Dynamic aggregation modeling method of wind farm based on virtual synchronous wind turbines

Abstract

The present invention discloses a method for dynamic aggregation modeling of a wind farm based on a virtual synchronous wind turbine generator set, comprising the following steps: using a virtual synchronous wind turbine model to simulate the external characteristics of a traditional synchronous generator; characterizing the electromechanical dynamic equations and terminal voltage steady-state equations in each synchronous wind turbine model after parallel confluence using a unified set of aggregation equations; providing a solution method for the virtual inertia, damping coefficient, and virtual synchronous impedance that should be set for the synchronous wind turbine model; and obtaining the wind turbine model's grid-connected frequency and voltage dynamic response characteristics that are consistent under the constraints of the virtual inertia and damping coefficient. The present invention proposes an aggregation method for virtual synchronous wind turbines within an electromechanical time constant through small-signal modeling of the unit, derives the terminal voltage equation, virtual synchronous shaft equation, and aggregation formula after parallel confluence of the wind turbine units, and characterizes the electromechanical dynamic equations and terminal voltage steady-state equations in the synchronous wind turbine model using a unified set of aggregation equations.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind farm control and modeling, and in particular relates to a wind farm dynamic aggregation modeling method based on a virtual synchronous wind turbine generator set. Background Art

[0002] To effectively address increasingly severe energy and environmental challenges, wind power generation is rapidly developing around the world. As the impact of large numbers of grid-connected wind farms on power system stability grows, wind farm dynamic aggregation models are attracting increasing attention from experts and scholars. Because wind farms typically consist of a large number of wind turbines connected in parallel, the spatial distribution and control variability between the units is significant. To avoid excessively high system order and nonlinearity, equivalent modeling of the paralleled units is necessary when studying the impact of grid-connected wind farms on power systems.

[0003] Currently, commonly used aggregation equivalence methods include capacity-weighted, interface-fitting, and coherent equivalence. The capacity-weighted method, based on the principle of parameter consistency, uses the capacity of each turbine as a weight, equating turbines of the same model and control mode to a single wind turbine. However, this method cannot effectively reduce the system order in wind farms with uneven wind speeds, wide geographical distribution, and a large number of wind turbine models. The interface-fitting method, to maximize order reduction, equates the characteristics of the wind farm's grid connection point (PCC) to a single generator, transforming the aggregation problem into a motor parameter determination and optimization problem. However, a single set of optimization parameters often fails to meet the requirements of complex wind farm operating conditions. The coherent equivalence method, based on the principle of synchronous generator aggregation, clusters turbines with similar terminal voltage and stator frequency variations. However, wind turbine types vary greatly, and their actual frequency and voltage dynamics are determined by their respective converter control parameters. Current maximum power point tracking (MPPT) control based on grid voltage vector orientation results in a lack of uniform dynamic response patterns across turbines. Under traditional vector control, the dynamic differences between each unit are obvious, and the wind speed and direction in the wind farm vary greatly. It is difficult to aggregate the wind farm into one unit and characterize the grid-connected characteristics under different operating conditions through clustering and grouping methods, making it difficult to provide an effective model for grid stability analysis under high wind energy penetration.

[0004] In recent years, the concept of virtual synchronous generator (VSG) has gradually gained attention. Among them, VSG wind turbines have the ability to simulate the external characteristics of synchronous generators, providing a control basis for the dynamic aggregation of wind farms based on traditional power systems. At the same time, the order and physical characteristics of the aggregated wind farm model are close to those of traditional synchronous motors, enhancing the integration of power electronic power supplies and power systems dominated by synchronous machines.

[0005] Taking into account that the power information of each wind turbine in the actual system can be measured locally in real time, and the steady-state power of each turbine can be obtained through SCADA low-speed communication, the wind turbines can be dynamically aggregated on this basis, so that the dynamic response characteristics of the grid-connected frequency and voltage of each wind turbine are consistent under the constraints of virtual inertia and damping coefficient, and the electromechanical dynamic process can be constrained by the control model and meet the aggregation conditions, thereby providing a feasible solution to the dynamic aggregation problem of wind farms during grid disturbances. Summary of the Invention

[0006] The present invention is proposed to solve the problems existing in the prior art, and its purpose is to provide a wind farm dynamic aggregation modeling method based on virtual synchronous wind turbines.

