Modeling method for parameter equivalence of wind turbine in large-scale direct-drive wind farm

Abstract

The application relates to the field of new energy power generation, and specifically provides a modeling method for parameter equivalence of wind turbine generators in large-scale direct-drive wind farms, which comprises the following steps: collecting operation power and capacity information of each wind turbine generator in the wind farm, constructing a wind speed backstepping model through backstepping of an MPPT curve, and calculating equivalent wind speed of the wind farm; integrating capacity-weighted electrical parameters and PI controller parameters of multiple wind turbine generators to construct an equivalent wind turbine parameter set; equivalently processing the impedance of a power collection line based on a current injection equivalence principle; connecting a capacitor compensation module in parallel with a grid connection point of the equivalent wind turbine model to correct reactive power response; and finally establishing an equivalent wind turbine generator model and performing dynamic simulation verification under a typical disturbance scenario. The method has a clear structure and a clear parameter acquisition approach, can significantly improve the simulation accuracy and dynamic response consistency of wind power modeling, and is suitable for modeling and control of various large-scale wind power systems.

Description

Technical Field

[0001] This invention relates to the field of new energy power generation technology, and in particular to a modeling method for equivalence of wind turbine parameters in large-scale direct-drive wind farms. Background Technology

[0002] my country's installed wind power capacity continues to climb, and wind power is gradually shifting from an auxiliary power source to a primary power source. In particular, the rapid deployment of large-scale centralized wind farms and offshore wind power bases has led to direct-drive permanent magnet synchronous wind turbines (PMSGs) becoming the mainstream grid connection method due to their low maintenance, high efficiency, and strong adaptability. In most modern wind farms, a single wind farm often contains dozens or even hundreds of wind turbines, with varying wind speeds, capacities, and control strategies, posing significant challenges to power system modeling and simulation.

[0003] In the dynamic simulation, stability assessment, and control strategy design of wind farms connected to the grid, it is necessary to construct an equivalent wind farm model with a certain level of accuracy while reducing the modeling dimensionality. Traditional wind farm modeling often employs the "representative turbine" method, the average parameter method, or the typical wind speed representation method. Their common characteristic is that they sacrifice some dynamic consistency and individual variability during the modeling process, leading to simulation results deviating from the actual operating state. These methods exhibit significant shortcomings in the following aspects:

[0004] 1. Lack of precise description of differences in wind turbine operation: Different wind turbines have different operating states and dynamic response characteristics due to differences in installation location, wind speed distribution, scheduling strategies, etc. The averaging method or simplified method is difficult to retain this difference.

[0005] 2. The parameter equivalence method is simple and crude, and lacks physical consistency: Current equivalent parameters are mostly processed by geometric mean or equal capacity merging, ignoring the capacity dependence of key dynamic characteristics such as wind turbine controller, converter, and inertial response, resulting in distortion of the overall response of the wind farm.

[0006] 3. Failure to consider the spatial distribution of wind speed and nonlinearity of wind energy conversion: MPPT (maximum power point tracking) and pitch adjustment make the nonlinear relationship between wind turbine output power and wind speed significant. The equivalent wind speed synthesized from different wind speeds needs to be back-calculated based on the power, which cannot be effectively handled by conventional methods.

[0007] 4. The complexity of the collector network topology is not reflected: In chain, radial or hybrid collector line structures, the traditional equivalent method does not reasonably preserve the impedance characteristics between the unit and the grid connection point, resulting in significant voltage and current injection errors.

[0008] In summary, existing equivalent modeling methods for wind turbines struggle to balance modeling accuracy, dynamic consistency, and physical rationality. This is especially true when modeling large-capacity direct-drive wind farms, necessitating the development of a more engineering-practical and mathematically rigorous parametric equivalence strategy. Summary of the Invention

[0009] The purpose of this invention is to provide a modeling method for equivalence of wind turbine parameters in large-scale direct-drive wind farms. This method can simultaneously handle multiple characteristics of wind turbines, such as electrical parameters, control parameters, output power, and wind speed, and combine them with the collector network structure for impedance equivalence, thereby improving the model's ability to accurately reflect the overall behavior of the wind farm and providing technical support for grid-connected analysis and system-level modeling of large-scale wind farms.

