A voltage sag evaluation method suitable for a doubly-fed wind farm accessing a power grid

Abstract

The application relates to a voltage sag evaluation method suitable for double-fed wind farm access to a power grid. The method comprises the following steps: analyzing the action characteristics of Crowbar protection of a double-fed wind turbine, establishing a wind farm equivalent model suitable for voltage sag analysis of a power grid side; establishing a wind farm iterative updating model based on the correction index of grid connection point voltage sag evaluation results; establishing a fault traversal model containing the wind farm equivalent, realizing fault simulation of the whole network; and finally realizing voltage sag evaluation of the double-fed wind farm access to the power grid.

Description

Technical Field

[0001] This invention belongs to the field of new energy, specifically relating to a voltage sag assessment method applicable to the grid connection of doubly fed wind farms. Background Technology

[0002] With the increasing global demand for clean energy, wind power, as a renewable energy source, has received widespread attention and application. Doubly fed wind farms, as one of the mainstream wind power generation technologies, offer advantages such as high efficiency, stability, and reliability, and have become an important development direction for the wind power industry. While the increasing integration of doubly fed wind farms into the power grid optimizes the grid's power supply structure, it also brings numerous challenges to power quality, supply reliability, and security. Furthermore, the impact of the reactive power support capability of wind turbines on the power grid during low-voltage ride-through has received widespread attention.

[0003] Voltage sag is currently the most frequent and serious power quality problem, causing significant economic losses to sensitive users. With the increasing integration of wind power and other renewable energy sources into the grid, the voltage sag problem has been alleviated to some extent. However, when a symmetrical fault occurs in the grid, wind farms enter a low-voltage ride-through state. During this period, the wind farm injects reactive power into the grid, providing voltage support to the grid connection point and surrounding nodes, which alters the voltage sag assessment results. Therefore, to achieve accurate voltage sag assessment for doubly-fed induction generator (DFIG) wind farms connected to the grid, a voltage sag assessment method that considers the reactive power support capability of wind farms is urgently needed.

[0004] Existing traditional system fault analysis methods may lead to overestimation or underestimation of voltage sag risk. Current voltage sag risk calculations often simplify renewable energy sources and power electronic equipment, ignoring the differences in their electrical, mechanical, and control characteristics compared to traditional power grids, resulting in overestimation or underestimation of voltage sag risk. For voltage sag assessment of wind farms connected to the grid, existing simulation-based assessment processes often neglect or simplify the reactive power characteristics analysis of the wind farm's low-voltage ride-through process, leading to inflated sag depth and duration results and overestimation of voltage sag risk. Therefore, voltage sag assessment requires modeling and analyzing wind farms. However, existing wind farm modeling often simplifies the wind farm using single-unit equivalent modeling, failing to accurately characterize the reactive power characteristics during low-voltage ride-through. Furthermore, detailed modeling of wind farms is complex and time-consuming. Therefore, a simplified wind farm model suitable for grid-side voltage sag analysis is needed for grid-side voltage sag assessment. Summary of the Invention

[0005] The purpose of this invention is to address the deficiencies in the prior art by providing a voltage sag assessment method suitable for doubly-fed induction generator (DFIG) wind farms connected to the power grid.

[0006] To achieve the above objectives, the technical solution of the present invention is: a voltage sag assessment method applicable to doubly-fed induction generator (DFIG) wind farm grid connection, comprising:

[0007] The operating characteristics of the Crowbar protection of the doubly fed wind turbine were analyzed, and an equivalent model of the wind farm suitable for voltage sag analysis on the grid side was established.

[0008] An iterative update model for wind farms was established, with the grid connection point voltage sag assessment results as the correction index.

[0009] A fault traversal model with wind farm equivalents was established to realize fault simulation of the entire network, and finally to realize voltage sag assessment of doubly-fed wind farms connected to the grid.

