Dynamic frequency control method of wind turbine considering weight coefficient and rotor energy state

Abstract

A dynamic frequency control method for a fan considering weight coefficient and rotor energy state includes the following steps: determining whether the fan participates in frequency regulation; if the system frequency deviation is greater than the set frequency regulation dead zone range and the fan rotor speed is greater than the minimum speed, virtual inertia control is started; otherwise, virtual inertia control is not started; virtual inertia control weight coefficient is introduced l d and virtual droop control weight coefficient l p , and according to △ f and d f / d t Adjust the size of the two according to the change; introduce the kinetic energy feedback coefficient and dynamically adjust the virtual inertia control coefficient according to the energy state of the fan rotor. K df and virtual droop control coefficient K pf According to the system frequency change, the fan virtual inertia control is locked, so that the fan can quickly restore the speed while ensuring the stability of the system frequency. The present invention can dynamically adjust the virtual inertia control parameters according to the system frequency change and the fan rotor kinetic energy reserve, so that the fan can fully participate in the frequency response while avoiding excessive deceleration of the rotor, effectively enhancing the system frequency stability capability.

Description

Technical Field

[0001] The present invention relates to the technical field of frequency control of wind power generator sets, and in particular to a method for dynamic frequency control of a wind turbine taking weight coefficients and rotor energy states into consideration. Background Art

[0002] As one of the fastest growing clean power sources, the large-scale grid connection of wind power has led to a reduction in grid inertia and a weakening of its ability to resist disturbances, and a significant increase in the risk of system frequency over-limit. In order to adapt to the future scenario of a high proportion of renewable energy, wind turbines need to have the ability to actively participate in grid frequency regulation. Therefore, some scholars have conducted research on wind power frequency control technology, and the virtual inertia control method of wind turbines that uses rotor energy to provide short-term power support has become a research hotspot. At present, the research focus on virtual inertia control is mainly concentrated on two aspects: ① How to determine the wind power virtual inertia control parameters to fully tap the frequency regulation potential of wind turbines and achieve better frequency control effects; ② How to set a reasonable speed recovery method so that the wind turbine can quickly restore the rotor speed after locking the virtual inertia control, while ensuring the system frequency stability performance.

[0003] Among them, for the study of virtual inertia control parameters, some scholars consider the frequency regulation capability of the fan itself, and adjust the virtual inertia control parameters according to the real-time rotor kinetic energy reserve of the fan, so that the frequency regulation capability of the fan is coupled with the kinetic energy storage to avoid excessive release of the rotor kinetic energy during the frequency response process, but this type of method does not consider the system frequency response requirements. When the system frequency deviation is large or the frequency change rate is too fast, the fan needs to provide a stronger frequency response capability. Therefore, some scholars consider dynamically adjusting the virtual inertia control parameters according to the frequency change situation, but there are few such literatures, and only consider the parameter adjustment method under the single control mode of droop control or inertia control, which cannot give full play to the frequency regulation capability of the fan.

[0004] The research on wind turbine speed recovery methods mainly focuses on optimizing the wind turbine operation trajectory to make its active output smoothly transition to the MPPT curve to avoid the impact of sudden changes in active power on the system frequency. However, changes in wind turbine active output will affect the system frequency. If a fixed speed recovery method is used, a secondary frequency drop may occur in a system with a high wind power penetration rate. At present, no relevant literature has considered this issue. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a dynamic frequency control method for a wind turbine that takes into account the weight coefficient and the rotor energy state, and can dynamically adjust the virtual inertia control parameters according to the system frequency change and the kinetic energy reserve of the wind turbine rotor, so that the wind turbine can fully participate in the frequency response while avoiding excessive deceleration of the rotor, so as to effectively enhance the system frequency stability capability. In addition, a virtual inertia control locking method is proposed, and relevant parameter adjustment schemes are formulated for different wind power penetration scenarios to ensure that the wind turbine can restore the optimal operating state while taking into account the system frequency stability performance.