[0010] To achieve the aforementioned objectives, the present invention employs the following technical solution: a modeling method for equivalence of wind turbine parameters in large-scale direct-drive wind farms, comprising the following steps:

[0011] (1) Wind speed inversion and equivalent processing:

[0012] Real-time operating power P of multiple wind turbines within the wind farm was collected. i With capacity S i By combining the wind power expression with the maximum power point tracking (MPPT) control relationship, an inverse function model is constructed. The equivalent wind speed of each unit is calculated; then, the overall equivalent wind speed v of the wind farm is calculated using a capacity-weighted method. eq :

[0013]

[0014] (2) Weighted integration of electrical parameters based on capacity:

[0015] Obtain the stator resistance r of each unit s Stator reactance x s Magnetizing inductance L f Moment of inertia H g Parameters, in terms of capacity S i The overall equivalent electrical parameters of the fan are obtained by weighted averaging:

[0016]

[0017] (3) PI controller parameter aggregation modeling:

[0018] The proportional gain K of the active and reactive current loop controllers for each wind turbine p Integral coefficient K i Perform capacity-weighted integration to build a PI controller with the same dynamic response as the original system:

[0019]

[0020] (4) Equivalent impedance of collector lines:

[0021] Based on the principle of maintaining consistent voltage and current at the grid connection point (PCC), the equal current method is used to combine the impedances of the collector network, equivalence multiple branches to a unified branch impedance Z. eq ,in:

[0022]

[0023] (5) Parallel reactive power compensation modeling:

[0024] A capacitor C is connected in parallel on the PCC side of the equivalent wind turbine model. comp Its capacitance value is adjusted according to the steady-state power balance condition, so that the reactive power response of the equivalent model is consistent with the response of the original wind farm after a fault or under low voltage ride-through conditions:

[0025]

[0026] (6) Modeling, simulation and error verification:

[0027] The constructed equivalent model was applied to the power system simulation platform, and the correlation between its active power error, PCC point voltage error and dynamic curve was evaluated by comparing the disturbance response curve under typical operating conditions with the original system modeling results.

[0028] Furthermore, the wind speed inverse model is based on the MPPT control relationship:

[0029]

[0030] Where A is the swept area of ​​the wind turbine. air density, For power coefficient, For the tip speed ratio, It is the propeller pitch angle.

[0031] Furthermore, the weighted electrical parameters include, but are not limited to, stator resistance, reactance, excitation inductance, permanent magnet flux linkage, and mechanical damping coefficient, and these parameters are applicable to the mathematical modeling of permanent magnet synchronous motors.

[0032] Furthermore, the PI controller aggregation is suitable for the following structure:

[0033] (1) Active current loop controller;

[0034] (2) Reactive current loop controller;

[0035] (3) Voltage regulation loop controller;

[0036] (4) Speed ​​ring regulating controller;

[0037] Each controller adopts a proportional-integral control structure, and the aggregated equivalent controller maintains the dynamic adjustment performance of the system.

[0038] Furthermore, the parallel capacitor C comp Dynamically adjust the PCC point under short-circuit disturbances by using sensitivity analysis or simulation optimization to reduce the reactive response error.

[0039] Furthermore, the modeling scheme is applicable to grid-connected direct-drive wind farms with no fewer than 10 wind turbine units and a total capacity of no less than 50MW, with turbine capacities ranging from 1.5MW to 6MW.

[0040] In practical applications, this invention can be widely used in scenarios such as wind farm master station modeling, power grid dispatch analysis, power system transient simulation, wind power control strategy testing, and fault ride-through capability assessment. For large-capacity, widely distributed, and complex wind-condition direct-drive wind farms, the parameter equivalent modeling method proposed in this invention can significantly reduce modeling complexity and computational overhead without sacrificing the dynamic characteristics of the system.