[0010] In one embodiment of the present invention, the specific implementation of analyzing the operating characteristics of the Crowbar protection of the doubly-fed induction generator (DFIG) and establishing an equivalent wind farm model suitable for grid-side voltage sag analysis is as follows:

[0011] Step 1: Perform theoretical analysis and calculation of the DIFG rotor current:

[0012] Based on stator flux linkage orientation, and to facilitate analysis while neglecting magnetic saturation, the reference direction is adopted according to the conventions for electric motors. The second-order differential equation for the rotor current is derived by simultaneously applying the DIFG electromagnetic transient equations of the doubly-fed induction generator (DFIG), the rotor current reference value obtained from reactive power priority control, and the rotor voltage reference value obtained from the inner loop control of the rotor-side current. The rotor current i is then solved to obtain the second-order differential equation. r :

[0013]

[0014] In the formula: i rn i is the natural component of the rotor current. rf1 i rf2 i is the forced component of the rotor current. r0 Let i be the initial value of the rotor current. r,ref =i rd,ref +ji rq,ref i is the reference value for rotor current. rd,ref i is the reactive component of the rotor current reference value. rq,ref The active component of the rotor current reference value. L s L r L m These are the stator inductance, rotor inductance, and mutual inductance, R. s R r These are the stator and rotor resistances, respectively; U s0 P is the terminal voltage vector during steady-state operation. s,ref I is the reference value for stator active power. r,maxK represents the maximum allowable value of the rotor-side converter current. q k is the reactive power gain coefficient. p k i These are the proportional and integral constants of the PI controller, ω. s For synchronous angular velocity, s = (ω s -ω r ) / ω r ω is the slip of the induction generator. r K is the rotor angular velocity. d The voltage drop coefficient is τ1=jω s +R s / L s is the decay time constant of the stator transient flux linkage.

[0015] Step 2: Analyze the factors affecting rotor current, formulate Crowbar protection action criteria, and analyze the impact of Crowbar protection action on reactive power characteristics;

[0016] From the rotor current expression, we know that i rf1 With active reference value P s,ref and voltage drop coefficient K d Related; i rf2 With rotor angular velocity ω r and voltage drop coefficient K d Relatedly, due to the different operating states of wind turbines during grid faults, the time it takes for the rotor current to reach the Crowbar protection activation threshold also varies. Statistical analysis shows that the Crowbar protection activation time differs very little under different operating states, and the peak rotor current i r,max Greater than or equal to action threshold i r,th The Crowbar protection will activate at all times, therefore the activation time is uniformly taken as the moment when the rotor current reaches its peak value, t. max , i.e. i r,max Represented as:

[0017] i r,max =f(P s,ref ,ω r ,K d )

[0018] Rotor angular velocity and active power reference value are expressed using wind speed:

[0019]

[0020]

[0021] In the formula: V wind Let λ represent wind speed, and k1, k2, k3, and λ3 be constants.

[0022] Therefore, the peak rotor current is written as a function of the voltage drop coefficient Kd and the wind speed V. wind The expression:

[0023]

[0024] In the formula:

[0025] Plot the independent variable K d V wind The dependent variable is i r,max The three-dimensional curve, each (K d V wind ) corresponds to an i r,max , take i r,max =i r,th Plot the Crowbar protection action characteristic curve, i.e., the action criterion is: when i r,max >i r,th Crowbar protection action, when i r,max <i r,th The Crowbar protection does not activate.

[0026] After the Crowbar protection is activated, the original AC excitation control will not function, and the unit will operate as an asynchronous generator. This will result in the unit being unable to compensate for reactive power or even absorb reactive power. Therefore, the Crowbar protection should not be activated for too long, that is, after ensuring that the power electronic devices have survived the impact of transient overcurrent, it should be deactivated before the fault is cleared. Based on the working principle of the Crowbar protection, the wind farm is divided into two groups: one with reactive power compensation and one without reactive power compensation, based on the Crowbar protection operation. The reactive power external characteristics of the wind farm are accurately characterized, and a grouped equivalent model suitable for grid-side voltage sag analysis is established.

[0027] Step 3, Wind Farm Parameter Equivalents:

[0028] (1) Wind speed is equal

[0029]

[0030] In the formula: V wind,eq f represents the equivalent wind speed at a wind farm. --1 Let P be the inverse function of the wind power curve, where n represents the n DFIG wind turbines in the wind farm. x This represents the active power of the xth generating unit;

[0031] (2) Equivalent value of Cowbar protection resistor

[0032] The capacity-weighted method is commonly used in equivalent modeling of wind farms and dynamic equivalence of traditional power systems.