[0006] The technical solution adopted by the present invention is:

[0007] The method for controlling the dynamic frequency of a wind turbine considering the weight coefficient and the energy state of a rotor comprises the following steps:

[0008] Step 1: Determine whether the fan participates in frequency regulation. If the system frequency deviation is greater than the set frequency regulation dead zone range and the fan rotor speed is greater than the minimum speed, start the virtual inertia control; otherwise, do not start;

[0009] Step 2: Introduce virtual inertia control weight coefficient λ d and virtual droop control weight coefficient λ p , and adjust the size of the two according to the changes of △f and df / dt to give full play to the control advantages of virtual inertia control and virtual droop control;

[0010] Step 3: Introduce kinetic energy adjustment coefficient K SOE , dynamically adjust the virtual inertia control coefficient K according to the wind turbine rotor energy state SOE df and virtual droop control coefficient K pf , so that the fan can effectively utilize the rotor energy while avoiding excessive rotor deceleration;

[0011] Step 4: According to the changes in system frequency, lock the fan virtual inertia control in a timely manner to enable the fan to quickly restore the speed while ensuring the stability of the system frequency.

[0012] In step 1, the virtual inertia control introduces an additional control loop of frequency deviation and frequency differential feedback in the active control link of the fan, so that the fan can release the rotor kinetic energy to provide power support when the system frequency deviates. The active reference power is expressed as:

[0013] P ref =P mppt +△P

[0014] =kω r 3 +K df ·df / dt+K pf ·△f

[0015] Where P mppt is the maximum power tracking value, △P is the additional power of the fan inertia control, K df is the virtual inertia control coefficient, K pf is the virtual droop control coefficient, and k is the maximum power tracking coefficient of the wind turbine.

[0016] In step 1, if the system frequency deviation is greater than the set frequency modulation dead zone range, and the fan rotor speed is greater than the minimum speed, the fan virtual inertia control is started, otherwise it is not started. The judgment conditions are:

[0017] |△f|>△f db &ω r >ω min

[0018] In the above formula, △f is the system frequency deviation, △f db is the primary frequency modulation dead zone threshold of the system, ω r is the rotor speed, ω min It is the lower limit of the fan rotor speed.

[0019] In the step 2, since the virtual droop control and virtual inertia control have good suppression effects on frequency deviation and frequency change rate respectively, the virtual droop control and virtual inertia control parameters should be adjusted in a targeted manner according to the changes in df / dt and △f during the system frequency response period. Therefore, the inertia control weight coefficient λ is considered to be introduced. d and droop control weight coefficient λ p , and always satisfy λ during the control process p +λ d =2, the additional power of fan inertia control is expressed as:

[0020] △P=λ d ·K df ·df / dt+λ p ·K pf ·△f

[0021] In the step 2, the frequency curve after the disturbance of the system is divided into a frequency deviation stage and a frequency recovery stage with the low frequency point as the boundary. The control strategies of each stage are as follows:

[0022] (1) Frequency deviation stage:

[0023] In the early stage of disturbance, the system frequency has a large change trend, that is, |df / dt| is large, while the frequency deviation is small, that is, |△f| is small. Considering that virtual inertia control has a stronger inhibitory effect on |df / dt| change, virtual inertia control should be dominant, and λ dSet to a larger value; after a certain delay, the synchronous unit speed control system gradually starts, |df / dt| gradually decreases, while |△f| is still increasing. Considering that the virtual droop control has a stronger hindering effect on the change of |△f|, the weight of the virtual inertia control should be weakened, so that the virtual droop control is dominant, and λ d Gradually reduce, when |df / dt|=0, the frequency deviation reaches the extreme value, λ d Reaching the minimum value λ d,min ;

[0024] (2) Frequency recovery phase:

[0025] The system frequency begins to recover gradually under the action of the speed regulator. At this time, the polarity of the system frequency change rate is opposite to that of the frequency deviation stage. Virtual inertia control will produce a "negative" power increment, which has a reverse effect on frequency recovery. Therefore, "negative" virtual inertia control should be used to promote rapid frequency recovery. Since |△f| is large in this stage, virtual droop control is still dominant, making λ d In a small range, it follows the synchronous change of |df / dt|, which is determined by λ d,min First increase and then gradually decrease.

