Wind turbine generator active frequency response method considering wind-fire power coordination

Abstract

The invention provides a wind turbine generator active frequency response method considering wind-fire power coordination, which comprises the following steps: establishing an improved SFR model considering the dynamic state of a speed regulator and high-proportion wind power, and solving the improved SFR model to obtain a time domain analytical solution of system frequency response; constructing a power track of a wind turbine generator frequency support stage by considering wind-fire power cooperation and frequency response peak-valley complementation under wind power frequency modulation power and load power disturbance, and solving parameters by using a time domain analytical solution of system frequency response; and designing a power track of the wind turbine generator frequency support quit stage based on a Logistic function. According to the method, the lowest point of system frequency drop under power disturbance can be improved, low-frequency load shedding can be prevented from being triggered or low-frequency load shedding turns can be reduced, and power supply reliability is guaranteed; according to the method, the initial frequency change rate can be improved, the off-network risk of distributed energy sources can be reduced, more online power generation capacity can be maintained, and cascading failures can be prevented; and popularization and application of the governor dynamic SFR model can be considered.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wind power frequency modulation, and particularly relates to a wind turbine active frequency response method considering wind-fire power coordination. BACKGROUND

[0002] The active frequency response control strategy is an important technical means to improve the primary frequency modulation capability of the wind turbine. However, the existing researches mainly focus on the design of the response mechanism of the wind turbine itself, and ignore the influence of the coordinated output of multiple energy sources on the frequency response characteristics. Under the promotion of the "double carbon" strategy, the proportion of wind power in the power structure is rising. However, the large-scale grid connection of new energy represented by wind power has caused serious challenges to the frequency stability of the power system. Therefore, it is particularly important to develop a reasonable wind power frequency modulation strategy to enable the wind power to provide frequency modulation service.

[0003] Currently, the wind power frequency modulation strategy can be divided into rotor kinetic energy control and power reserve control according to the energy source. Rotor kinetic energy control includes closed-loop control such as virtual inertia control, droop control, and comprehensive inertia control, as well as open-loop control such as step inertia control and constant slope control. Although the above rotor kinetic energy control has achieved certain improvement on system frequency dynamics, it does not consider the coordination of wind power and thermal power active power output. The literature [Liu J Z, Yao Q, Liu Y, et al. Research on primary frequency control strategy of wind farm in wind-thermal combined dispatching [J]. Proceedings of the CSEE, 2017, 37(12): 3462-3469.] proposed the concept of combined primary frequency control, which takes thermal power frequency modulation as the main and wind power frequency modulation as the auxiliary. The literature [Sun M, Sun Y, Chen L, et al. Novel temporary frequency support control strategy of wind turbine generator considering coordination with synchronous generator [J]. IEEE Transactions on Sustainable Energy, 2022, 13(2): 1011-1020.] established an optimal control function optimization model considering the coordination of wind turbine generators and synchronous generators. The literature [Xu G, Liu F, Hu J, et al. Coordination of wind turbines and synchronous generators for system frequency control [J]. Renewable energy, 2018, 129: 225-236.] realized coordinated frequency modulation by introducing wind power frequency modulation power into synchronous generators. The literature [Sun M, Min Y, Chen L, et al. Optimal auxiliary frequency control of wind turbine generators and coordination with synchronous generators [J]. CSEE Journal of Power and Energy Systems, 2021, 7(1): 78-85.] showed that appropriate slowing down of the response speed of wind turbine generator droop control can improve the frequency secondary drop depth.The documents [Xiang Z. Considering Delayed Support for Primary Frequency Control Strategy of Wind Turbines[D]. Nanjing University of Technology, 2021.] and [Gu W, Chen Z, Li Q, et al. Torque Limit-based Inertial Control Method Based on Delayed Support for Primary Frequency Control of Wind Turbines[J]. Journal of Modern Power Systems and Clean Energy, 2024, 12(2): 561-570.] discuss the influence of wind turbine delayed response on the frequency primary drop valley under various control strategies. Existing research reveals the potential of wind-fire cooperation in improving the frequency stability of power systems. However, current research on wind and fire response coordination is still limited, and existing research mainly focuses on the effect of wind-fire cooperation in reducing the frequency drop depth, while ignoring its important role in reducing the frequency change rate at the initial disturbance.

[0004] Improper design of the control strategy for the frequency support exit stage of wind turbines will lead to the problem of frequency secondary drop (FSD). In the literature [Ullah N R, Thiringer T, Karlsson D. Temporary primary frequency control support by variable speed wind turbines—Potential and applications[J]. IEEE Transactions on Power Systems, 2008, 23(2): 601-612.] and [Kang M, Kim K, Muljadi E, et al. Frequency control support of a doubly-fed induction generator based on the torque limit[J]. IEEE Transactions on Power Systems, 2016, 31(6): 4575-4583.], the FSD problem is solved by reducing the drop power. Although the FSD phenomenon is alleviated to some extent, it cannot fundamentally solve the problem. In the literature [Tang Y F, Yang P, Yang Y. Frequency response control strategy of wind turbines considering frequency secondary drop[J]. Power System Automation, 2023, 47(09): 166-174.], a linear exit scheme of wind turbines is designed, which effectively solves the FSD problem. However, the system has a secondary disturbance of frequency in the exit stage, and the speed recovery is slow. In the literature [Tao Y K, Yang F F, He P, et al. Flexible frequency response strategy of wind turbines considering frequency secondary drop and speed recovery[J]. Power System Automation, 2024, 48(13): 60-68.], the Logistic function is applied to design the output of wind turbines in the exit stage. The proposed scheme solves the FSD problem and has a faster speed recovery. However, this scheme only considers the wind power penetration rate and does not take into account the important factor of wind speed.