[0033]

[0034] In the formula: R c,eq R represents the equivalent resistance of Crowbar after clustering and aggregation. x Let W be the resistance value of the Crowbar of the x-th wind turbine in the wind farm. x S refers to the weighting coefficient of the xth wind turbine unit. x Refers to the rated capacity of the xth wind turbine unit;

[0035] (3) Equivalent parameters of the collector line

[0036] The parameters of the collector lines are calculated using the equivalent loss power method;

[0037]

[0038] In the formula: Z eq Z is the equivalent line impedance. x Let x be the impedance from the xth wind turbine to the grid connection point.

[0039] In one embodiment of the present invention, the specific implementation of establishing the wind farm iterative update model based on the grid connection point voltage sag assessment result as the correction index is as follows:

[0040] Step 1: Perform initial grouping and equivalence of the wind farm.

[0041] A single fault simulation was conducted using synchronous turbine units of the same capacity to replace the wind farm, yielding the grid voltage sag assessment result, i.e., the node voltage sag amplitude. The voltage sag amplitude at the wind farm's grid connection point was then calculated. As a voltage drop condition that needs to be considered for wind farm equivalence, the initial cluster equivalence of wind farms is performed in combination with wind farm data, and the wind farm equivalence model is connected to the power grid to form a wind farm grid-connected model.

[0042] Step 2: Perform fault simulation on the wind farm grid connection model, and use the sag assessment results to correct and update the wind farm equivalent model;

[0043] Simulations of the same fault from step one were performed on the wind farm grid-connected model to obtain voltage sag assessment results considering wind farm grid connection. The voltage sag amplitude at the current grid connection point was then calculated. Compared with the previous grid connection point voltage sag If a comparison is made,

[0044]

[0045] That is, the difference between the two evaluation results cannot be ignored. This would result in more units being assigned to the group where Crowbar protection is not activated, thus requiring further grouping and equalization; if

[0046]

[0047] That is, the difference in cluster partitioning caused by the difference between the two evaluation results can be ignored. The grouping results as voltage sag conditions and The clustering results used as voltage sag conditions are approximately consistent, and the voltage sag assessment results under the grid connection of this wind farm are then output.

[0048] Where, Δε k The error limit refers to the maximum value at which the difference between the evaluation results of the k-th iteration does not cause a significant change in the cluster partitioning.

[0049] In one embodiment of the present invention, Δε k The method for obtaining the value is as follows:

[0050] (1) Each wind turbine in the initial group corresponds to a (K) d,0 V wind ), K d,0 This represents the initial voltage drop coefficient at the generator terminals, where the wind speed V is at that time. wind The voltage drop coefficient corresponding to the Crowbar protection action characteristic curve is K. dc That is, the initial error limit represents

[0051] Δε n,1 =K dn,0 -K dnc

[0052] In the formula: Δε n,1 K represents the initial error limit of the nth wind turbine. dn,0 K represents the initial terminal voltage drop coefficient of the nth wind turbine. dnc This represents the voltage drop coefficient corresponding to the Crowbar protection action characteristic curve for the wind speed of the nth fan.

[0053] (2) Since K is in each iteration d All of these change, and the error limit changes in each iteration, but the wind speed of the same fan remains constant, i.e., K. dc The error limit remains unchanged, therefore it is expressed as:

[0054] Δε n,k =Kd n,k-1 -Kd nc

[0055] In the formula: Δε n,k K represents the error limit for the k-th iteration of the n-th wind turbine. dn,k This represents the coefficient for the k-th voltage drop at the terminal of the nth wind turbine;

[0056] (3) Take Δε k=min(Δε) 1,k ,Δε 2,k ,...,Δε n,k ), Δε k This represents the error limit for the k-th iteration of the entire wind farm.