[0026] Based on the above control strategy, a fuzzy logic controller is established. df / dt and △f are used as input variables of the fuzzy logic controller. max and ∆f max As the reference value, normalization is performed to obtain the domain interval of the two to be [-1,1], and the inertia weight coefficient λ d As an output variable, the domain interval is [0,2].

[0027] In step 3, the ratio of the real-time rotor kinetic energy of the wind turbine to the rotor kinetic energy at the rated speed is the rotor energy state SOE:

[0028]

[0029] In the above formula, ω max ,ω min They are respectively the upper and lower limits of the fan rotor speed operating range.

[0030] Since the fan frequency response is achieved by utilizing the rotor kinetic energy, in order to make full use of the rotor energy and avoid excessive deceleration of the rotor, the kinetic energy adjustment coefficient K is introduced. SOE , and adjust K according to the change of SOE SOE , K SOE The corresponding relationship with SOE is:

[0031]

[0032] Virtual inertia control coefficient Kdf and virtual droop control coefficient K pf The corresponding expression is:

[0033]

[0034] In the above formula, K do is the basic adjustment coefficient of virtual inertia control, K po It is the basic adjustment coefficient of virtual droop control.

[0035] In step 4, after the wind turbine uses the rotor kinetic energy to participate in the frequency response, the rotor speed needs to be restored to the initial value to return the wind turbine to the maximum power tracking state and ensure the maximum power generation efficiency. The existing control method directly cuts off △P, which will cause the wind turbine active power to instantly jump to the MPPT curve. The joint action of a large number of wind turbines will inevitably lead to the problem of secondary frequency disturbance. Based on this, consider gradually locking the virtual inertia control to make the wind turbine smoothly transition to the initial operating point;

[0036] (1) Locking control coefficient μ:

[0037] The locking strategy is implemented by gradually reducing the additional power △P by introducing the locking control coefficient μ, thereby dynamically correcting the wind turbine operating point trajectory so that it can be smoothly restored to the MPPT curve. The corresponding wind turbine active reference power is:

[0038] P ref =kω r 3 +△P·μ

[0039] Among them, the expression of μ is:

[0040]

[0041] In the formula, k o Slope coefficient of a linear function; t off is the start time of locking control; t re is the duration of the locking process; α is the wind power penetration rate of the system.

[0042] On the one hand, after the locking control is started, the wind turbine will reduce the active output to restore the rotor speed, and this process has a reverse effect on the system frequency recovery. As the wind power penetration rate increases, the total active power change per unit time of the wind farm increases, and the impact on the system frequency is stronger. Therefore, t re It should be appropriately extended as the wind power penetration level of the system increases to reduce the impact on the system frequency; on the other hand, t re The increase of k will reduce the fan speed recovery rate and prolong the time to return to the initial operating point. Therefore, it is necessary to consider both aspects. o Take a reasonable value, which is 150 in the present invention.

[0043] (2) Locking control start conditions:

[0044] Starting the locking control at different times will affect the system frequency performance. Most existing control methods use the convergence time of the fan speed as the locking control start time, but in fact this time may occur near the extreme point of frequency change. In this case, starting the locking control will cause the fan power to drop rapidly, which may cause the system frequency to continue to drop. Therefore, the locking control start time should be determined according to the system frequency change. The present invention considers that when the system frequency reaches the extreme point, the delay T r s restarts the locking control, and the frequency extreme point determination condition is:

[0045]

[0046] In the formula, f t 、f t-1 are the real-time frequency of the system and the frequency of the previous sampling cycle respectively; △f t , △f t-1 are the real-time frequency deviation of the system and the frequency deviation of the previous sampling period, respectively, n is the rated frequency of the system and the sampling period is 50ms.

[0047] The present invention provides a method for controlling the dynamic frequency of a fan taking into account the weight coefficient and the rotor energy state, and the technical effects are as follows:

[0048] 1) For a power system with large-scale wind power access, the control method of the present invention is adopted, and the wind turbine can dynamically adjust the virtual inertia control parameters according to the system frequency change and its own rotor kinetic energy reserve, so as to more effectively participate in frequency regulation and avoid excessive release of rotor kinetic energy.