[0005] System frequency response (SFR) model is one of the important tools for studying the active frequency response of wind turbines. Existing researches are often based on the classical SFR model. However, simulation studies in the literature [Shi Q, Li F, Cui H. Analytical method to aggregate multi-machine SFR model with applications in power system dynamic studies [J]. IEEE Transactions on Power Systems, 2018, 33(6): 6355-6367.] show that the classical SFR model will produce more than 10% error compared with the full-order model. The literature [Ju Ping, Qin Chuan, Luo Jianyu. Modern power system frequency safety analysis and control [M]. Beijing: Science Press, 2022.] studies the SFR model based on measured data using identification method. The research results show that compared with the classical SFR model, the improved SFR model containing the dynamic of the governor can significantly improve the accuracy of the system frequency response curve, and ensure that the model parameters are in a reasonable range of values. SUMMARY

[0006] In view of the technical problems that the traditional frequency modulation strategy of wind turbines cannot realize the improvement of the initial frequency change rate and the lowest point of the first frequency drop, and the limitation in the frequency secondary drop governance, the present application proposes a wind turbine active frequency response method considering wind-fire power coordination, based on the improved SFR model, the frequency support scheme designed by the wind-fire power coordination and the frequency response component peak-valley complementary principle realizes the comprehensive improvement of the initial frequency change rate and the first frequency drop depth; the frequency support exit scheme considering wind speed and wind power penetration rate can realize the improvement of the frequency secondary drop problem, in summary, the proposed strategy can realize the improvement of the system frequency drop lowest point (the smaller one in the first drop and the secondary drop valley value) under power disturbance, which is beneficial to prevent the triggering of low-frequency load shedding or reduce the number of low-frequency load shedding, avoid or reduce user power outage, and ensure power supply reliability; at the same time, reduce the risk of distributed energy off-line, maintain more power generation capacity online, and prevent the occurrence of cascading failures.

[0007] In order to achieve the above purpose, the technical scheme of the present application is as follows: a wind turbine active frequency response method considering wind-fire power coordination, comprising the following steps:

[0008] Step 1: Establish an improved SFR model considering the dynamic of the governor and high proportion of wind power, and obtain the time-domain analytical solution of the system frequency response by solving the improved SFR model;

[0009] Step two: considering the wind and thermal power synergy and the frequency response peak valley complementation of wind power frequency modulation power and load power disturbance, the power trajectory of the wind turbine frequency support stage is constructed, and the time domain analytical solution of the system frequency response is used to solve the parameters;

[0010] Step three: the power trajectory of the wind turbine frequency support exit stage is designed based on the Logistic function.

[0011] Preferably, the transfer function of the speed governor dynamics is described as a first-order inertia link: In the formula, T G is the inertia time constant of the speed governor, the equivalent regulation coefficient 1 / R=K m / R′, K m is the mechanical power gain coefficient, 1 / R′ is the regulation coefficient, and s is the Laplace operator.

[0012] The transfer function of the improved SFR model from the power disturbance input ΔP(s) to the frequency change output Δf(s) is:

[0013]

[0014] In the formula: ΔP(s) represents the disturbance power ΔP L or the frequency modulation power ΔP W of the wind turbine, D is the system damping coefficient, and ρ is the wind power penetration level; H SYS is the equivalent inertia time constant of the system; D is the system damping coefficient; F H is the high-pressure cylinder work proportion of the thermal power unit; T R is the reheating time constant of the thermal power unit.

[0015] Preferably, the time domain analytical solution of the system frequency response is:

[0016]

[0017] In the formula:

[0018]

[0019] In the formula: f1, f2, f3, δ, ζ, η, b1, b2, b3, Ω, x, A m are intermediate coefficients.

[0020] Preferably, the solving method of the time domain analytical solution of the system frequency response is that the transfer function G( S ) of the improved SFR model is abbreviated as:

[0021]

[0022] Wherein: d0, d1, d2, e0, e1, e2, e3 are intermediate coefficients, and

[0023]

[0024] Considering the system disturbance power in the form of step ΔP(t) = ΔP A ε(t), ΔP A represents the disturbance amplitude, and ε(t) is a unit step function; the Laplace transform is performed to obtain: When the disturbance occurs, the system frequency response Δf(s) is:

[0025]

[0026] The characteristic equation of the transfer function G( S ) is a cubic parameter equation, and the equation is solved by using the Cardano formula; the time-domain expression of the system frequency response Δf(t) is obtained by performing partial fraction expansion and Laplace inverse transform on the system frequency response Δf(s).

[0027] Preferably, the expected power trajectory on the P-ω plane in the frequency support stage is an arc segment from an initial operating point A to a support power peak point B', and a segment from the support power peak point B' to a critical point C' for exiting the support stage, wherein the point A corresponds to an initial rotational speed ω0 and an initial power P0, the point B' corresponds to a rotational speed ω B' and a power P0+ ΔP0, and the point C' corresponds to a rotational speed ω C' and a power P0.

[0028] In the frequency support stage, the wind turbine output active power P W linearly increases with time to a support power peak value at a time t B′ , and the wind turbine output active power P B′ linearly decreases after the time t W .

[0029] Preferably, the expression of the wind turbine output active power P W is:

[0030]

[0031] Wherein: d is the power rising segment length, t0 is the frequency support starting time, t C′ is the frequency support ending time, at which the wind turbine output active power is equal to the initial power P0; ΔP0 is the initial power increment, P MPPT (ω0) is the output power corresponding to the initial rotational speed ω0 in the maximum power tracking operating mode, and P MPPT (ω0) = P0, the wind turbine output active power P W[ω(t)] is a functional of the rotational speed ω(t), which is a function of time t.

[0032] Preferably, in the P-ω plane, the wind turbine output active power P W [ω(t)] is expressed as:

[0033]

[0034] wherein Δω is the rotational speed variation amount generated in the wind power frequency modulation power rising stage, P MPPT (ω min ) is the wind power corresponding to the minimum rotational speed in the MPPT mode;

[0035] The wind turbine mechanical power P m is ignored, and the power rising stage has the following relationship:

[0036]

[0037] wherein H W is the wind turbine inertia time constant;

[0038] The calculation formula of the rotational speed variation amount Δω is obtained by arrangement:

[0039]

[0040] The power rising stage duration d is designed according to the complementary concept of the frequency response peak and valley under the wind power frequency modulation power and the load power disturbance.

[0041] Preferably, the method for solving the parameters by using the time domain analytical solution of the system frequency response is:

[0042] In order to obtain the system frequency response peak time generated by the wind power increase, the wind turbine power variation amount ΔP W in the frequency support stage is approximately processed:

[0043]

[0044] wherein, represents the approximate power variation amount of the wind turbine, u(t) is a unit step function, K r is the slope of the descending segment , and t off is the end time of the descending segment .