[0057] (4) In one embodiment of the present invention, the specific implementation of establishing a fault traversal model with wind farm equivalents to realize fault simulation of the entire network is as follows:

[0058] Step 1: Perform iterative evaluation of a single fault to obtain the optimal cluster of the wind farm and store the optimal evaluation results;

[0059] Step 2: Establish a fault traversal model;

[0060] (1) Set the fault set R i , i = 1, 2, 3, ..., N;

[0061] (2) Perform the operation in step one for fault i;

[0062] (3) Forming fault traversal criteria i <N;

[0063] If i < N, let i = i + 1 and repeat step two; if i ≥ N, end the fault traversal.

[0064] Step 3: Calculate the expected value of the node voltage sag:

[0065]

[0066] In the formula: U sag,j U represents the expected value of the voltage sag at grid node j. i,j Let N represent the voltage sag at node j in the i-th fault, and N be the total number of faults.

[0067] Compared with the prior art, the present invention has the following beneficial effects: The method of the present invention performs grouping and equivalent evaluation of doubly-fed wind farms based on the operation of Crowbar protection, and continuously updates and corrects the equivalent model of wind farms according to the voltage sag evaluation results, thereby realizing the accurate characterization of the reactive external characteristics of doubly-fed wind farms, and thus accurately characterizing the support capability of wind farms for grid voltage, and finally realizing the accurate evaluation of voltage sag of doubly-fed wind farms connected to the grid. Attached Figure Description

[0068] Figure 1 This is a flowchart illustrating the voltage sag assessment process for doubly-fed wind farms connected to the power grid according to the present invention.

[0069] Figure 2 This is a flowchart illustrating the equivalent wind farm process of this invention. Detailed Implementation

[0070] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings.

[0071] This invention proposes a voltage sag assessment method applicable to doubly-fed induction generator (DFIG) wind farm grid connection, comprising:

[0072] (1) By analyzing the operating characteristics of Crowbar protection, an equivalent model of wind farm suitable for voltage sag analysis on the grid side is established;

[0073] (2) Conduct voltage sag assessment of the wind power grid connection model and establish an iterative update model for wind farms with the voltage sag assessment results at the grid connection point as the correction index.

[0074] (3) Establish a fault traversal model with wind farm equivalents to realize fault simulation of the whole network.

[0075] like Figure 1 As shown, the method of the present invention is specifically implemented as follows:

[0076] (I) Establishing an equivalent model for wind farms suitable for grid-side voltage sag analysis

[0077] Step 1: Perform theoretical analysis and calculation of the DIFG rotor current.

[0078] Based on stator flux linkage orientation, and to facilitate analysis while neglecting magnetic saturation, the reference direction is adopted according to the conventions for electric motors. The second-order differential equation for the rotor current is derived by simultaneously applying the DIFG electromagnetic transient equations of the doubly-fed induction generator (DFIG), the rotor current reference value obtained from reactive power priority control, and the rotor voltage reference value obtained from the inner loop control of the rotor-side current. The rotor current i is then solved to obtain the second-order differential equation. r :

[0079]

[0080] In the formula: i rn i is the natural component of the rotor current. rf1 i rf2 i is the forced component of the rotor current. r0 Let i be the initial value of the rotor current. r,ref =i rd,ref +ji rq,ref i is the reference value for rotor current. rd,ref i is the reactive component of the rotor current reference value. rq,ref The active component of the rotor current reference value. L s L r L m These are the stator inductance, rotor inductance, and mutual inductance, R. s R r These are the stator and rotor resistances, respectively; U s0P is the terminal voltage vector during steady-state operation. s,ref I is the reference value for stator active power. r,max K represents the maximum allowable value of the rotor-side converter current. q k is the reactive power gain coefficient. p k i These are the proportional and integral constants of the PI controller, ω. s For synchronous angular velocity, s = (ω s -ω r ) / ω r ω is the slip of the induction generator. r K is the rotor angular velocity. d The voltage drop coefficient is τ1=jω s +R s / L s is the decay time constant of the stator transient flux linkage.

[0081] Step 2: Analyze the factors affecting rotor current, formulate the Crowbar protection action criteria, and analyze the impact of Crowbar protection action on reactive power characteristics.