[0041] The equivalent parameters of the wind turbine obtained by the wind speed back-calculation and capacity weighting methods are highly representative. The aggregation processing of the PI controller ensures the consistency of the control response. The equivalent impedance of the collector line and the parallel capacitor compensation measures together construct an equivalent model that closely matches the behavior of the original system. Simulation results show that this method has high accuracy, strong adaptability, and good dynamic consistency under typical scenarios such as wind speed disturbances and grid faults. It can effectively replace detailed modeling schemes and be used for engineering analysis and online simulation of large-scale wind power systems.

[0042] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0043] 1. A strategy for obtaining equivalent parameters combining capacity weighting and power conservation principles is proposed, which not only preserves the dynamic characteristics of wind turbines but also avoids information loss during simplification, significantly improving the accuracy and engineering applicability of the simulation model.

[0044] 2. A wind speed inverse model based on the maximum power point tracking (MPPT) inverse function is introduced, which realizes accurate calculation of wind speed nonlinear synthesis under complex wind conditions and solves the dynamic response deviation problem caused by the traditional capacity weighting method.

[0045] 3. A collector network parameter aggregation method based on the principle of constant injection current at PCC points is adopted, which takes into account the influence of wind farm topology on electrical characteristics and improves the accuracy of the equivalent model under fault conditions.

[0046] 4. A parallel capacitor compensation mechanism is added to finely adjust the reactive power output of the equivalent model, significantly reducing the errors in voltage and current injection before and after equivalence. Actual measurements show that under conditions of wind speed fluctuations and uneven capacity distribution, the voltage error of the wind farm outlet bus modeled using this method is less than 1%, demonstrating good dynamic consistency and scalability. Attached Figure Description

[0047] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0048] Figure 1 This is an overall flowchart of wind farm clustering and parameter equivalence in this invention.

[0049] Figure 2 This is an equivalent schematic diagram of the wind farm collector line in this invention.

[0050] Figure 3 This is a schematic diagram showing the location of the centralized compensation capacitor in this invention. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. Of course, the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0052] Example 1

[0053] See Figure 1 This embodiment provides a technical solution: a modeling method for equivalent wind turbine parameters in large-scale direct-drive wind farms. Based on clustering of the direct-drive wind farm, to obtain the equivalent wind farm model, it is necessary to perform equivalent calculations on wind turbines within the same group and determine the aggregation parameters of the collector network. This chapter proposes a method for deriving equivalent wind speed based on the principle of constant power, a method for obtaining equivalent wind turbine parameters based on the capacity-weighted method, and a method for aggregating collector network parameters based on the principle of constant injected current. It also proposes reducing reactive power errors by using parallel compensation capacitors.

[0054] The capacity-weighted method approximates a parameter of an equivalent wind turbine unit by multiplying that parameter by its capacity weight. Using equivalent components of a wind farm cluster, various electrical and mechanical parameters of the wind turbine units can be accurately calculated. The equivalent wind turbine unit accurately reflects and represents the external characteristics of the wind farm. Since wind farms typically consist of wind turbines of the same capacity and model, the parameters of the equivalent wind turbine are related to the number of turbines in the same group. Looking from the wind turbine towards the grid, the parameters that need to be equivalent mainly include the equivalent wind turbine unit's capacity, impedance, inertia, and controller parameters.

[0055] The equivalent parameters of the direct-drive wind turbine model are as follows:

[0056] (1)

[0057] In the formula: S is the capacity of the wind turbine generator; This represents the number of direct-drive fans in the same cluster before the equivalence; the subscript eq represents the parameter after the equivalence. For the magnetizing reactance of PMSG; and These are the PMSG stator reactance and rotor reactance, respectively. and These are the stator resistance and rotor resistance of the PMSG, respectively.

[0058] The equivalent inertia time constant of the transmission system is:

[0059] (2)

[0060] In the formula, Let be the time constant of the inertia of the i-th wind turbine. Let be the time constant of the inertia of the i-th generator.