[0007] The technical solution of the present invention is: a wind farm dynamic aggregation modeling method based on a virtual synchronous wind turbine generator set, comprising the following steps:

[0008] A. Use a virtual synchronous wind turbine model to simulate the external characteristics of a traditional synchronous generator;

[0009] B. The electromechanical dynamic equations in the parallel-connected synchronous wind turbine models are represented by a unified set of aggregate equations;

[0010] C. The steady-state voltage equations of the terminal of each synchronous wind turbine model after parallel confluence are represented by a unified aggregate equation group;

[0011] D. The solution method for the virtual inertia, damping coefficient and virtual synchronous impedance that should be set for each synchronous wind turbine model is given;

[0012] E. The grid-connected frequency and voltage dynamic response characteristics of each synchronous wind turbine model obtained are consistent under the constraints of virtual inertia and damping coefficient.

[0013] Furthermore, in step A, a virtual synchronous wind turbine model is used to simulate the external characteristics of a traditional synchronous generator. Each synchronous wind turbine model in the wind farm operates in a unity power factor output mode, and each synchronous wind turbine model is connected in parallel.

[0014] Furthermore, in step C, the steady-state equations of the terminal voltages in the parallel-connected synchronous wind turbine models are represented by a unified set of aggregate equations. The specific process is as follows:

[0015] First, Unit 1 and Unit 2 are two units located at the ends of the busbar;

[0016] Then, using the MPPT power values ​​P1 and P2 of unit 1 and unit 2 at the current wind speed as the weights of the equation, the terminal voltage expression at the power confluence of the two units is obtained.

[0017] Then, the terminal voltage expression u1 is as follows:

[0018] P1·u1=-P1·j(X s1 +X l1 )i1+P1·e o1

[0019] P2·u1=-P2·jX s2 i2+P2·e o2

[0020] Where, X s is the equivalent impedance at the machine end;

[0021] X lk is the bus impedance corresponding to the kth unit;

[0022] e ok is the terminal potential of the kth wind turbine;

[0023] i k is the output current of the kth unit;

[0024] Finally, the aggregation voltage equation is obtained based on the terminal voltage expression.

[0025] Furthermore, the power convergence point of the unit 1 and the unit 2 is the first convergence point in the convergence line.

[0026] Furthermore, the aggregation voltage equation is obtained based on the terminal voltage expression. The specific process is as follows:

[0027] First, the impedance matching degree of the parallel group determines the dynamic aggregation of its terminal voltage. Then, let P1(X s1 +X l1 )=P2X s2 , the aggregation voltage equation is:

[0028]

[0029] Furthermore, step C represents the terminal voltage steady-state equations in the models of the synchronous wind turbines after parallel confluence with a unified aggregate equation group, and also includes obtaining the voltage equation of any confluence point based on the above aggregate voltage equations.

[0030] Furthermore, the voltage equation of any busbar is obtained by the following process:

[0031] First, assuming that there are n generators on each busbar, the voltage aggregation condition at the h-1th busbar is:

[0032]

[0033] Where, 2≤h≤n;

[0034] Then, since the system has only n units, the voltage equation for any busbar can be obtained:

[0035]

[0036] Furthermore, step C represents the terminal voltage steady-state equations of each synchronous wind turbine model after parallel converging with a unified aggregate equation group, which also includes the following:

[0037] First, the equivalent synchronous reactance at the grid connection point needs to be hour;

[0038] Then, according to the aggregation conditions, the equivalent reactance X of each unit is calculated in reverse order. s It is necessary to satisfy the iterative law formula, which makes the total voltage drop caused by the output power of each unit on the equivalent output impedance and bus line reflect the wind farm grid-connected power. Voltage drop on .

[0039] Furthermore, the iterative rule formula is as follows:

[0040]

[0041] The beneficial effects of the present invention are as follows:

[0042] This invention addresses the diverse frequency and voltage dynamics of various units within existing wind farms. Wind farm aggregation models typically rely solely on steady-state output power for equivalence, lacking an effective dynamic aggregation mechanism. Virtual synchronous wind turbines, however, are capable of simulating the external characteristics of traditional synchronous generators. Their frequency dynamic response is primarily influenced by virtual inertia and damping coefficient, while their voltage dynamic response is determined by excitation regulator parameters. Therefore, the dynamic aggregation methods used for traditional synchronous generators offer some relevance for these types of wind turbines.