[0082] From the rotor current expression, we know that i rf1 With active reference value P s,ref and voltage drop coefficient K d Related; i rf2 With rotor angular velocity ω r Voltage drop coefficient K at the terminal d This is relevant. Because the operating conditions of wind turbines (wind speed, voltage drop) differ during grid faults, the time it takes for the rotor current to reach the Crowbar protection activation threshold also varies. Statistical analysis shows that the Crowbar protection activation time differs very little under different operating conditions, and the peak rotor current i r,max Greater than or equal to action threshold i r,th The Crowbar protection will activate at all times, therefore the activation time is uniformly taken as the moment when the rotor current reaches its peak value, t. max , i.e. i r,max Represented as:

[0083] i r,max =f(P s,ref ,ω r ,K d )

[0084] Rotor angular velocity and active power reference value are expressed using wind speed:

[0085]

[0086]

[0087] In the formula: V wind Let λ represent wind speed, and k1, k2, k3, and λ3 be constants.

[0088] Therefore, the peak rotor current can be expressed as a function of the voltage drop coefficient Kd and the wind speed V. wind The expression:

[0089]

[0090] In the formula:

[0091] Plot the independent variable K d V wind The dependent variable is i r,max The three-dimensional curve, each (K d V wind ) corresponds to an i r,max , take i r,max =i r,th Plot the Crowbar protection action characteristic curve, i.e., the action criterion is: when i r,max >i r,th Crowbar protection action, when i r,max <i r,th The Crowbar protection does not activate.

[0092] After the Crowbar protection is activated, the original AC excitation control will become ineffective, and the unit will operate as an asynchronous generator. This will prevent the unit from compensating for reactive power or even absorbing reactive power. Therefore, the Crowbar protection should not be activated for too long; that is, it should be deactivated before the fault is cleared, after ensuring that the power electronic devices have survived the transient overcurrent impact. Generally, 60ms is used. Based on the working principle of the Crowbar protection, the wind farm can be divided into two groups: one with reactive power compensation and one without. This allows for an accurate characterization of the reactive power external characteristics of the wind farm, and the establishment of a grouped equivalent model suitable for grid-side voltage sag analysis.

[0093] Step 3: Equivalent values ​​of wind farm parameters.

[0094] (1) Wind speed is equal

[0095]

[0096] In the formula: V wind,eq f represents the equivalent wind speed at a wind farm. --1 Let P be the inverse function of the wind power curve, where n represents the n DFIG wind turbines in the wind farm. x This represents the active power of the xth generating unit.

[0097] (2) Equivalent value of Cowbar protection resistor

[0098] The capacity weighting method is commonly used in equivalent modeling of wind farms and dynamic equivalence of traditional power systems.

[0099]

[0100] In the formula: R c,eq R represents the equivalent resistance of Crowbar after clustering and aggregation. x Let W be the resistance value of the Crowbar of the x-th wind turbine in the wind farm. x S refers to the weighting coefficient of the xth wind turbine unit. x This refers to the rated capacity of the xth wind turbine unit.

[0101] (3) Equivalent parameters of the collector line

[0102] The parameters of the collector line are calculated using the equivalent loss power method.

[0103]

[0104] In the formula: Z eq Z is the equivalent line impedance. x Let P be the impedance from the xth wind turbine to the grid connection point. x This represents the active power output of the xth wind turbine unit.

[0105] (II) Establishing an iterative update model for wind farms with grid connection point voltage sag assessment results as the correction index.

[0106] Step 1: Perform initial clustering and equivalence of the wind farm. The wind farm clustering and equivalence process is as follows: Figure 2 As shown.

[0107] A single fault simulation was conducted using synchronous turbine units of the same capacity to replace the wind farm, yielding the grid voltage sag assessment results (i.e., node voltage sag amplitude). The voltage sag amplitude at the wind farm's grid connection point was then calculated. As a voltage drop condition that needs to be considered when equipping wind farms, the initial cluster equivalence of wind farms is performed by combining wind farm data, and the wind farm equivalence model is connected to the power grid to form a wind farm grid-connected model.