[0049] 2) By adopting the control method of the present invention, on the one hand, the wind turbine can dynamically correct the operating point trajectory through locking control to return it to the MPPT mode, thereby improving the power generation efficiency of the wind farm; on the other hand, the locking control related parameters can be adjusted according to the wind power penetration rate of the system, so that the wind turbine can quickly restore the speed while ensuring the system frequency stability performance.

[0050] 3) The present invention can effectively enhance the frequency response capability of the fan and improve the frequency stability performance of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a flow chart of an embodiment of the present invention.

[0052] Figure 2 Schematic diagram of system frequency response range division.

[0053] Figure 3 This is the fuzzy logic reasoning result diagram.

[0054] Figure 4 It is the membership function relationship diagram corresponding to the system frequency change rate.

[0055] Figure 5 This is the membership function relationship diagram corresponding to the system frequency deviation.

[0056] Figure 6 It is the membership function relationship diagram corresponding to the inertia weight coefficient.

[0057] Figure 7 is the kinetic energy feedback coefficient K SOE Corresponding relationship diagram with SOE.

[0058] Figure 8 is the blocking control coefficient and blocking control start time △t=tt under different wind power penetration rates off The corresponding relationship diagram.

[0059] Fig. 9 4 is a simulation system diagram of an embodiment of the present invention.

[0060] Figure 10(a) is the simulation experiment diagram (system frequency) when wind speed = 9m / s;

[0061] FIG10( b ) is a simulation experiment diagram (rotor speed) when the wind speed is 9 m / s.

[0062] Figure 11(a) is a simulation experiment diagram (system frequency) at wind speed = 11m / s;

[0063] FIG11( b ) is a simulation experiment diagram (rotor speed) at a wind speed of 11 m / s.

[0064] Fig.12 This is the simulation experiment diagram when wind power penetration rate = 25%.

[0065] Fig.13 This is the simulation experiment diagram when wind power penetration rate = 35%. DETAILED DESCRIPTION

[0066] The present invention considers the weight coefficient and the rotor energy state of the fan dynamic frequency control method. Based on the fan's own frequency regulation capability and the system frequency response requirements, a method for dynamically correcting the virtual inertia control parameters according to the rotor energy state and the system frequency change is proposed. By introducing the virtual inertia control weight coefficient λ d , virtual droop control weight coefficient λ p, in order to give full play to the control advantages of inertial control and droop control in each stage of frequency response, while taking into account the operation of the wind turbine, the kinetic energy adjustment coefficient is introduced to constrain the virtual inertia control parameters, so as to effectively use the rotor kinetic energy to participate in frequency regulation. In addition, from the perspective of optimizing the operating state of the wind turbine, a locking control method after the frequency response is completed is proposed, and relevant parameter adjustment plans are formulated for different wind power penetration scenarios to ensure that the wind turbine can restore the optimal state while taking into account the system frequency stability performance. Finally, through simulation verification, compared with the traditional virtual inertia control method, the flow chart is as follows Figure 1 The method comprises the following steps:

[0067] Step 1: Determine whether the fan participates in frequency regulation based on the system frequency deviation and the fan operation status. If the system frequency deviation is greater than the set frequency regulation dead zone range, and the fan rotor speed is greater than the minimum speed, the fan virtual inertia control is started, otherwise it is not started. The judgment conditions are:

[0068] |△f|>△f db &ω r >ω min

[0069] In the above formula, △f is the system frequency deviation, △f db is the primary frequency modulation dead zone threshold of the system, ω r is the rotor speed, ω min It is the lower limit of the fan rotor speed.

