replace the average wind speed of the current period with Substituting the real-time wind speed v, we get: Simplifying formula (14) we can get: In formula (15), K1 is the gain coefficient of the average value of the power fluctuation at the source end; S43, calculate the power fluctuation ΔP at the source end wind : ΔP wind =ΔP avg ·a′ (16) In formula (16), a′ is the corrected acceleration of the real-time wind speed; The average power fluctuation ΔP at the source end in formula (15) avg Substituting into formula (16), we can get the power fluctuation ΔP at the source end: wind for:

6. The wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load according to claim 1 or 2 is characterized by: The load end power fluctuation ΔP in step S5 L The formula for obtaining is: ΔP L =P L1 -P L0 (18) In formula (18), ΔP L P is the load power fluctuation, which means the difference between two adjacent collected grid load power values; L0 is the power load value of the power grid collected last time; P L1 It is the power value of the power grid load collected last time.

7. The wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load according to claim 1 or 2 is characterized by: The source-load comprehensive change ΔP in step S6 s The formula for obtaining is: ΔP s =λ L ΔP L -(1+a)·ΔP wind (19) In formula (19), λ L is the weight coefficient of the power fluctuation at the source end, which is 2; ΔP L is the load end power fluctuation; ΔP wind is the power fluctuation at the source end; a is the real-time acceleration of the real-time wind speed, ranging from [-1, 1], in m / s; (1+a) is the weight coefficient of the power fluctuation at the source end.

8. The wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load according to claim 1 or 2 is characterized by: The value range of the error e in step S7 is (-3, 3), and the value range of the error rate ec is (-3, 3).

9. The wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load according to claim 1 or 2 is characterized by: The droop coefficient update amount K in step S8 d The value range of ′ is (0,0.6).

10. The wind power fuzzy variable droop control strategy considering the random fluctuation characteristics of source and load according to claim 1 or 2, characterized in that: The droop additional power update amount ΔP in step S9 droop The formula for obtaining ′ is: ΔP droop ′=K d ′(f N -f) (20) In formula (20), ΔP droop ′ is the droop additional power update amount; K d ′ is the updated value of droop coefficient; f N is the rated frequency of the system; f is the actual operating frequency of the system; The control instruction P of the electromagnetic power output by the wind turbine in step S9 is eref The formula for obtaining is: P eref =ΔP droop ′+P max (21) In formula (21), P eref P is the control instruction of electromagnetic power output by wind turbine; max It is the maximum active reference value under MPPT control.

Images