[0045] Based on the consideration that the system frequency peak time generated by the wind power variation lags behind but is close to the wind turbine mechanical power P W peak time, the power variation descending segment adopts a linear approximation.

[0046] The slope Kr is equal to the arc segment of the power-time curve At t B′ , the tangent slope is:

[0047]

[0048] P′ W is the power of the arc segment , K P-ω is the slope of the wind turbine output power curve in the P-ω plane, which is derived from the rotor motion equation;

[0049] The approximate power change of the wind turbine under the action of the system frequency response is:

[0050]

[0051] h ramp (t) is the unit ramp response function of the transfer function G(s);

[0052] Since the frequency response peak time lags behind the t B′ time, two cases are analyzed: t off ≤t off and t>t ramp .

[0053]

[0054] Taking the derivative of the above equation gives:

[0055]

[0056] Since h′ step (t) = h step (t), h step (t) is the unit step response function of the transfer function G(s), the above equation is rewritten as:

[0057]

[0058] h A (t) = Δf(t) / ΔP W , let the derivative Δf′ MPPT (t) = 0 of the time-domain system frequency response change, and take the frequency peak time generated by the wind power increase to be equal to the valley time of the system frequency response under the action of the load power disturbance, to obtain the value of the power rise segment d.

[0059] Preferably, the active power output of the wind turbine in the frequency support exit stage is designed as:​

[0060]

[0061] In the formula: P MPPT is the wind power in the maximum power tracking mode, k1 is a participation factor based on the Logistic function, b is a shape coefficient, δP W is the exit rotating speed ω C′ corresponding to the power step, the power step δP W shows a linear characteristic with the wind speed, P MPPT (ω C′ ) represents the maximum power tracking output power corresponding to the rotating speed at the end of the frequency support, and b is a shape coefficient.

[0062] Preferably, the value of the shape coefficient b monotonically decreases with the increase of the penetration rate and the wind speed, and the shape coefficient b is:

[0063]

[0064] In the formula: a, k2, and k3 are design rule coefficients, v W is the wind speed, and ρ is the wind power penetration rate; the wind speed reference value v base and the penetration rate reference value ρ base are selected near the middle values in the respective intervals.

[0065] Compared with the prior art, the present application has the following beneficial effects: firstly, the improved system frequency response model is established by considering the wind power frequency modulation effect and the governor dynamics, and the system frequency response time domain analytical solution under power disturbance is derived. Compared with the classic SFR model, the improved SFR model expands the power supply type (from pure thermal power units to coexistence of thermal power and wind power) and improves the accuracy of the model. The system derivation of the SFR model frequency response time domain analytical solution considering the governor dynamics makes up for the deficiency of the existing literature in the aspect of the analytical solution of the model; secondly, the power trajectory model of the wind turbine frequency support stage, i.e., the short-time frequency support scheme, is constructed by following the design principle of complementary peak and valley of wind and fire power coordination and frequency response components, and compared with the traditional peak and valley matching compensation wind and fire coordination strategy, the proposed strategy shows smaller initial frequency change rate and higher first drop minimum point; in addition, the influence of the wind speed and the wind power penetration rate is comprehensively considered, the Logistic function is introduced to adaptively design the power trajectory of the frequency support exit stage, i.e., the adaptive support exit scheme, which improves the system frequency secondary drop problem and the wind turbine itself has a faster rotating speed recovery speed; finally, the feasibility and effectiveness of the proposed strategy in improving the frequency dynamic characteristics are verified by simulation. The present application has the following advantages:

[0066] (1) Considering the wind power frequency modulation effect and the improvement of the SFR model and the time domain analytical solution of system frequency response, on the one hand, the SFR model improves the power supply type in the classical model from single thermal power unit to wind-thermal coexistence power supply structure, and improves the reproduction accuracy of the SFR model to the frequency response; On the other hand, the system derives the time domain analytical solution of the frequency response of the 3-order SFR model.

[0067] (2) The short-term frequency support strategy based on the design concept of wind-thermal power coordination and peak-valley complementation realizes the comprehensive improvement of the initial frequency change rate and the frequency first drop depth. The initial frequency change rate is closely related to the grid protection of distributed energy, reducing the initial frequency change rate will reduce the risk of distributed energy off-grid, maintain more power generation capacity online, and prevent the occurrence of cascading failures.

[0068] (3) The support exit scheme based on the design of Logistic function has the characteristics of wind speed and wind power penetration self-adaptation, which can effectively reduce the FSD depth under different wind power penetration and wind speed conditions, and takes into account the demand of wind turbine speed rapid recovery to the initial operating point. Combined with advantage (2), the present application can realize the improvement of the system frequency drop minimum point (the larger one in the first drop and the second drop) under power disturbance. The frequency minimum point is closely related to the low-frequency load shedding starting value, and improving the frequency minimum point is beneficial to prevent triggering low-frequency load shedding or reducing low-frequency load shedding rounds, avoiding or reducing user power outage, and ensuring power supply reliability.

[0069] The present application integrates the improvement of classical SFR model, the derivation of system frequency time domain analytical solution, the short-term frequency support of wind turbine and the design of frequency second drop suppression strategy, which is helpful to the further popularization and application of SFR model considering the governor dynamics, and is helpful to the improvement of the system frequency drop minimum point and the initial frequency change rate under power disturbance. BRIEF DESCRIPTION OF DRAWINGS

[0070] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating laborious work.

[0071] Figure 1 The flowchart of the present application.

[0072] Figure 2 The principle diagram of the improved SFR model of the power system containing high proportion of wind power of the present application.

[0073] Figure 3 The comparison diagram of the influence of the governor on the system frequency response.

[0074] Figure 4 A comparison chart of the active frequency response power trajectory of the present application and the traditional strategy, wherein (a) is the power-speed relationship, and (b) is the power-time relationship.

[0075] Figure 5 A step amount δP of the present application W A relationship curve between the wind speed and the wind turbine output power.

[0076] Figure 6 A simulation topology chart of a 4-machine 2-area power system.

