Virtual impedance control method and system of converter adaptive to parallel operation of multiple wind turbines

Abstract

This invention relates to the field of wind power grid-connected control technology, and provides a method and system for controlling the virtual impedance of a converter adapted to the parallel operation of multiple wind turbines. Instead of using a fixed virtual impedance value, it acquires two key dynamic indicators in real time: the normalized active power characterizing the converter's own operating conditions and the estimated short-circuit ratio characterizing the external power grid. These two dynamic indicators are used as inputs to a fuzzy logic controller, which uses a fuzzy inference mechanism to comprehensively determine the current complex operating conditions, thereby dynamically and online generating adjustment coefficients for virtual resistance and virtual inductance. Based on these adjustment coefficients, the final virtual impedance value is adjusted in real time, enabling it to automatically adapt to power fluctuations and changes in grid strength, ensuring that the wind turbines maintain good damping characteristics and power output performance across a wide operating range.

Description

Technical Field

[0001] This invention relates to the field of wind power grid connection control technology, and in particular to a converter virtual impedance control method and system adapted to the parallel operation of multiple wind turbines. Background Technology

[0002] Large-scale wind farms typically consist of numerous wind turbine generators, which operate in parallel via their respective converters and are connected to the grid's point of common coupling (PCC). To achieve stable operation, reasonable power distribution, and suppression of circulating currents between parallel units in a multi-converter parallel system, virtual impedance control technology has been introduced into the converter's control strategy. This technology simulates a virtual resistor and inductor in the control algorithm, making them equivalently exhibit specific impedance characteristics at the converter output, thereby improving the system's grid connection performance.

[0003] However, most existing virtual impedance control methods use fixed parameters. Specifically, when the wind turbine is operating at low power output, the system is not sensitive to power loss, and the main control objective is to ensure stable operation and suppress potential oscillations. In this case, the system needs a relatively large virtual resistance to provide sufficient damping. When the wind turbine is operating at high power output, the main control objective shifts to maximizing power output and minimizing power loss. In this case, it is desirable to minimize the virtual impedance to reduce internal voltage drop and energy loss. Traditional methods using fixed virtual impedance parameters cannot simultaneously meet the conflicting control requirements under high and low power conditions. Fixed parameters are essentially a compromise, leading to a mismatch between the virtual impedance parameters and the real-time operating point under dynamically changing conditions. This mismatch may limit maximum power output at high power and cause insufficient system damping at low power, affecting the operational stability and power quality of wind farms under complex and variable conditions. Furthermore, the operating conditions of the power grid are not constant, and fixed parameters are also difficult to adapt to changes in external grid characteristics. Summary of the Invention

[0004] The present invention aims to solve at least one of the problems existing in the prior art, and to provide a converter virtual impedance control method and system adapted to the parallel operation of multiple wind turbines.

[0005] One aspect of the present invention provides a converter virtual impedance control method adapted to the parallel operation of multiple wind turbines, comprising: The three-phase voltage at the grid connection point is collected from the voltage transformer at the grid connection point, and the three-phase current at the output of the converter is collected from the current transformer on the output side of the converter. A grid strength assessment is performed on the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain the normalized active power and estimate the short-circuit ratio; The normalized active power and the estimated short-circuit ratio are fuzzified by a fuzzy logic controller to obtain the normalized active power membership vector and the estimated short-circuit ratio membership vector. Based on the fuzzy rule base, fuzzy reasoning is performed on the normalized active power membership vector and the estimated short-circuit ratio membership vector to obtain the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient. Based on the reference virtual resistance and reference virtual inductance, the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient are defuzzified to obtain the virtual resistance value and the virtual inductance value. Voltage reference command correction is performed based on virtual resistance and virtual inductance values.