[0108] Step 2: Perform fault simulation on the wind farm grid connection model, and use the transient assessment results to correct and update the wind farm equivalent model.

[0109] Simulations of the same fault from step one were performed on the wind farm grid-connected model to obtain voltage sag assessment results considering wind farm grid connection. The voltage sag amplitude at the current grid connection point was compared with the previous grid connection point voltage sag amplitude. If...

[0110]

[0111] That is, the difference between the two evaluation results cannot be ignored. This would result in more units being assigned to the group where Crowbar protection is not activated, thus requiring further grouping and equalization; if

[0112]

[0113] That is, the difference in cluster partitioning caused by the difference between the two evaluation results can be ignored. The grouping results as voltage sag conditions and The clustering results used as voltage sag conditions are approximately consistent, and the voltage sag assessment results for this wind farm grid connection are then output.

[0114] Where, Δε k The error limit refers to the maximum value at which the difference between the evaluation results of the k-th iteration does not cause a significant change in the cluster partitioning, Δε. k This can be obtained from the Crowbar protection action characteristic curve. Δε k The method for obtaining the value is as follows:

[0115] (1) Each wind turbine in the initial group corresponds to a (K) d,0 V wind ), at this time V wind The voltage drop coefficient corresponding to the Crowbar protection action characteristic curve is K. dc That is, the initial error limit represents

[0116] Δε n,1 =K dn,0 -K dnc

[0117] In the formula: Δε n,1 K represents the initial error limit of the nth wind turbine. dn,0 K represents the initial terminal voltage drop coefficient of the nth wind turbine. dnc This represents the voltage drop coefficient corresponding to the Crowbar protection action characteristic curve for the wind speed of the nth fan.

[0118] (2) Since K is in each iteration d All of these change, and the error limit changes in each iteration, but the wind speed of the same fan remains constant, i.e., K. dc Since it remains unchanged, the error limit can be expressed as:

[0119] Δε n,k =Kd n,k-1 -Kd nc

[0120] In the formula: Δε n,k K represents the error limit for the k-th iteration of the n-th wind turbine. dn,kThis represents the coefficient for the k-th voltage drop at the terminal of the nth wind turbine.

[0121] (3) Take Δε k =min(Δε) 1,k ,Δε 2,k ,...,Δε n,k ), Δε k This represents the error limit for the k-th iteration of the entire wind farm.

[0122] (III) Establish a fault traversal model with wind farm equivalents to realize multi-fault simulation of the whole network.

[0123] The voltage sag assessment proposed in this invention for doubly fed wind farms connected to the grid should calculate faults at various locations across the entire grid. However, the degree of voltage sag at the grid connection point caused by faults at different locations varies, resulting in different equivalent values ​​for wind farm clusters due to faults at different locations. Therefore, this invention establishes a fault traversal model that includes equivalent values ​​for wind farms to achieve multi-fault simulation across the entire grid.

[0124] Step 1: Use the method in (II) to perform iterative evaluation of a single fault to obtain the optimal cluster of the wind farm and store the optimal evaluation results.

[0125] Step 2: Establish a fault traversal model.

[0126] (1) Set the fault set R i i = 1, 2, 3, ..., N

[0127] (2) Perform step one operation on fault i.

[0128] (3) Forming fault traversal criteria i <N

[0129] If i < N, let i = i + 1 and repeat step two; if i ≥ N, end the fault traversal.

[0130] Step 3: Calculate the expected value of the node voltage sag.

[0131]

[0132] In the formula: U sag,j U represents the expected value of the voltage sag at grid node j. i,j Let N represent the voltage sag at node j in the i-th fault, and N be the total number of faults.