[0077] Figure 7 A simulation result comparison chart under scenario 1, wherein (a) is the system frequency, (b) is the wind turbine speed, (c) is the wind turbine output power, and (d) is the total wind and thermal output power.

[0078] Figure 8 A simulation result comparison chart under scenario 2, wherein (a) is the system frequency, (b) is the wind turbine speed, (c) is the wind turbine output power, and (d) is the total wind and thermal output power.

[0079] Figure 9 A simulation result comparison chart under scenario 3, wherein (a) is the system frequency, (b) is the wind turbine speed, (c) is the wind turbine output power, and (d) is the total wind and thermal output power.

[0080] Figure 10 A comparison chart of the actual frequency response and the SFR model output.

[0081] Figure 11 A simulation result comparison chart under scenario 4, wherein (a) is the system frequency, (b) is the wind turbine speed, and (c) is the wind turbine output power. DETAILED DESCRIPTION

[0082] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the present application.

[0083] As shown in Figure 1 A wind turbine active frequency response method considering wind-thermal power coordination, comprising the following steps:

[0084] Step 1: An improved SFR model considering the governor dynamics and high proportion of wind power is established, and a time-domain analytical solution of the system frequency response is obtained by solving the improved SFR model.

[0085] The time constant of the governor of thermal power units has a great influence on the valley of the frequency primary drop curve, but has no influence on the maximum frequency change rate and the steady-state frequency deviation. The classic SFR model, which does not consider the dynamic of the governor, underestimates the primary drop depth and the valley time. Based on the classic SFR model, an improved SFR model considering the dynamic of the governor and high penetration of wind power is established. The time-domain analytical solution of the system frequency response is obtained by solving the high-order model, which provides a theoretical basis for designing the active frequency response strategy of wind power units based on wind-thermal power coordination. The influence of the dynamic of the governor is studied in the s-domain and the time-domain based on the improved SFR model.

[0086] The classic SFR model uses the static regulation coefficient to model the governor of thermal power units, ignoring the dynamic process of the governor. The equivalent governor dynamic model G gov (s) can be described by a first-order inertia link:

[0087]

[0088] In the formula: T G is the inertia time constant of the governor, the equivalent regulation coefficient 1 / R = K m / R', K m is the mechanical power gain coefficient, 1 / R' is the regulation coefficient, and s is the Laplace operator.

[0089] The improved SFR model of the power system composed of high-penetration wind power units and thermal power units is shown in Figure 2 , in which the dynamic model of the governor is introduced in the feedback path of the thermal power unit. In the figure: ΔP L is the disturbance power, and ΔP L > 0 when the power generation increases, and ΔP L < 0 when the load increases; ΔP w is the frequency modulation power of the wind power unit; ΔP T is the frequency modulation power of the thermal power unit; ρ is the wind power penetration level; H SYS is the equivalent inertia time constant of the system; D is the system damping coefficient; F H is the high-pressure cylinder work proportion of the thermal power unit; T R is the reheating time constant of the thermal power unit.

[0090] As shown in Figure 2 , the transfer function of the improved SFR model from the power disturbance input ΔP(s) to the frequency change output Δf(s) is defined as:

[0091]

[0092] In the formula: ΔP(s) represents the disturbance power ΔP LOr the frequency modulation power ΔP of the wind turbine W .

[0093] The formula (2) can be simply written as:

[0094]

[0095] In the formula, d0, d1, d2, e0, e1, e2 and e3 are intermediate coefficients,

[0096]

[0097] The step form of disturbance is the most common and has the greatest impact in the power system, therefore, the present application considers the step form of system disturbance power, i.e. ΔP(t) = ΔP A ε(t), ΔP A represents the disturbance amplitude, and ε(t) is a unit step function. The Laplace transform is carried out as follows:

[0098]

[0099] Then, when the disturbance occurs, the system frequency response Δf(s) is:

[0100]

[0101] In the formula, 1 / s is the Laplace transform of the unit step disturbance.

[0102] It is known from the formula (3) that the characteristic equation of the transfer function G( S ) is a cubic parameter equation, and the equation can be solved by using the Cardano formula. The partial fraction expansion and the inverse Laplace transform are carried out on the frequency domain system frequency response Δf(s) i.e. the formula (5), and the time domain expression of the system frequency response can be obtained as shown in the formula (6).

[0103] The frequency domain system frequency response is:

[0104] In the formula, the intermediate coefficients are

[0105]

[0106] The characteristic equation of the transfer function G( S ) is H1(s) = e0+ e1s+ e2s 2 + e3s 3 , and the discriminant of the characteristic root is:

[0107] delta = (q / 2) 2 + (p / 3) 3

[0108] In the formula, the intermediate coefficients are

[0109]

[0110] The Cardano formula is used to solve the characteristic equation H1(s). When the discriminant delta < 0, there are three different real roots; when the discriminant delta = 0, there are three real roots. When the intermediate coefficients p and q are not equal to 0, there are two real roots and one single root; when the discriminant delta > 0, there is one real root and two conjugate complex roots. In summary, the characteristic equation H1(s) has at least one real root in all cases. Considering the intermediate coefficient e3 = 1, the characteristic equation H1(s) can be rewritten as:

[0111] H1(s) = e0+ e1s + e2s 2 +s 3 = (s + δ) (s 2 + ζs + η)

[0112]

[0113] Since e0≠ 0 and δ ≠ 0, we have

[0114]

[0115] Expansion:

[0116] In the formula, the intermediate coefficient

[0117]

[0118] The expanded is brought into the frequency domain system frequency response, and the Laplace inverse transform is performed to obtain the time domain expression of the system frequency response:

[0119]

[0120] In the formula:

[0121]

[0122] In the formula: f1, f2, f3, δ, ζ, η, b1, b2, b3, Ω, x, A m are intermediate coefficients.