[0043] The present invention takes the self-synchronization mechanism of the unit under the complex network of the Kuramoto model as its mathematical basis, and proposes an aggregation method of virtual synchronous wind turbines within the electromechanical time constant through unit small signal modeling, that is, a wind farm dynamic aggregation method based on virtual synchronous wind turbines. The terminal voltage equation, virtual synchronous axis equation and mathematical aggregation formula after multiple wind turbines are connected in parallel are derived, the electromechanical dynamic equations and terminal voltage steady-state equations in the models of each synchronous wind turbine after parallel confluence are represented by a unified aggregation equation group, and a solution method for the virtual inertia, damping coefficient and virtual synchronous impedance that should be set for each wind turbine is given.

[0044] Since the order and physical characteristics of the aggregated wind farm model of the present invention are close to those of traditional synchronous motors, this method also enhances the integration of power electronic power supplies with traditional power systems dominated by synchronous machines, which is conducive to the power system to better analyze the dynamic stability problems of power grids containing large-scale wind farms. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a schematic diagram of the wind farm architecture and its aggregation model in the present invention;

[0046] Figure 2 It is a schematic diagram of a small signal model of a virtual synchronous fan in the present invention;

[0047] Figure 3 It is a comparison diagram of the actual wind farm model and the dynamic aggregation model in the present invention;

[0048] Figure 4 are the short-term wind speed curves corresponding to the busbars in the present invention;

[0049] Figure 5 It is a comparison diagram of the actual wind farm model and the dynamic aggregation model in the present invention. DETAILED DESCRIPTION

[0050] Hereinafter, the present invention will be described in detail with reference to the accompanying drawings and embodiments:

[0051] like Figures 1 to 5 As shown, a wind farm dynamic aggregation modeling method based on virtual synchronous wind turbine generator sets includes the following steps:

[0052] A. Use a virtual synchronous wind turbine model to simulate the external characteristics of a traditional synchronous generator;

[0053] B. The electromechanical dynamic equations in the parallel-connected synchronous wind turbine models are represented by a unified set of aggregate equations;

[0054] C. The steady-state voltage equations of the terminal of each synchronous wind turbine model after parallel confluence are represented by a unified aggregate equation group;

[0055] D. The solution method for the virtual inertia, damping coefficient and virtual synchronous impedance that should be set for each synchronous wind turbine model is given;

[0056] E. The grid-connected frequency and voltage dynamic response characteristics of each synchronous wind turbine model obtained are consistent under the constraints of virtual inertia and damping coefficient.

[0057] In step A, a virtual synchronous wind turbine model is used to simulate the external characteristics of a traditional synchronous generator. Each synchronous wind turbine model in the wind farm operates in a unity power factor output mode and is connected in parallel.

[0058] Step C represents the steady-state voltage equations of the terminal of each synchronous wind turbine model after parallel confluence using a unified aggregate equation group. The specific process is as follows:

[0059] First, Unit 1 and Unit 2 are two units located at the ends of the busbar;

[0060] Then, using the MPPT power values ​​P1 and P2 of unit 1 and unit 2 at the current wind speed as the weights of the equation, the terminal voltage expression at the power confluence of the two units is obtained.

[0061] Then, the terminal voltage expression u1 is as follows:

[0062] P1·u1=-P1·j(X s1 +X l1 )i1+P1·e o1

[0063] P2·u1=-P2·jX s2 i2+P2·e o2

[0064] Where, X s is the equivalent impedance at the machine end;

[0065] X lk is the bus impedance corresponding to the kth unit;

[0066] e ok is the terminal potential of the kth wind turbine;

[0067] i k is the output current of the kth unit;

[0068] Finally, the aggregation voltage equation is obtained based on the terminal voltage expression.

[0069] The power convergence point of the unit 1 and the unit 2 is the first convergence point in the convergence line.

[0070] The aggregation voltage equation is obtained based on the terminal voltage expression. The specific process is as follows:

[0071] First, the impedance matching degree of the parallel group determines the dynamic aggregation of its terminal voltage.

[0072] Then, let P1(X s1 +X l1 )=P2X s2 , the aggregation voltage equation is:

[0073]

[0074] Step C represents the terminal voltage steady-state equations in the parallel-connected synchronous wind turbine models with a unified aggregate equation group, and also includes obtaining the voltage equation of any confluence point based on the above aggregate voltage equations.