[0061] Chopper circuit equivalent resistance value for:

[0062] (3)

[0063] The equivalent parameters of each PI controller within the PMSG are:

[0064] (4)

[0065] In the formula, These are the named values ​​of the proportional and integral constants of the PI controller for the i-th direct-drive fan, respectively. The equivalent capacity and impedance values ​​of the step-up transformer connected to the equivalent fan are:

[0066] (5)

[0067] In the formula, Let i be the capacity of the i-th step-up transformer. Let be the short-circuit impedance value of the i-th step-up transformer.

[0068] For different turbine models, averaging wind speed by capacity is a common approach. However, the wind speed of each PMSG in a direct-drive wind farm of the same model is significantly affected by time lag and wake effects. Using the capacity-weighted method to calculate the average wind speed input leads to inaccurate equivalent results, affecting the accuracy of subsequent dynamic analysis. Therefore, based on the principle of constant input power and employing a reverse calculation method, using MPPT curves to determine the equivalent turbine wind speed is more accurate. The steady-state power of the i-th PMSG unit can be calculated from the MPPT curve. Based on the principle of power conservation within the same group before and after the equivalent value, the output power of the equivalent PMSG unit can be calculated. The equivalent wind speed was obtained based on the MPPT curve. for:

[0069] (6)

[0070] In the formula, f is the MPPT function of the direct-drive wind turbine. This represents its inverse function.

[0071] Collector network parameter aggregation is a crucial part of the dynamic equivalence process for wind farms. Collector network parameters can alter power flow distribution, power loss, and turbine terminal voltage distribution within the wind farm. If the equivalent collector network parameters are inaccurate, the error before and after equivalence will be too large when a fault occurs near the wind farm outlet. Therefore, a more accurate collector network equivalence method is needed. The most commonly used methods for collector network equivalence include the REI equivalence method and the WARD equivalence method. Both of these schemes rely on maintaining the grid connection point voltage and power unchanged before and after equivalence, resulting in cumbersome equivalence structures and relatively poor accuracy. Therefore, this chapter proposes a method based on the principle of constant injected current at the PCC point to obtain the collector network parameters.

[0072] In large-scale wind farms, the output power of each wind turbine is collected and transmitted to the substation via collector lines. Depending on the wind conditions and terrain characteristics of the wind farm, different topologies are used between the wind turbines, with the basic wiring configurations being radial and chain-type. Therefore, during the wind farm equivalence process, the line impedance between the turbines can affect the accuracy of the equivalence. This paper selects a chain-type collector system, and the topology diagram of the collector network aggregation principle is shown below. Figure 2 As shown.

[0073] Suppose that a certain group has Each PMSG unit is equivalent to one equivalent unit W for all units in the group. eq W eq The connection impedance to the PCC point is First, the steady-state active power of each PMSG unit is calculated using the MPPT curve and related formulas based on the given wind speed. and reactive power Generally, the default is Treating the node containing PMSG as a PQ node, the internal power flow of the wind farm is calculated. Based on the principle of constant injected current, the equivalent machine W... eq Current injected into the grid connection point for:

[0074] (7)

[0075] In the formula, conj is the conjugate operator. Equivalent wind turbine W eq Voltage at the node for:

[0076] (8)

[0077] From this, the equivalent impedance of the wind farm collector network can be obtained. for:

[0078] (9)

[0079] In the formula: This refers to the voltage at the wind farm's grid connection point. The above method has strict physical meaning and is simple to calculate, easily derived from the steady-state power flow before the fault. .

[0080] It should be noted that the PMSG output differs significantly before and during the fault. During the fault, the dynamic process of the unit's output exhibits noticeable oscillations and decreases, resulting in different power flows within the wind farm. Therefore, the power flow calculated based on the power flow before the fault is not accurate. This approach is not entirely applicable during fault periods. Considering the significant impact of the fault steady-state voltage on the dynamics of the PMSG, it can be approximated as... The fault steady-state voltage is:

[0081] (10)

[0082] like The actual voltage is too low; it can be increased appropriately. ,Increase Electrical distance from the fault point to increase Conversely, it decreases. and reduce Electrical distance from the fault point.