[0070] Step 2: Introduce weight coefficients to adjust the wind turbine output according to the system frequency changes. In order to strengthen the frequency support effect of wind turbines, so that they can effectively suppress the frequency change rate and improve the frequency deviation extreme value as much as possible, the wind turbine virtual droop control and virtual inertia control parameters should be adjusted in a targeted manner according to the changes in df / dt and △f during the system frequency response period. Therefore, the inertia control weight coefficient λ is considered to be introduced. d and droop control weight coefficient λ p , and always satisfy λ during the control process p +λ d =2. The additional power of fan inertia control can be expressed as:

[0071] △P=λ d ·K df ·df / dt+λ p ·K pf ·△f

[0072] like Figure 2 As shown, the frequency curve after the system disturbance is divided into the frequency deviation stage and the frequency recovery stage with the low frequency point as the boundary. The control strategies for each stage are as follows:

[0073] (1) Frequency deviation stage: In the early stage of system disturbance, the system frequency has a large change trend, that is, |df / dt| is large, while the frequency deviation is small, that is, |△f| is small. Considering that virtual inertia control has a stronger inhibitory effect on |df / dt| changes, virtual inertia control should be dominant, and λ d After a certain delay, the synchronous unit speed control system gradually starts, |df / dt| gradually decreases, while |△f| is still increasing. Considering that the virtual droop control has a stronger effect on the change of |△f|, the weight of the virtual inertia control should be weakened, so that the virtual droop control is dominant, and λ d Gradually reduce, when |df / dt|=0, the frequency deviation reaches the extreme value, λ d Reaching the minimum value λ d,min .

[0074] (2) Frequency recovery stage: The system frequency begins to recover gradually under the action of the speed regulator. At this time, the polarity of the system frequency change rate is opposite to that of the frequency deviation stage. Virtual inertia control will produce a "negative" power increment, which has a reverse effect on frequency recovery. Therefore, "negative" virtual inertia control should be used to promote rapid frequency recovery. Since |△f| is large in this stage, droop control is still dominant, making λ d In a small range, it follows the synchronous change of |df / dt|, which is determined by λ d,min First increase and then gradually decrease.

[0075] Based on the above control idea, the fuzzy logic control rules are established as shown in Table 1.

[0076] Table 1 Fuzzy logic control rules

[0077]

[0078]

[0079] The inference result is obtained by Mamdani algorithm. Figure 3 As shown in the figure, df / dt and △f are used as the input variables of the fuzzy controller, and df / dt max and ∆f max As the reference value, normalization is performed and the domain interval of the two is obtained as [-1,1], including NL (negative large), NM (negative medium), NS (negative small), ZO (zero), PS (positive small), PM (positive medium), PL (positive large), a total of 7 levels, and the corresponding membership function is as follows Figure 4 and Figure 5 As shown. d As the output variable, the domain interval is [0,2], including VS (very small), RS (small), MO (medium), RL (large), VL (very large), a total of 5 levels, the corresponding membership function is as follows Figure 6 shown.

[0080] Step 3: Adjust the virtual inertia control parameters of the fan according to the rotor energy state. Since the fan participates in the frequency response by utilizing the rotor kinetic energy, the virtual inertia control and virtual droop control coefficients need to be adjusted according to the rotor kinetic energy reserve, so that the fan can fully utilize the rotor kinetic energy while avoiding excessive rotor deceleration and machine cutting.

[0081] The ratio of the real-time rotor kinetic energy of a wind turbine to the rotor kinetic energy at rated speed is the rotor energy state SOE:

[0082]

[0083] In the above formula, ω max ,ω min They are respectively the upper and lower limits of the fan rotor speed operating range.

[0084] In the existing control method, the wind turbine virtual inertia control coefficient K is usually adjusted directly according to the SOE size. df and virtual droop control coefficient K pf However, this method will still significantly weaken the inertia response strength of the fan when the SOE is high. Therefore, this paper considers introducing the kinetic energy adjustment coefficient K SOE , dynamically adjust K according to SOE changes SOE , and then adjust K df and K pf , thereby giving full play to the fan frequency response capability.