[0075] The specific process of obtaining the voltage equation of any confluence point is as follows:

[0076] First, assuming that there are n generators on each busbar, the voltage aggregation condition at the h-1th busbar is:

[0077]

[0078] Where, 2≤h≤n;

[0079] Then, since the system has only n units, the voltage equation for any busbar can be obtained:

[0080]

[0081] Step C represents the terminal voltage steady-state equations of each synchronous wind turbine model after parallel confluence using a unified aggregate equation group, which also includes the following:

[0082] First, the equivalent synchronous reactance at the grid connection point needs to be hour;

[0083] Then, according to the aggregation conditions, the equivalent reactance X of each unit is calculated in reverse order. s It is necessary to satisfy the iterative law formula, which makes the total voltage drop caused by the output power of each unit on the equivalent output impedance and bus line reflect the wind farm grid-connected power. Voltage drop on .

[0084] The iterative rule formula is as follows:

[0085]

[0086] Specifically, step B represents the electromechanical dynamic equations in the parallel-connected synchronous wind turbine models using a unified aggregate equation set. The specific process is as follows:

[0087] First, under the dynamic power per unit system, the frequency characteristics of each virtual synchronous control unit can be expressed by the electromechanical equations in its virtual synchronous machine control model:

[0088]

[0089] Then, in the dynamic power per unit system, the control system of each unit sets the same virtual inertia under the MPPT power value per unit. and damping coefficient At the output mechanical torque and electromagnetic torque Under the action of , the angular acceleration of each unit and the frequency change rate within the same time are equal.

[0090] Finally, let the angular acceleration and frequency change rate after aggregation be α a and Δω a , because the angular velocity and its rate of change satisfy the following conditions in the dynamic process within the same time, the electromechanical equations can be aggregated, as follows:

[0091]

[0092] Specifically, the electromechanical dynamic equations and terminal voltage equations of each wind turbine controlled by virtual synchronous control in combination with step B and step C can be represented by a complete set of electromechanical equations and terminal voltage equations, wherein the electromechanical equations reflect their frequency characteristics, the terminal voltage equations reflect their voltage characteristics, and the machine-end equivalent impedance represents the synchronous reactance in the virtual synchronous machine model, reflecting the dynamic current distribution during system disturbances and the armature reaction of the virtual synchronous machine in steady state. The specific expression is as follows:

[0093]

[0094]

[0095]

[0096] Where u apcc Indicates the voltage at the common connection point.

[0097] Specifically, the relationship between the electromechanical dynamic equation and the steady-state electromagnetic equation is related through the torque expression of each unit, specifically:

[0098]

[0099] where θ o and θ h-1 They are the phase angle of the unit's back electromotive force and the phase angle of the corresponding bus voltage of the unit.

[0100] Specifically, during the dynamic process, the frequency change rate of each unit is close, and the normalized virtual inertia and damping coefficient The terminal voltage characteristic is determined by the normalized excitation adjustment parameter, and the weighting coefficient P h It is represented by the current MPPT power of each h-th unit.

[0101] Specifically, based on the aggregation model, it is also necessary to make equivalent values ​​for the distributed wind turbines and use them as the prime movers of the synchronous generator sets.

[0102] Specifically, since the rotor inertia time constants of wind turbines of the same power level are close and much larger than the active-frequency closed-loop time constant of the virtual synchronous motor control, a set of equivalent inertia time constants H w Characterize the wind turbine rotor as follows:

[0103]

[0104] According to the system capacity S n , rated frequency f n The equivalent inertia J can be determined by the number of rotor pole pairs p w The wind turbine in the system mainly determines the closed-loop time of MPPT, including the power following during wind speed disturbances and the MPPT recovery process after grid-side disturbances.

[0105] The equivalent wind speed is used to represent the wind energy that the wind farm can currently absorb and serves as the input variable of the equivalent wind turbine.

[0106] Example 1

[0107] Under the same operating conditions, an actual synchronous generator with a rated power of 3 MW is used to characterize the two virtual synchronous wind turbines using the PWM average model. The parameters of the synchronous generator sets in the aggregated model are shown in Table 1.

[0108] Table 1 Dynamic aggregation synchronous generator system parameters

[0109]

[0110] Considering that the large power grid is composed of large synchronous machines connected in parallel, a 30MW synchronous generator set is used to replace the infinite power grid model with an initial load of 30MW.