[0123] The indexes for evaluating the frequency response dynamics of the power system after active disturbance mainly include three: the lowest point of system frequency Δf max , the maximum frequency change rate df / dt| max , and the steady-state frequency deviation Δf setBased on the above indicators, the influence of the speed regulator time constant on the system frequency response is studied. The maximum frequency change rate and steady-state frequency deviation of the system can be obtained by the Laplace transform initial value theorem and final value theorem respectively:

[0124]

[0125]

[0126] Analyzing formula (7), we can get that the maximum frequency change rate of the system is only related to the disturbance power amplitude ΔP A size, wind power penetration level ρ and system equivalent inertia time constant H SYS It is related to the inertia time constant of the speed regulator. From formula (8), we can see that the steady-state frequency offset Δf set Only with the disturbance power amplitude ΔP A The size of the speed regulator is related to the equivalent regulation coefficient 1 / R of the speed regulator, the wind power penetration level ρ and the system damping coefficient D, but has nothing to do with the inertia time constant of the speed regulator.

[0127] Considering that power shortage poses a more severe challenge to the power system than power surplus, this paper studies the system frequency stability under the condition of power shortage caused by load increase. In the improved SFR model that introduces the speed regulator dynamic model, the time when the system frequency reaches the lowest point under the action of step power is It is closely related to the time constant of the speed regulator, which can be obtained by setting the first-order derivative of formula (6) equal to 0. In order to simplify the calculation, the present invention conducts a qualitative analysis of the lowest frequency point through numerical simulation. The simulation results of the classic SFR model and the improved SFR model are shown in the figure below. Figure 3 As shown. Figure 3 It can be seen that the improved SFR model that considers the speed regulator dynamics has an earlier frequency minimum point and a smaller valley value. This shows that the classic SFR model underestimates the maximum frequency drop depth.

[0128] Step 2: Considering the synergy between wind and thermal power and the peak-valley complementarity of frequency response under wind power frequency regulation power and load power disturbance, the power trajectory of the wind turbine in the frequency support phase is constructed, and the parameters are solved using the time domain analytical solution of the system frequency response.

[0129] Through mechanism analysis, the defects of the traditional fixed slope active frequency response strategy and its improved version are revealed. To address these defects, a designed active frequency response strategy is proposed, which designs the output power of wind turbines in the frequency support stage and the frequency support exit stage respectively.

[0130] The traditional slope setting active frequency response strategy is referred to as traditional strategy 1 in the present application. Under the control of traditional strategy 1, the step increase of active power of the wind turbine after the disturbance can significantly reduce the initial frequency change rate. However, it needs to be pointed out that the initial frequency change rate of the actual power grid is difficult to measure, and usually the average frequency change rate within 200 ms or 500 ms after the disturbance is concerned. When the step amplitude is large, the improvement of the average frequency change rate within the measurement time window can be achieved. Although traditional strategy 1 can effectively suppress the frequency change rate after the disturbance, the large-scale and rapid release of the rotor kinetic energy of the wind turbine will significantly weaken the dynamic response capability of the primary frequency modulation of the thermal power unit.

[0131] On the basis of traditional strategy 1, the response start time of the wind turbine is considered, and a delay setting slope support strategy is proposed by using the peak-valley matching compensation principle of the frequency response under the action of the wind power frequency modulation power and the load power disturbance. The strategy is referred to as traditional strategy 2 in the present application. Based on the control of traditional strategy 1, the time when the system frequency reaches the peak value under the action of the wind turbine frequency modulation power is earlier than the time when the valley value is reached under the action of the load power disturbance This makes the primary minimum frequency not be improved sufficiently. In order to improve the primary minimum frequency, the wind turbine is delayed to respond to the disturbance in traditional strategy 2, and the delay time The strategy only relies on the increased active power of the thermal power unit to respond to the power disturbance in the initial stage of the disturbance (the delay time , and the potential of the coordinated output of the multi-energy power to participate in frequency modulation is not considered, which will significantly worsen the initial frequency change rate in the high-penetration new energy power system.

[0132] The traditional strategy has the following problems:

[0133] (1) Support strategy: Traditional strategy 1 and traditional strategy 2 fail to start from the perspective of the total power change amount when designing the frequency modulation power trajectory of the wind turbine. The two designs focus on improving the initial frequency change rate and the primary frequency drop minimum point respectively, and fail to comprehensively consider the improvement of the two aspects.

[0134] (2) Exit strategy: The step power reduction scheme is adopted when the frequency support exits in the two traditional strategies, and the contradiction between the FSD depth and the speed recovery speed is balanced by selecting a smaller power drop amount ΔP down when the frequency support exits. However, the relatively slow response speed of the thermal power unit makes it unable to make up for the power shortage of the system caused by the step reduction of the wind power, and inevitably leads to the problem of the secondary frequency drop of the system.

[0135] In view of the problems existing in the traditional strategy, the power trajectory in the frequency support stage of the present application is shown as curve A-B'-C' in (a) of Figure 4 , and the power trajectory in the frequency support exit stage is shown in Figure 4The power trajectory in (a) is the expected trajectory, which is also the target trajectory designed later. The arc segment from the initial operating point A to the support power peak point B', and the arc segment from the support power peak point B' to the critical point C' of the exit support phase, where point A corresponds to the initial speed ω0 and initial power P0, and point B' corresponds to the speed ω B' With power P0+ΔP0, point C' corresponds to the speed ω C' With power P0. Figure 4 Medium P MPPT is the wind turbine output power in maximum power tracking mode, P Tlim is the output power corresponding to the torque limit, ω min is the minimum speed, ω C' and ω C are the rotation speeds of the proposed strategy and the traditional strategy when exiting the support phase, ω D' and ω E are the speeds when the exit powers of the proposed strategy and the traditional strategy reach the maximum power tracking output power, ω B' is the speed when the supporting power reaches the peak value, ω0 is the initial speed, t B' is the time for the support power to reach the peak value, t D' The time it takes for the wind turbine output power to reach the maximum power tracking output.

[0136] In the frequency support stage, the output power of the wind turbine increases linearly with time after the active frequency response is started to t B′ time, t B′ Output power P after time W It decreases linearly according to the speed. In this stage, the wind turbine outputs active power P W The expression is:

[0137]

[0138] Where: d is the duration of the power rise period, t0 is the start time of frequency support, t C′ is the frequency support exit moment (the moment when wind power is equal to the initial power P0). ΔP0 is the initial power increment, P MPPT (ω0) is the output power corresponding to the initial speed ω0 in the maximum power tracking operation mode. In the present invention, P MPPT (ω0)=P0, wind turbine output active power P W [ω(t)] is a functional of the rotational speed ω(t), which is a function of time t.