[0133] The above are preferred embodiments of the present invention. Any changes made to the technical solution of the present invention that do not exceed the scope of the technical solution of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for assessing voltage sags in doubly-fed induction generator (DFIG) wind farms connected to the power grid, characterized in that, include: The operating characteristics of the Crowbar protection of the doubly fed wind turbine were analyzed, and an equivalent model of the wind farm suitable for voltage sag analysis on the grid side was established. An iterative update model for wind farms was established, with the grid connection point voltage sag assessment results as the correction index. Establish a fault traversal model with wind farm equivalents to realize fault simulation of the entire network, and finally realize the voltage sag assessment of doubly fed wind farms connected to the grid; The specific implementation of the wind farm iterative update model based on the grid connection point voltage sag assessment result as the correction index is as follows: Step 1: Perform initial clustering and equivalence analysis on the wind farm; A single fault simulation was conducted using synchronous turbine units of the same capacity to replace the wind farm, yielding the grid voltage sag assessment result, i.e., the node voltage sag amplitude. The voltage sag amplitude at the wind farm's grid connection point was then calculated. As a voltage drop condition that needs to be considered for wind farm equivalence, the initial cluster equivalence of wind farms is performed in combination with wind farm data, and the wind farm equivalence model is connected to the power grid to form a wind farm grid-connected model. Step 2: Perform fault simulation on the wind farm grid connection model, and use the sag assessment results to correct and update the wind farm equivalent model; Simulations of the same fault from step one were performed on the wind farm grid-connected model to obtain voltage sag assessment results considering wind farm grid connection. The voltage sag amplitude at the current grid connection point was then calculated. Compared with the previous grid connection point voltage sag value If a comparison is made, That is, the difference between the two evaluation results cannot be ignored. This will cause more units to be assigned to the group where the Crowbar protection is not activated, so it is necessary to regroup and equalize them; like That is, the difference in cluster partitioning caused by the difference between the two evaluation results can be ignored. The grouping results as voltage sag conditions and The clustering results used as voltage sag conditions are approximately consistent, and the voltage sag assessment results under the grid connection of this wind farm are then output. in, The error limit refers to the maximum value at which the difference between the evaluation results of the k-th iteration does not cause a significant change in the cluster partitioning; The method for obtaining the value is as follows: (1) Each wind turbine in the initial group corresponds to a (K) d,0 V wind ), K d,0 This represents the initial voltage drop coefficient at the generator terminals, where the wind speed V is at that time. wind The voltage drop coefficient corresponding to the Crowbar protection action characteristic curve is K. dc That is, the initial error limit represents In the formula: This represents the initial error limit for the nth wind turbine. This represents the initial terminal voltage drop coefficient of the nth wind turbine. This represents the voltage drop coefficient corresponding to the Crowbar protection action characteristic curve for the wind speed of the nth fan. (2) Since K is in each iteration d All of these change; the error limit changes in each iteration, but the wind speed of the same fan remains constant, i.e., K. dc The error limit remains unchanged, therefore it is expressed as: In the formula: This represents the error limit for the k-th iteration of the n-th wind turbine. This represents the coefficient for the k-th voltage drop at the terminal of the nth wind turbine; (3) Take , This represents the error limit for the k-th iteration of the entire wind farm.