[0085] Kinetic energy adjustment coefficient K SOE The corresponding relationship with SOE is as follows Figure 7 The SOE variation range is divided into two intervals, and the wind power regulation coefficient K is adjusted differently according to the SOE situation. SOE , so as to give full play to the frequency response capability of the fan. When SOE is low, SOE≦0.6, the fan will release the rotor kinetic energy excessively, which will easily increase the risk of instability and machine tripping. Therefore, K SOE It should decrease synchronously with the decrease of SOE; when SOE is high, SOE>0.6, the fan should fully release the rotor kinetic energy to provide active support, K SOE The maximum value should be taken. In this case, K SOE The expression is:

[0086]

[0087] It is worth noting that when the system frequency is positively offset, the wind turbine reduces its active output, the rotor speed increases, and the SOE increases. When the SOE reaches the maximum value, the wind turbine can reduce the amount of wind energy captured by controlling the pitch angle, thereby limiting the rotor speed within the operating range. Therefore, there is no need to adjust the inertia control coefficient according to the SOE change, K SOE Always 1.

[0088] The expressions corresponding to the virtual inertia control coefficient and virtual droop control coefficient of the fan are:

[0089]

[0090] In the above formula, K do is the basic adjustment coefficient of virtual inertia control, K po is the basic adjustment coefficient of virtual droop control

[0091] Step 4: According to the system frequency change, the virtual inertia control is locked in time to quickly restore the speed of the fan while ensuring the stability of the system frequency. The existing control method directly cuts off △P, which will cause the active power of the fan to jump to the MPPT curve instantly. The joint action of a large number of fans will inevitably lead to the problem of secondary frequency disturbance. Based on this, consider gradually locking the inertia control to make the fan smoothly transition to the initial operating point.

[0092] (1) Locking control coefficient μ:

[0093] The locking strategy is implemented by gradually reducing the additional power △P by introducing the locking control coefficient μ, thereby dynamically correcting the wind turbine operating point trajectory so that it can be smoothly restored to the MPPT curve. The corresponding wind turbine active reference power is:

[0094] P ref =kω r 3 +△P·μ

[0095] Among them, the expression of μ is:

[0096]

[0097] In the above formula, k o Slope coefficient of a linear function; t off is the start time of locking control; t re is the duration of the locking process; α is the wind power penetration rate of the system.

[0098] On the one hand, after the locking control is started, the wind turbine will reduce the active output to restore the rotor speed, and this process has a reverse effect on the system frequency recovery. As the wind power penetration rate increases, the total active power change per unit time of the wind farm increases, and the impact on the system frequency is stronger. Therefore, t reIt should be appropriately extended as the wind power penetration level of the system increases to reduce the impact on the system frequency; on the other hand, t re The increase of k will reduce the fan speed recovery rate and prolong the time to return to the initial operating point. Considering both aspects, o When the value is 150, the blocking control coefficient and blocking control start time △t=tt under different wind power penetration rates are obtained. off The corresponding relationship is as follows Figure 8 shown.

[0099] (2) Locking control start conditions:

[0100] Starting the locking control at different times will affect the system frequency performance. Most existing control methods use the convergence time of the fan speed as the locking control start time, but in fact this time may occur near the extreme point of frequency change. In this case, starting the locking control will cause the fan power to drop rapidly, which may cause the system frequency to continue to drop. Therefore, the locking control start time should be determined according to the system frequency change. The present invention considers that when the system frequency reaches the extreme point, the delay T r s restarts the locking control, and the frequency extreme point determination condition is:

[0101]

[0102] In the above formula, f t 、f t-1 are the system real-time frequency and the frequency of the previous sampling cycle respectively; △f t , △f t-1 are the real-time frequency deviation of the system and the frequency deviation of the previous sampling period, respectively, n is the rated frequency of the system and the sampling period is 50ms.

[0103] Step 5: The proposed wind turbine dynamic frequency control strategy considering the weight coefficient and rotor energy state is verified to be effective and superior through simulation examples.

[0104] In the Matlab / simulink environment, we established Fig. 9 The improved four-machine two-area simulation system shown in the figure contains three synchronous generators and one wind farm. A frequency drop event is simulated by setting a 10% load surge at t=50s, and the following three control methods are used for the wind turbines for comparison: ①MPPT control; ②traditional virtual inertia control; ③the method of the present invention.