[0139] On the P-ω plane, the wind turbine outputs active power P W [ω(t)] can be expressed as:

[0140]

[0141] Where: Δω is the speed change caused by the wind power frequency modulation power rising stage, P MPPT (ω min ) is the wind power corresponding to the minimum speed in MPPT mode. When the wind speed remains constant and the speed changes slightly, the mechanical power of the wind turbine P m The change of can be ignored, then the following relationship exists in the power rising stage:

[0142]

[0143] Where: H W is the inertia time constant of the wind turbine.

[0144] The calculation formula of the speed change Δω can be rearranged as follows:

[0145]

[0146] The power rise period d is designed based on the concept of peak-valley complementarity of frequency response under wind power frequency regulation power and load power disturbance. In order to obtain the peak moment of system frequency response caused by wind power increase, the power change of wind turbines ΔP in the frequency support stage is calculated. W Make an approximation:

[0147]

[0148] in, Represents the approximate power change of the wind turbine, u(t) is the unit step function, K r It is a descending segment The slope, t off It is a descending segment Ending moment.

[0149] The peak moment of the system frequency generated by the change of wind power lags behind but is close to the mechanical power P of the wind turbine W Taking the peak moment into consideration, a linear approximation is used for the power variation decrease segment.

[0150] Take the slope K r equal Figure 4 (b) The arc segment In t B′ The tangent slope at the moment. Simulation research found that Figure 4 (b) The arc segment and The difference in slope increases slightly with time, and the difference in the initial stage is negligible, so

[0151]

[0152] P' = P - ω W P = P - ω K = P - ω P-ω P = P - ω It can be derived from the rotor motion equation.

[0153] P = P - ω The time-domain expression of the system frequency response under the action of the wind power ramp is:

[0154]

[0155] h (t) = h (t) - h (t) ramp (t) is the unit ramp response function of the transfer function G(s).

[0156] Since the frequency response peak time lags behind the time t B′ , the analysis is carried out from two cases: t B′ <t≤t off and t>t off .

[0157]

[0158] u(t) is the unit step function, because t B′ ,t off are greater than t0, so the unit step function u(t) is equal to 1. The derivative of the above equation is:

[0159]

[0160] Since h′ ramp (t) = h step (t), h step (t) is the unit step response function of the transfer function G(s), the above equation can be rewritten as:

[0161]

[0162] h (t) = h (t) - h (t) step (t) = Δf(t) / ΔP A , that is, equation (6) is divided by the power disturbance amplitude ΔP A . Let the derivative Δf W ′(t) = 0 of the time-domain system frequency response change, and take (that is, let the frequency peak time generated by the wind power ramp be equal to the valley time of the system frequency response under the action of the load power disturbance), that is, the value of the power rising section length d can be obtained.

[0163] By obtaining the power ramp duration d, we can determine the speed change Δω (Equation (12)), and then complete the support trajectory design. Regarding the startup method of the wind turbine frequency response, when the frequency change exceeds the frequency regulation dead zone (±0.03Hz to ±0.1Hz, the specific dead zone is determined by the grid demand), the wind turbine outputs active power according to the preset trajectory and performs active frequency response.

[0164] Step 3: Design the power trajectory of the wind turbine frequency support exit phase based on the Logistic function.

[0165] The output active power of wind turbines in the frequency support exit phase is designed to be:

[0166]

[0167] Where: k1 is the participation factor constructed based on the Logistic function, b is the shape coefficient, δP W is the exit speed ω C′ The corresponding power step amount, power step δP W The relationship between the wind speed and the wind speed is linear, such as Figure 5 As shown, P MPPT (ω C′ ) represents the maximum power tracking output power corresponding to the rotational speed at the end of support.

[0168] From formula (18) and Figure 2 It can be seen that in the frequency support exit stage, k1δP W Both ρ and wind power penetration rate are closely related to system frequency disturbances (caused by wind power). Figure 5 It shows that the power step δP W The wind power penetration rate is not only proportional to the total wind power disturbance power, but also positively correlated with the frequency disturbance magnitude due to the change in the system dynamic characteristics. To simplify the design, the present invention assumes that the wind power penetration rate is linearly related to the equivalent disturbance power. Based on this, the present invention proposes a design rule for the shape coefficient b as shown in formula (19), which makes the b value monotonically decrease with the increase of penetration rate and wind speed. In order to balance the adjustment effect within the wind turbine operating range and the target penetration rate range, the wind speed reference value v base and the permeability reference value ρ base Select values ​​near the middle of their respective intervals.

[0169]

[0170] Where: a, k2, k3 are design rule coefficients, v W is the wind speed, ρ is the wind power penetration rate. For the wind turbine selected in the simulation study of the present invention, a=0.54, k2=0.4, k3=0.04, v base =9.6m / s,ρbase = 20%.

[0171] In summary, the design of the active frequency response trajectory of the wind turbine generator set is completed in combination with formulas (9), (10), (12), (17)-(19). When the wind turbine generator set real-time monitoring system frequency offset is applied, when the frequency modulation dead zone is exceeded (±0.03 Hz to ±0.1 Hz, the specific dead zone is determined according to the grid demand), the wind turbine generator set outputs active power according to the preset trajectory to support the frequency. When the output power P W of the wind turbine generator set drops to equal the initial power P0, the frequency support exit phase is entered, and the rotating speed is restored until the initial rotating speed ω0.

[0172] In order to verify the strategy, the simulation model is built in MATLAB / Simulink, and the proposed strategy is compared with traditional strategy 1 and traditional strategy 2. Four simulation scenarios are set, and scenarios 1-3 are implemented based on the simulation model shown in Figure 2 Table 1. Scenario 4 is implemented based on a 4-machine 2-area system simulation model, Figure 6 The 4-machine 2-area simulation system topology built to verify the proposed strategy is given, and the generator set parameters are shown in Table 2. Figure 6 The G1-G3 are thermal power units, and the wind farm composed of 400 1.5 MW doubly-fed wind turbine generators is connected to bus 4.