[0083] In the equivalent process based on PCC principles such as constant current, the distributed capacitance of the collector lines is not considered, which can lead to a significant deviation between the reactive power output of the equivalent system and the detailed model. Therefore, to maintain a constant reactive power at the grid connection point of the PMSG wind farm after the equivalent process, a compensation capacitor can be connected in parallel at the grid connection point. To improve the reactive power accuracy of the equivalent model, the location of the compensation capacitor is as follows: Figure 3 As shown.

[0084] Compensation capacitor The specific formula for calculating the value is as follows:

[0085] (11)

[0086] In the formula, These represent the reactive power injected into the grid connection point of the wind farm before and after the equivalent values.

[0087] Example 2

[0088] Based on Example 1, this embodiment further verifies the modeling method for equivalent wind turbine parameters in large-scale direct-drive wind farms described in this invention. It demonstrates the modeling accuracy and dynamic response consistency under wind speed disturbances and fault conditions, and compares it with traditional modeling methods, highlighting the engineering application value and simulation advantages of the technical solution of this invention.

[0089] I. Experimental System Setup

[0090] A large-scale wind farm simulation model containing 30 direct-drive permanent magnet synchronous wind turbine generators was constructed, with the following main configuration:

[0091] (1) Capacity of a single wind turbine: 2 MW, with a rotor radius of 82 m;

[0092] (2) Electrical structure: PMSG + full power converter structure;

[0093] (3) Control method: MPPT control strategy is adopted, and the current loop is adjusted by the PI controller in a closed loop;

[0094] (4) Grid connection method: The power is collected to the 35kV bus via the collection line, and then connected to the power grid via the step-up transformer;

[0095] (5) Simulation platform: MATLAB / Simulink (Simscape PowerSystems);

[0096] (6) Simulation time: 10s, step size: 100 μs.

[0097] II. Comparison of Modeling Methods

[0098] Method 1: Traditional Equivalent Modeling Method (referred to as "Traditional Method")

[0099] The parameters of all fans were obtained by averaging the values ​​of the fans at equal capacities.

[0100] Ignore actual wind speed differences and differences in inertia, reactance, and controller parameters;

[0101] Without considering reactive power compensation and current disturbance coordination, only an equivalent single-machine model is constructed.

[0102] Method 2: The modeling method of this invention (hereinafter referred to as "this method")

[0103] The following operations are performed sequentially according to the steps in the claims of this invention:

[0104] (1) Wind speed inverse calculation and equivalent calculation: by inversely solving the MPPT relation Obtain representative wind speed v eq Calculate the dynamic average wind speed using the wind turbine capacity weighted formula;

[0105] (2) Parameter capacity weighted integration:

[0106] (3) PI controller parameter aggregation: ;

[0107] (4) Equivalent impedance of collector network (equal current method): ;where U eq I is the equivalent fan voltage. eq The sum of the current injected into each fan.

[0108] (5) Parallel capacitor compensation: A parallel capacitor CcompC_{comp}Ccomp is connected on the PCC side of the equivalent model to correct the reactive response of the output of the equivalent model.

[0109] III. Operating Conditions and Disturbance Types

[0110] Operating Condition A: Wind Speed ​​Disturbance Response Test

[0111] Initial steady-state wind speed: 10 m / s;

[0112] At t=1.5 seconds, the wind speed suddenly changed to 13 m / s;

[0113] Observe the active / voltage response after modeling.

[0114] Operating Condition B: Short Circuit Fault Disturbance Test

[0115] At t=2.5 seconds, a three-phase short circuit fault occurs at point PCC;

[0116] The fault lasted for 0.1 seconds, and the ability of the equivalent model to track the fault response was examined.

[0117] IV. Evaluation indicators are defined as shown in Table 1 below.

[0118] Table 1

[0119]

[0120] V. Simulation comparison results are shown in Table 2.

[0121] Table 2 six,

[0123] VII. Comprehensive Analysis

[0124] This embodiment shows that:

[0125] Excellent dynamic consistency: The method of this invention maintains a high DCC value under fault disturbances and sudden wind speed changes, and the response curve is highly consistent with the real system.