2. The voltage sag assessment method for doubly-fed induction generator (DFIG) wind farms connected to the power grid according to claim 1, characterized in that, The analysis of the operating characteristics of the Crowbar protection of the doubly-fed induction generator (DFIG) and the establishment of an equivalent wind farm model suitable for grid-side voltage sag analysis are specifically implemented as follows: Step 1: Perform theoretical analysis and calculation of the DIFG rotor current: Based on stator flux linkage orientation, and to facilitate analysis while neglecting magnetic saturation, the reference direction is adopted according to the conventions for electric motors. The second-order differential equation for the rotor current is derived by simultaneously applying the DIFG electromagnetic transient equations of the doubly-fed induction generator (DFIG), the rotor current reference value obtained from reactive power priority control, and the rotor voltage reference value obtained from the inner loop control of the rotor-side current. The rotor current i is then solved to obtain the second-order differential equation. r : In the formula: i rn i is the natural component of the rotor current. rf1 i rf2 i is the forced component of the rotor current. r0 This is the initial value of the rotor current. This is the reference value for rotor current. The reactive component of the rotor current reference value. The active component of the rotor current reference value. , , , , L s L r L m These are the stator inductance, rotor inductance, and mutual inductance, R. s R r These are the stator and rotor resistances, respectively; U s0 The terminal voltage vector during steady-state operation. This is a reference value for stator active power. K represents the maximum allowable value of the rotor-side converter current. q k is the reactive power gain coefficient. p k i These are the proportional and integral constants of the PI controller, respectively. For synchronous angular velocity, The slip of the induction generator. K is the rotor angular velocity. d This is a voltage drop coefficient. The decay time constant of the stator transient flux linkage; Step 2: Analyze the factors affecting rotor current, formulate Crowbar protection action criteria, and analyze the impact of Crowbar protection action on reactive power characteristics; From the rotor current expression, we know that i rf1 With active reference value P s,ref and voltage drop coefficient K d Related; i rf2 With rotor angular velocity and voltage drop coefficient K d Relatedly, due to the different operating states of wind turbines during grid faults, the time it takes for the rotor current to reach the Crowbar protection activation threshold also varies. Statistical analysis shows that the Crowbar protection activation time differs very little under different operating states, and the peak rotor current i r,max Greater than or equal to action threshold i r,th The Crowbar protection will activate at all times, therefore the activation time is uniformly taken as the moment when the rotor current reaches its peak value, t. max , i.e. i r,max Represented as: Rotor angular velocity and active power reference value are expressed using wind speed: In the formula: V wind Indicates wind speed, k1, k2, k3, It is a constant; Therefore, the peak rotor current is written as a function of the voltage drop coefficient Kd and the wind speed V. wind The expression: In the formula: , , , ; Plot the independent variable K d V wind The dependent variable is The three-dimensional curve, each (K d V wind ) corresponds to one ,Pick Plot the Crowbar protection action characteristic curve, i.e., the action criterion is: when Crowbar protection action, when The Crowbar protection does not activate. After the Crowbar protection is activated, the original AC excitation control will not function, and the unit will operate as an asynchronous generator. This will result in the unit being unable to compensate for reactive power or even absorb reactive power. Therefore, the Crowbar protection should not be activated for too long, that is, after ensuring that the power electronic devices have survived the impact of transient overcurrent, it should be deactivated before the fault is cleared. Based on the working principle of the Crowbar protection, the wind farm is divided into two groups: one with reactive power compensation and one without reactive power compensation, based on the Crowbar protection operation. The reactive power external characteristics of the wind farm are accurately characterized, and a grouped equivalent model suitable for grid-side voltage sag analysis is established. Step 3, Wind Farm Parameter Equivalents: (1) Wind speed equivalent In the formula: V wind,eq f represents the equivalent wind speed at a wind farm. --1 Let P be the inverse function of the wind power curve, where n represents the n DFIG wind turbines in the wind farm. x This represents the active power of the xth generating unit; (2) Equivalent value of Cowbar protection resistor The capacity-weighted method is commonly used in equivalent modeling of wind farms and dynamic equivalence of traditional power systems. In the formula: R c,eq R represents the equivalent resistance of Crowbar after clustering and aggregation. x Let W be the resistance value of the Crowbar of the x-th wind turbine in the wind farm. x S refers to the weighting coefficient of the xth wind turbine unit. x Refers to the rated capacity of the xth wind turbine unit; (3) Equivalent parameters of the collector line The parameters of the collector lines are calculated using the equivalent loss power method; In the formula: Z eq Z is the equivalent line impedance. x Let x be the impedance from the xth wind turbine to the grid connection point.

3. The voltage sag assessment method for doubly-fed induction generator (DFIG) wind farms connected to the power grid according to claim 1, characterized in that, The specific implementation of establishing a fault traversal model with equivalent wind farm values ​​to achieve fault simulation across the entire network is as follows: Step 1: Perform iterative evaluation of a single fault to obtain the optimal cluster of the wind farm and store the optimal evaluation results; Step 2: Establish a fault traversal model; (1) Set the fault set R i , i=1,2,3,…,N; (2) Perform step one on fault i; (3) Forming fault traversal criteria i <N; like Let i = i + 1, and repeat step two; if End fault traversal; Step 3: Calculate the expected value of the node voltage sag: In the formula: This represents the expected value of the voltage sag at grid node j. Let N represent the voltage sag at node j in the i-th fault, and N be the total number of faults.

Images