[0173] Table 1 Simulation parameters

[0174]

[0175] Table 2 Generator set parameters

[0176]

[0177] Scenario 1: Set the wind speed to 9.6 m / s, the wind power penetration rate to 20%, and the power disturbance ΔP L A 0.1 p.u. step disturbance occurs at 70 s, Figure 7The simulation results of the application and two traditional strategies are given. The traditional strategy 1 immediately increases the output of the wind turbine after the disturbance occurs. At the end of the frequency support, the system appears a secondary frequency drop phenomenon due to the sharp change of power. In this strategy, the lowest point of the first frequency drop is the smallest among the three strategies, reaching 49.4551 Hz, and the secondary drop amplitude is 0.0366 Hz. The traditional strategy 2 introduces a delay in the framework of the traditional strategy 1, and adjusts the starting time of the wind turbine response. This adjustment improves the lowest point of the first frequency drop, which is 3.96% (49.4767 Hz) higher than that of the traditional strategy 1. However, this strategy gives up the potential of wind-fire coordination to improve the frequency change rate at the initial stage of the disturbance, and only the output response of the thermal power unit to the system power disturbance, the 200ms average frequency change rate is the largest among the three strategies (0.757 Hz / s). The lowest point of the first frequency drop of the application is 49.4946 Hz, which is 7.25% and 3.42% higher than that of the traditional strategy 1 and the traditional strategy 2 respectively; the secondary drop amount is 0.0145 Hz, which is 60.38% and 60.49% higher than that of the traditional strategy 1 and the traditional strategy 2 respectively. The reason for the large improvement of the secondary drop is that the strategy proposed in the application slows down the speed of the wind turbine output reduction at the initial stage of the support exit, which gives the thermal power unit time to respond to this power disturbance, so that the total output power of the system does not change suddenly, as shown in (d) of Figure 7 In , the total output of the wind and thermal power of the strategy proposed in the application is greater than that of the traditional strategy 2, and the increase of the total output of the system power realizes the improvement of the initial average frequency change rate of the strategy proposed in the application over the traditional strategy 2, and the 200ms average frequency change rate is improved by 4.47%.

[0178] Scenario 2: The wind power penetration rate in scenario 1 is increased to 30%, and the rest of the conditions remain unchanged, and the simulation results are shown in Figure 8 and Table 3. With the increase of the wind power penetration rate, the 200ms average initial change rate of the traditional strategy 2 increases from 0.757 Hz / s to 0.869 Hz / s, which is consistent with the trend revealed by the maximum frequency change rate that the maximum frequency change rate will increase with the increase of the wind power penetration rate. Because the traditional strategy 2 is a delayed response, only the thermal power unit responds at the initial stage of the disturbance. Compared with scenario 1, the 200ms average initial change rate of the strategy proposed in the application increases by 11.31% in scenario 2, which is smaller than the 14.72% of the traditional strategy 2, thereby verifying the effectiveness of the application in changing the initial frequency change rate. It is worth noting that the data in Table 3 show that the traditional strategy 1 has the ability to change the initial change rate which increases with the increase of the penetration rate.

[0179] Table 3 Influence of different strategies on the frequency response dynamic process

[0180]

[0181] Scenario 3: This scenario increases the wind speed in Scenario 2 to 10.8m / s, while other conditions remain unchanged. The simulation results are as follows: Figure 9 As shown. Under higher wind speed and wind power penetration rate, the traditional strategy exacerbates the secondary drop phenomenon of the system frequency, and the secondary drop amount increases from 0.0366Hz in scenario 1 to 0.0649Hz. Reasonable setting of the shape coefficient b can balance the contradiction between the depth of the secondary drop of frequency and the speed recovery speed. According to the adjustment of the shape coefficient b value, the secondary drop amount of the system frequency under the control of the strategy proposed in this invention is 0.0237Hz. From the total output power curves of wind power and thermal power in scenarios 1-3, Figure 7 (d) Figure 8 (d) and Figure 9 As shown in (d), the proposed strategy achieves maximum total system power output power at the end of the ramp-up process. The proposed strategy enables a faster ramp-up of the total frequency-modulated power of the system power supply, better achieving coordination between wind and fire power.

[0182] Table 4 SFR model transfer function parameter identification results

[0183]

[0184] Scenario 4: The wind power penetration rate in this scenario is 21% and the wind speed is 9.6m / s. At t=20s, a load power step disturbance of 196MW occurs at node 7. The parameters of the 4-machine 2-zone system transfer function (Equation (3)) used in this scenario are obtained by identification using the MATLAB parameter identification toolbox (System Identification Toolbox) based on the measured data of the power disturbance and frequency response dynamic process. The parameter identification results are shown in Table 4, and the comparison between the identification SFR model output and the measured frequency response curve is shown in Figure 10 (The present invention identifies the per-unit value of the measured data, Figure 10 Result obtained by multiplying by the reference value 50Hz). The simulation results of scenario 4 are as follows Figure 11 The analysis of the proposed strategy for improving frequency response dynamics (initial frequency change rate, primary drop depth, and secondary drop depth) in scenarios 1, 2, and 3 also applies to scenario 4. Furthermore, the wind turbine speed curves in each of the above scenarios demonstrate that the proposed strategy achieves its goal of improving the SFD problem without sacrificing speed recovery.

[0185] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A wind turbine active frequency response method considering wind-thermal power coordination, characterized in that: The steps include: Step 1: Establish an improved SFR model that considers the governor dynamics and high proportion of wind power, and solve the improved SFR model to obtain the time domain analytical solution of the system frequency response; Step 2: Considering the synergy between wind and thermal power, as well as the peak-valley complementarity of frequency response under wind power frequency regulation power and load power disturbance, the power trajectory of the wind turbine frequency support phase is constructed, and the parameters are solved using the time domain analytical solution of the system frequency response. Step 3: Design the power trajectory of the wind turbine frequency support exit phase based on the Logistic function.