[0126] Voltage deviation has been significantly reduced: the maximum voltage deviation has decreased from 5.47% to 1.68%, and the equivalent model is better able to reflect the characteristics of power grid disturbance transmission;

[0127] Significant reactive power response correction: Through the parallel compensation mechanism, reactive power output deviation is effectively controlled, making steady-state reactive power closer to the behavior of the full model;

[0128] High feasibility of engineering: all parameters are generated through capacity weighting and structural modeling, which has portability and computational efficiency, and is suitable for scenarios such as wind power simulation, protection configuration, and scheduling analysis.

[0129] 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 within the protection scope of the present invention.

Claims

1. A modeling method for equivalence of wind turbine parameters in large-scale direct-drive wind farms, characterized in that, Includes the following steps: (1) Wind speed inversion and equivalent processing: Real-time operating power P of multiple wind turbines within the wind farm was collected. i With capacity S i By combining the wind power expression with the MPPT control relationship, an inverse function model is constructed. The equivalent wind speed of each unit is calculated; then, the overall equivalent wind speed v of the wind farm is calculated using a capacity-weighted method. eq : ； The wind speed inverse model is based on the MPPT control relationship: ； Where A is the swept area of ​​the wind turbine. air density, For power coefficient, For the tip speed ratio, The pitch angle; (2) Weighted integration of electrical parameters based on capacity: Obtain the stator resistance r of each unit s Stator reactance x s Magnetizing inductance L f Moment of inertia H g Parameters, in terms of capacity S i The overall equivalent electrical parameters of the fan are obtained by weighted averaging: ； (3) PI controller parameter aggregation modeling: The proportional gain K of the active and reactive current loop controllers for each wind turbine p Integral coefficient K i Perform capacity-weighted integration to build a PI controller with the same dynamic response as the original system: ； (4) Equivalent impedance of collector lines: Based on the principle of maintaining consistent voltage and current at the grid connection point (PCC), the equal current method is used to combine the impedances of the collector network, equivalence multiple branches to a unified branch impedance Z. eq ,in: ； U eq I is the equivalent fan voltage. eq The sum of the current injected into each fan; The PI controller aggregation is applicable to the following structure: (1) Active current loop controller; (2) Reactive current loop controller; (3) Voltage regulation loop controller; (4) Speed ​​ring regulating controller; Each controller adopts a proportional-integral control structure, and the aggregated equivalent controller maintains the dynamic adjustment performance of the system. (5) Parallel reactive power compensation modeling: A capacitor C is connected in parallel on the PCC side of the equivalent wind turbine model. comp Its capacitance value is adjusted according to the steady-state power balance condition, so that the reactive power response of the equivalent model is consistent with the response of the original wind farm after a fault or under low voltage ride-through conditions: ； (6) Modeling, simulation and error verification: The constructed equivalent model was applied to the power system simulation platform, and the correlation between its active power error, PCC point voltage error and dynamic curve was evaluated by comparing the disturbance response curve under typical operating conditions with the original system modeling results.

2. The modeling method for equivalent wind turbine parameters in large-scale direct-drive wind farms according to claim 1, characterized in that, The weighted electrical parameters include stator resistance, reactance, excitation inductance, permanent magnet flux linkage, and mechanical damping coefficient. These parameters are applicable to the mathematical modeling of permanent magnet synchronous motors.

3. The modeling method for equivalence of wind turbine parameters in large-scale direct-drive wind farms according to claim 1, characterized in that, The parallel capacitor C comp Dynamically adjust the PCC point under short-circuit disturbances by using sensitivity analysis or simulation optimization to reduce the reactive response error.

4. The modeling method for equivalence of wind turbine parameters in large-scale direct-drive wind farms according to claim 1, characterized in that, The modeling scheme is applicable to grid-connected direct-drive wind farms with no fewer than 10 wind turbine units and a total capacity of no less than 50MW, with turbine capacities ranging from 1.5MW to 6MW.

Images