[0111] In the simulation, the corresponding wind speeds of Units 1 and 2 increased from 9m / s and 8.8m / s to 9.5m / s and 9.2m / s respectively at 2s. Since the wind speed variation law in this example is relatively simple, it can be calculated that the equivalent wind speed of the prime mover in the wind farm aggregation model with a rated power of 3MW and a wind rotor equivalent inertia time constant of 5s increased from 8.93m / s to 9.38m / s. The dynamic response of the aggregation model during the grid disturbance was simulated by a 5% load increase at 12s, and the grid-connected power and system frequency of each unit and wind farm were observed. Figure 2As shown in the figure, with increasing wind speed, the system frequency of the actual model increases slightly after 2 seconds. Due to the slow increase in wind turbine rotor speed during this process, the system frequency change rate is small. A 5% load surge at 12 seconds causes a sudden increase in the electromagnetic power of the 30MW synchronous generator representing the power grid. Constrained by the synchronous machine rotor inertia and the virtual inertia of the virtual synchronous machine, the system frequency drops significantly and enters a new stable operating point as the wind turbine resumes MPPT. The aggregated model's frequency support effect on the system is consistent with the actual model.

[0112] Figure 2 Where PI stands for proportional integral controller, ω n Represents the rated angular frequency of the AC system, E s and θ s Indicates the modulation voltage amplitude and phase angle, P ref With P g Represents active power command and actual active power, Q ref With Q g Represents reactive power command and actual reactive power, and Δ represents small signal increment.

[0113] However, in actual wind farms, wind speed fluctuations are random, and it is usually impossible to calculate the accurate equivalent wind speed of the wind farm. Moreover, if all wind turbines use a refined model, the system order will be too high, making it impossible to verify the effectiveness of the aggregation model in large wind farms.

[0114] In Case 2, the MPPT command values ​​of each unit transmitted by the SCADA system replace the output mechanical power of the prime mover in the aggregation model. The output internal potential of the virtual synchronous machine model serves as a controlled voltage source, replacing the wind turbine's PWM model. This demonstrates the feasibility of the proposed aggregation method in a large-scale wind farm using an actual wind farm with 33 wind turbines. The wind farm busbar parameters are shown in the Appendix. The per-unit values ​​of the synchronous generator parameters in the aggregation model are consistent with those in Table 1, with only the rated power increased to 49.5 MW. The main grid is represented by a conventional synchronous generator with a rated capacity of 200 MW, and the initial load is 200 MW.

[0115] The wind speed curve used in the simulation contains a step component that changes in seconds and a random component that changes in Hz. The wind speed curves corresponding to the three busbars A, B, and C in the simulation are as follows: Figure 4As shown. The MPPT data of the three busbars uploaded by the SCADA system are further superimposed as the mechanical drive power of the synchronous generator in the wind farm dynamic aggregation model, and the zero-order holder and delay module are used to simulate the 300ms communication time required to upload data in the SCADA system. In the simulation of the aggregation model, the grid-connected dynamic characteristics of the aggregation model are also verified through the 5% load surge at 2s and the load recovery process at 12s. The grid-connected power waveform of the aggregation model under this working condition is compared with the total grid-connected power of the three busbars at the PCC point in the actual wind farm model. The results are shown as follows. Figure 5 As shown in (a), the system frequency is Figure 5 (b) shown.

[0116] When the grid-side load suddenly increases by 5% at 2s of simulation, all three busbars demonstrate frequency support function. Since the total steady-state output power of busbars A and B before the disturbance is close, the dynamic support power borne by the two busbars is also similar. However, since the steady-state power of busbar C before the disturbance is smaller, the dynamic support power shared by the steady-state power is also smaller. However, the power dynamic adjustment time of the three busbars is basically the same. When the system returns to rated load after 12s, the power dynamic characteristics of the three busbars still maintain a high degree of consistency. At the same time, Figure 5 The comparison curve of total grid-connected power in (a) shows that due to factors such as communication delays caused by the SCADA system and the failure to consider line losses during the convergence of each unit, the output power of the aggregate model and the actual model deviate by less than 10%. However, the maximum power tracking trend remains consistent throughout the simulation. Furthermore, in both simulations, the load disturbance occurs at the same moment, and the dynamic response of each model's output power is completely dependent on its own electromechanical dynamic model constraints. It can be seen that the steady-state power error of the two models at the same moment is small, and the dynamic consistency of the output power of the dynamic aggregate model and the actual wind farm model when participating in frequency support is also high. Figure 5 The system frequency change process in (b) can reflect the contribution of the wind farm to the inertia of the large power grid. During the load surge, the lowest system frequency of the actual model and the aggregated model is 49.98 Hz, while the highest frequency of both reaches 50.07 Hz during the load recovery process. The frequency adjustment time is consistent. The time when the lowest frequency point of the aggregated model occurs is slightly delayed due to the influence of the communication cycle. The above simulation results also show that the dynamic aggregated model better reflects the contribution of the virtual inertia of the wind turbine in the actual model to the power grid.