2. The method for active frequency response of wind turbines considering wind-thermal power coordination according to claim 1, characterized in that: The dynamic transfer function of the speed regulator is described using a first-order inertia link as: Where: T G Is the inertia time constant of the speed regulator, equivalent adjustment coefficient 1 / R=K m / R′,K m is the mechanical power gain coefficient, 1 / R′ is the adjustment coefficient, and s is the Laplace operator; The transfer function of the improved SFR model from the power disturbance input ΔP(s) to the frequency change output Δf(s) is: Where: ΔP(s) represents the disturbance power ΔP L Or the frequency modulation power of the wind turbine ΔP W , D is the system damping coefficient, ρ is the wind power penetration level; H SYS is the system equivalent inertia time constant; D is the system damping coefficient; F H is the work ratio of the high-pressure cylinder of the thermal power unit; T R is the reheat time constant of the thermal power unit.

3. The active frequency response method of wind turbines considering wind-thermal power coordination according to claim 2, characterized in that: The time domain analytical solution of the system frequency response is: Where: In the formula: f1, f2, f3, δ, ζ, η, b1, b2, b3, Ω, x,A m is the intermediate coefficient.

4. The active frequency response method for wind turbines considering wind-thermal power coordination according to claim 3 is characterized in that: The time domain analytical solution of the system frequency response is solved by improving the SFR model transfer function G( S ) is abbreviated as: Where: d0, d1, d2, e0, e1, e2, e3 are intermediate coefficients, and Considering the system disturbance power of step form ΔP(t)=ΔP A ε(t),ΔP A represents the disturbance amplitude, ε(t) is the unit step function; Performing Laplace transform yields: Then when the disturbance occurs, the system frequency response Δf(s) is: Transfer function G( S ) is a cubic parametric equation, and the Cardano formula is used to solve the equation; Perform partial fraction expansion and inverse Laplace transform on the system frequency response Δf(s) to obtain the time domain expression Δf(t) of the system frequency response.

5. The method for active frequency response of a wind turbine generator system considering wind-thermal power coordination according to any one of claims 1 to 4, characterized in that: The expected power trajectory of the P-ω plane in the frequency support stage is: the arc segment from the initial operating point A to the support power peak point B', and the support power peak point B' to the critical point C' of the exit support stage, where point A corresponds to the initial speed ω0 and initial power P0, and point B' corresponds to the speed ω B' With power P0+ΔP0, point C' corresponds to the speed ω C' With power P0. During the frequency support phase, the wind turbine outputs active power P W After the active frequency response is started, it increases linearly with time until the support power reaches the peak time t B′ , time t B′ The wind turbine output active power P W Decreases linearly with speed.

6. The method for active frequency response of wind turbines considering wind-thermal power coordination according to claim 5, characterized in that: The wind turbine outputs active power P W The expression is: Where: d is the duration of the power rise period, t0 is the starting time of frequency support, t C′ is the end time of frequency support, at which time the wind turbine output active power is equal to the initial power P0; ΔP0 is the initial power increment, P MPPT (ω0) is the output power corresponding to the initial speed ω0 in the maximum power tracking operation mode, and P MPPT (ω0)=P0, wind turbine output active power P W [ω(t)] is a functional of the rotational speed ω(t), which is a function of time t.

7. The method for active frequency response of wind turbines considering wind-thermal power coordination according to claim 6, characterized in that: On the P-ω plane, the wind turbine outputs active power P W [ω(t)] is expressed as: Where: Δω is the speed change caused by the wind power frequency modulation power rising stage, P MPPT (ω min ) is the wind power corresponding to the minimum speed in MPPT mode; Wind turbine mechanical power P m When the change of is ignored, the following relationship exists during the power increase phase: Where: H W is the inertia time constant of the wind turbine; The calculation formula for the speed change Δω is: The duration d of the power rise segment is designed based on the concept of peak-valley complementarity of frequency response under wind power frequency regulation power and load power disturbance.

8. The method for active frequency response of wind turbines considering wind-thermal power coordination according to claim 7, characterized in that: The method for solving the parameters using the time domain analytical solution of the system frequency response is: In order to obtain the peak moment of system frequency response caused by the increase in wind power, the power change of wind turbines in the frequency support stage ΔP W Make an approximation: in, Represents the approximate power change of the wind turbine, u(t) is the unit step function, K r It is a descending segment The slope, t off It is a descending segment Ending moment; The peak moment of the system frequency generated by the change of wind power lags behind but is close to the mechanical power P of the wind turbine W Considering the peak moment, the power variation decrease section adopts a linear approximation; Take the slope K r Equal to the arc segment in the power-time relationship curve In t B′ The slope of the tangent line at the moment: Where: P′ W For arc segments Power, K P-ω is the slope of the wind turbine output power curve in the P-ω plane, It is derived from the rotor motion equation; Approximate power change of wind turbines The time domain expression of the system frequency response under the action of is: Where: h ramp (t) is the unit ramp response function of the transfer function G(s); Since the peak moment of the frequency response lags behind t B′ moment, therefore, from t B′ <t≤t off With t>t off Two situations are analyzed: Derivative of the above formula: Since h′ ramp (t) = h step (t), h step (t) is the unit step response function of the transfer function G(s), and the above formula can be rewritten as: Where: Unit step response function h step (t) = Δf(t) / ΔP A , let the derivative of the frequency response change of the time domain system Δf′ W (t) = 0, and the frequency peak moment generated by the additional wind power generation is equal to the valley moment of the system frequency response under the action of load power disturbance, and the value of the power rise period length d is obtained.

9. The method for active frequency response of a wind turbine generator system considering wind-thermal power coordination according to any one of claims 6 to 8, characterized in that: The output active power of the wind turbine in the frequency support exit stage is designed to be: Where: P MPPT is the wind power in the maximum power tracking mode, k1 is the participation factor constructed based on the Logistic function, b is the shape coefficient, δP W is the exit speed ω C′ The corresponding power step amount, power step δP W It shows linear characteristics with wind speed, P MPPT (ω C′ ) represents the maximum power tracking output power corresponding to the rotational speed at the end of frequency support, and b is the shape coefficient.

10. The method for active frequency response of wind turbines considering wind-thermal power coordination according to claim 9, characterized in that: The value of the shape coefficient b decreases monotonically with the increase of permeability and wind speed. The shape coefficient b is: Where: a, k2, k3 are design rule coefficients, v W is the wind speed, ρ is the wind power penetration rate; wind speed reference value v base and the permeability reference value ρ base Select values ​​near the middle of their respective intervals.

Images