[0105] The simulation parameters are as follows:

[0106] The control parameters of the strategy proposed in the present invention are as follows: △f db 0.033Hz; df / dtmax and ∆f max They are 0.5Hz / s and 0.2Hz respectively; the basic adjustment coefficient of virtual inertia and virtual droop control is K po and K do 20 and 10 respectively; upper and lower limits of fan rotor speed ω max and ω min They are 1.2pu and 0.7pu respectively; the delay time T r 5s.

[0107] Synchronous generator parameters (G2, G3, G4): The adjustment coefficient is 0.05, the inertia time constant is 6.5s, and the rated capacity is 900MW·A. The parameters of the doubly fed wind turbine are shown in Table 2.

[0108] Table 2 Parameters of double-fed wind turbine

[0109]

[0110] The simulation projects include:

[0111] 1) In a system with a wind power penetration rate of 15%, the system frequency stability performance is verified when the wind speed is 9m / s and 11m / s. This project is verified by Figure 10(a), Figure 10(b) and Figure 11(a), Figure 11(b);

[0112] 2) Under wind speed = 11m / s, verify the system frequency stability performance when wind power penetration rate is 25% and 35%. Fig.12 and Fig.13 verify.

[0113] It can be seen from Figure 10(a) and Figure 10(b) that when the wind turbine adopts MPPT control, the system frequency drops the fastest, and the corresponding lowest frequency is 49.927Hz. In Method 2-3, the wind turbine increases power by using rotor energy, effectively enhancing the system equivalent inertia, thereby slowing down the rapid decline of the system frequency. When Method 2 is used, the system frequency is at a low point of 49.93Hz, but when the system frequency returns to the steady-state value, the fan rotor speed still has a deviation of 0.006pu from the initial value, and no longer continues to recover. The essence is that the existence of additional power makes the fan reach the steady-state operation point in advance. When Method 3 is used, the system frequency control effect is the best, the system frequency is only 49.932Hz, and after the fan participates in the frequency support, the speed can be restored to the initial speed through locking control, thereby improving the system frequency stability performance and the wind turbine power generation efficiency.

[0114] As shown in Figure 11(a) and Figure 11(b), when three control methods are used respectively, method 3 still has the best frequency control effect, followed by method 2, and method 1 is the worst. Comparing Figure 10(a) and Figure 10(b), method 3 has a significantly improved frequency drop suppression effect than method 2. This is because under this wind condition, the initial speed of the wind turbine is relatively high and has sufficient rotational kinetic energy to support the frequency response process. Therefore, method 3 can adjust the inertia control and droop control weight coefficients to enable the wind turbine to obtain a larger reference power during the frequency response period, thereby being able to utilize more rotor energy and show a better frequency control effect.

[0115] Depend on Fig.12 and Fig.13 It can be seen that with the increase of wind power penetration rate, method 3 can still play a good frequency control effect, and the effect is more significant. In addition, when the wind turbine lockout control is started, the system does not have an obvious secondary frequency drop problem because the corresponding lockout control duration is adjusted according to the wind power penetration rate.