[0117] In Case 2, although the input mechanical power of the aggregation model requires MPPT data uploaded by the SCADA system, it can be used as the driving power of the synchronous generator in the aggregation model by simply adding it up. The electromechanical dynamic consistency of each unit is guaranteed by its own virtual synchronous machine control. Compared with the aggregation method that requires optimal solution and interface fitting of relevant data in the host computer, this model can more quickly and accurately obtain the electromechanical dynamic characteristics of the wind farm output power. It was also found that even though a certain communication delay was taken into account in Example 2, because the MPPT instruction itself has been filtered by the rotor inertia and its rate of change is less than the fluctuation frequency of the wind speed, the impact of communication delay on the accuracy of the output power of the aggregation model is further reduced. Moreover, because the electromechanical dynamic equations of the aggregation model are the same as the virtual synchronous axis equations, the dynamic frequency response mechanisms of the two models during grid disturbances are similar. Therefore, this aggregation method is suitable for the aggregation modeling of large-scale wind farms.

[0118] This invention addresses the diverse frequency and voltage dynamics of various units within existing wind farms. Wind farm aggregation models typically rely solely on steady-state output power for equivalence, lacking an effective dynamic aggregation mechanism. Virtual synchronous wind turbines, however, are capable of simulating the external characteristics of traditional synchronous generators. Their frequency dynamic response is primarily influenced by virtual inertia and damping coefficient, while their voltage dynamic response is determined by excitation regulator parameters. Therefore, the dynamic aggregation methods used for traditional synchronous generators offer some relevance for these types of wind turbines.

[0119] The present invention takes the self-synchronization mechanism of the unit under the complex network of the Kuramoto model as its mathematical basis, and proposes an aggregation method of virtual synchronous wind turbines within the electromechanical time constant through unit small signal modeling, that is, a wind farm dynamic aggregation method based on virtual synchronous wind turbines. The terminal voltage equation, virtual synchronous axis equation and mathematical aggregation formula after multiple wind turbines are connected in parallel are derived, the electromechanical dynamic equations and terminal voltage steady-state equations in the models of each synchronous wind turbine after parallel confluence are represented by a unified aggregation equation group, and a solution method for the virtual inertia, damping coefficient and virtual synchronous impedance that should be set for each wind turbine is given.

[0120] Since the order and physical characteristics of the aggregated wind farm model of the present invention are close to those of traditional synchronous motors, this method also enhances the integration of power electronic power supplies with traditional power systems dominated by synchronous machines, which is conducive to the power system to better analyze the dynamic stability problems of power grids containing large-scale wind farms.

Claims

1. A wind farm dynamic aggregation modeling method based on virtual synchronous wind turbines, characterized by: The following steps are involved: (A) Using a virtual synchronous wind turbine model to simulate the external characteristics of a traditional synchronous generator; (B) The electromechanical dynamic equations in the parallel-connected synchronous wind turbine models are represented by a unified set of aggregate equations. (C) The steady-state voltage equations at the terminals of the synchronous wind turbine models after parallel connection are represented by a unified set of aggregate equations; (D) The solution method for the virtual inertia, damping coefficient and virtual synchronous impedance that should be set for each synchronous wind turbine model is given; (E) The obtained grid-connected frequency and voltage dynamic response characteristics of each synchronous wind turbine model are consistent under the constraints of virtual inertia and damping coefficient; Step (C) represents the terminal voltage steady-state equations of each synchronous wind turbine model after parallel confluence using a unified aggregate equation group, which also includes the following: First, the equivalent synchronous reactance at the grid connection point needs to be hour; Then, according to the aggregation conditions, the equivalent reactance of each unit is calculated by reverse recursion. X s It is necessary to satisfy the iterative law formula, which makes the total voltage drop caused by the output power of each unit on the equivalent output impedance and bus line reflect the wind farm grid-connected power. Voltage drop across The iterative rule formula is as follows: ； Specifically, step B represents the electromechanical dynamic equations in the parallel-connected synchronous wind turbine models using a unified aggregate equation set. The specific process is as follows: First, under the dynamic power per unit system, the frequency characteristics of each virtual synchronous control unit can be expressed by the electromechanical equations in its virtual synchronous machine control model: ； Then, in the dynamic power per unit system, the control system of each unit sets the same virtual inertia under the MPPT power value per unit. and damping coefficient , at the output mechanical torque and electromagnetic torque Under the action of , the angular acceleration of each unit and the frequency change rate within the same time are equal.

2. The wind farm dynamic aggregation modeling method based on virtual synchronous wind turbine generator sets according to claim 1 is characterized by: In step (A), a virtual synchronous wind turbine model is used to simulate the external characteristics of a traditional synchronous generator. Each synchronous wind turbine model in the wind farm operates in a unity power factor output mode and is connected in parallel.

3. The wind farm dynamic aggregation modeling method based on virtual synchronous wind turbine generator sets according to claim 1 is characterized by: Step (C) represents the terminal voltage steady-state equations of each synchronous wind turbine model after parallel confluence using a unified aggregate equation group. The specific process is as follows: First, Unit 1 and Unit 2 are two units located at the ends of the busbar; Then, use the MPPT power values ​​of unit 1 and unit 2 at the current wind speed P 1. P 2 as the weight of the equation, the terminal voltage expression of the power confluence of the two units is obtained, Then, the terminal voltage expression is u 1 The details are as follows: ； Where, X s is the equivalent impedance at the machine end; X lk For the k The bus impedance corresponding to each unit; e ok For the k The internal potential of each fan; i k For the k Output current of each unit; Finally, the aggregation voltage equation is obtained based on the terminal voltage expression.

4. The wind farm dynamic aggregation modeling method based on virtual synchronous wind turbine generator sets according to claim 3 is characterized by: The power convergence point of the unit 1 and the unit 2 is the first convergence point in the convergence line.

5. The wind farm dynamic aggregation modeling method based on virtual synchronous wind turbine generator sets according to claim 3 is characterized by: The aggregation voltage equation is obtained based on the terminal voltage expression. The specific process is as follows: First, the impedance matching degree of the parallel group determines the dynamic convergence of its terminal voltage. Then, let P 1( X s1+ X l1 )= P 2 X s2 , the aggregation voltage equation is:

6. The wind farm dynamic aggregation modeling method based on virtual synchronous wind turbine generators according to claim 5 is characterized by: Step (C) represents the terminal voltage steady-state equations in the parallel-connected synchronous wind turbine models using a unified aggregate equation group, and also includes obtaining the voltage equation of any confluence point based on the above aggregate voltage equations.

7. The wind farm dynamic aggregation modeling method based on virtual synchronous wind turbine generator sets according to claim 6 is characterized in that: The specific process of obtaining the voltage equation of any confluence point is as follows: First, assume that each bus line has n Unit, then h -The voltage aggregation condition of a busbar is: ； Where, 2≤ h ≤ n ; Then, since the system only has n The voltage equation of any junction point can be obtained for each unit: 。 8. The wind farm dynamic aggregation modeling method based on virtual synchronous wind turbine generator sets according to claim 1 is characterized by: Step (C) represents the terminal voltage steady-state equations of each synchronous wind turbine model after parallel confluence using a unified aggregate equation group, which also includes the following: First, the equivalent synchronous reactance at the grid connection point needs to be hour; Then, according to the aggregation conditions, the equivalent reactance of each unit is calculated by reverse recursion. X s It is necessary to satisfy the iterative law formula, which makes the total voltage drop caused by the output power of each unit on the equivalent output impedance and bus line reflect the wind farm grid-connected power. Voltage drop on .

9. The wind farm dynamic aggregation modeling method based on virtual synchronous wind turbine generator sets according to claim 8, characterized in that: The iterative rule formula is as follows: 。

Images