Claims

1. A method for controlling the dynamic frequency of a fan considering weight coefficients and rotor energy states, characterized in that The following steps are involved: Step 1: Determine whether the fan participates in frequency regulation. If the system frequency deviation is greater than the set frequency regulation dead zone range and the fan rotor speed is greater than the minimum speed, start the virtual inertia control; otherwise, do not start; Step 2: Introduce virtual inertia control weight coefficient λ d and virtual droop control weight coefficient λ p , and adjust the size of the two according to the changes of △f and df / dt; Step 3: Introduce kinetic energy adjustment coefficient K SOE , dynamically adjust the virtual inertia control coefficient K according to the wind turbine rotor energy state SOE df and virtual droop control coefficient K pf ; Step 4: According to the change of system frequency, lock the virtual inertia control of the fan to quickly restore the fan speed while ensuring the stability of the system frequency; In the step 2, the frequency curve after the disturbance of the system is divided into a frequency deviation stage and a frequency recovery stage with the low frequency point as the boundary. The control strategies of each stage are as follows: (1) Frequency deviation stage: In the early stage of disturbance, |df / dt| is large, while |△f| is small. The virtual inertia control is dominant and λ d Set to a larger value; after a certain delay, the synchronous unit speed control system gradually starts, |df / dt| gradually decreases, while |△f| is still increasing, weakening the weight of virtual inertia control, making virtual droop control dominant, and setting λ d Gradually reduce, when |df / dt|=0, the frequency deviation reaches the extreme value, λ d Reaching the minimum value λ d,min ; (2) Frequency recovery phase: Since |△f| is large in this stage, virtual droop control is dominant, making λ d In a small range, it follows the synchronous change of |df / dt|, which is determined by λ d,min First increase and then gradually decrease; Based on the above control strategy, a fuzzy logic controller is established. df / dt and △f are used as input variables of the fuzzy logic controller. max and ∆f max As the reference value, normalization is performed to obtain the domain interval of the two to be [-1,1], and the inertia weight coefficient λ d As an output variable, the domain interval is [0,2]; In step 3, the ratio of the real-time rotor kinetic energy of the wind turbine to the rotor kinetic energy at the rated speed is the rotor energy state SOE: In the above formula, ω max ,ω min They are the upper and lower limits of the fan rotor speed operating range respectively; Introducing kinetic energy adjustment coefficient K SOE , and adjust K according to the change of SOE SOE , K SOE The corresponding relationship with SOE is: Virtual inertia control coefficient K df and virtual droop control coefficient K pf The corresponding expression is: In the above formula, K do is the basic adjustment coefficient of virtual inertia control, K po It is the basic adjustment coefficient of virtual droop control.

2. The method for controlling the dynamic frequency of a fan considering the weight coefficient and the rotor energy state according to claim 1, characterized in that: In step 1, the virtual inertia control introduces an additional control loop of frequency deviation and frequency differential feedback in the active control link of the fan, so that the fan can release the rotor kinetic energy to provide power support when the system frequency deviates. The active reference power is expressed as: P ref =P mppt +△P =kω r 3 +K df ·df / dt+K pf ·△f Where P mppt is the maximum power tracking value, △P is the additional power of the fan inertia control, K df is the virtual inertia control coefficient, K pf is the virtual droop control coefficient, and k is the maximum power tracking coefficient of the wind turbine.

3. The method for controlling the dynamic frequency of a fan considering the weight coefficient and the rotor energy state according to claim 1, characterized in that: In the step 1, if the system frequency deviation is greater than the set frequency modulation dead zone range, and the fan rotor speed is greater than the minimum speed, the fan virtual inertia control is started, otherwise it is not started; the judgment conditions are: △f|>△f db &oh r >oh min In the above formula, △f is the system frequency deviation, △f db is the primary frequency modulation dead zone threshold of the system, ω r is the rotor speed, ω min It is the lower limit of the fan rotor speed.

4. The method for controlling the dynamic frequency of a fan considering the weight coefficient and the rotor energy state according to claim 1, characterized in that: In step 2, the inertia control weight coefficient λ is introduced according to the changes of df / dt and Δf during the system frequency response period. d and droop control weight coefficient λ p , and always satisfy λ during the control process p +λ d =2, the additional power of fan inertia control is expressed as: △P=λ d ·K df ·df / dt+λ p ·K pf ·△f。 5. The method for controlling the dynamic frequency of a fan considering the weight coefficient and the energy state of the rotor according to claim 1, characterized in that: In step 4, the locking control coefficient μ is introduced to gradually reduce the additional power △P, thereby dynamically correcting the wind turbine operating point trajectory so that it can be smoothly restored to the MPPT curve. The corresponding wind turbine active reference power is: P ref =kω r 3 +△P·μ Among them, the expression of μ is: In the formula, k o Slope coefficient of a linear function; t off is the start time of locking control; t re is the duration of the locking process; α is the wind power penetration rate of the system; When the system frequency reaches the extreme point, the delay T r s restarts the locking control, and the frequency extreme point determination condition is: In the formula, f t 、f t-1 are the real-time frequency of the system and the frequency of the previous sampling cycle respectively; △f t , △f t-1 are the real-time frequency deviation of the system and the frequency deviation of the previous sampling period, respectively, n is the rated frequency of the system.

Images