[0006] Another aspect of the present invention provides a converter virtual impedance control system adapted to the parallel operation of multiple wind turbines, comprising: The voltage and current acquisition module is used to acquire the three-phase voltage at the grid connection point from the voltage transformer at the grid connection point, and to acquire the three-phase current output from the current transformer on the output side of the converter. The grid strength assessment module is used to assess the grid strength of the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain normalized active power and estimate the short-circuit ratio. The parameter fuzzification module is used to fuzzify the normalized active power and the estimated short-circuit ratio through a fuzzy logic controller to obtain the normalized active power membership vector and the estimated short-circuit ratio membership vector. The fuzzy inference module is used to perform fuzzy inference on the normalized active power membership vector and the estimated short-circuit ratio membership vector based on the fuzzy rule base to obtain the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient. The defuzzification module is used to defuzzify the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient based on the reference virtual resistance and the reference virtual inductance to obtain the virtual resistance value and the virtual inductance value. The reference instruction correction module is used to correct the voltage reference instruction based on the virtual resistance value and the virtual inductance value.

[0007] Compared with existing technologies, the present invention provides a converter virtual impedance control method and system adapted to the parallel operation of multiple wind turbines. Instead of using a fixed virtual impedance value, it acquires two key dynamic indicators in real time: the normalized active power characterizing the converter's own operating conditions and the estimated short-circuit ratio characterizing the external power grid. These two dynamic indicators are used as inputs to a fuzzy logic controller. The fuzzy inference mechanism is used to comprehensively determine the current complex operating conditions, thereby dynamically and online generating adjustment coefficients for virtual resistance and virtual inductance. The final virtual impedance value is adjusted in real time based on the adjustment coefficients of virtual resistance and virtual inductance, enabling it to automatically adapt to power fluctuations and changes in power grid strength, ensuring that the wind turbines maintain good damping characteristics and power output performance over a wide operating range. Attached Figure Description

[0008] One or more embodiments are illustrated by way of example with the corresponding pictures in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0009] Figure 1 A flowchart of a converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to an embodiment of the present invention; Figure 2 This is a data flow diagram illustrating a converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to an embodiment of the present invention. Figure 3 A flowchart is provided for a converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to an embodiment of the present invention, which performs grid strength assessment on the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain normalized active power and estimate the short-circuit ratio. Figure 4 This is a flowchart illustrating the virtual resistance and virtual inductance values ​​obtained by defuzzifying the virtual resistance and virtual inductance adjustment coefficients to obtain virtual resistance and virtual inductance values, based on a reference virtual resistance and reference virtual inductance, according to an embodiment of the present invention, for a converter virtual impedance control method adapted to parallel operation of multiple wind turbines. Figure 5 A flowchart of a converter virtual impedance control method adapted to parallel operation of multiple wind turbines according to an embodiment of the present invention; Figure 6 This is a block diagram of a converter virtual impedance control system adapted to the parallel operation of multiple wind turbines according to an embodiment of the present invention. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the various embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details are presented in the various embodiments of the present invention to facilitate a better understanding of the invention. However, the technical solutions claimed in the present invention can be implemented even without these technical details and with various variations and modifications based on the following embodiments. The division of the various embodiments below is for ease of description and should not constitute any limitation on the specific implementation of the present invention. The various embodiments can be combined with and referenced by each other without contradiction.

[0011] As indicated in the specification and claims of this invention, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0012] While this invention makes various references to certain modules in systems according to embodiments of the invention, any number of different modules can be used and run on user terminals and / or servers. The modules described are merely illustrative, and different aspects of the systems and methods may use different modules.

[0013] This invention uses flowcharts to illustrate the operations performed by a system according to embodiments of the invention. It should be understood that preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously, as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0014] Most existing converter virtual impedance control methods employ fixed parameter settings. However, the output power of wind turbine generators fluctuates drastically due to changes in wind speed, and the grid strength at the connection point may also change. This causes fixed virtual impedance parameters to be unable to adapt to dynamically changing operating conditions, resulting in parameter mismatch and consequently affecting system stability and power output efficiency. Therefore, this invention proposes a converter virtual impedance control method adapted to the parallel operation of multiple wind turbines. This method no longer uses fixed virtual impedance values ​​but instead constructs a dynamic adjustment mechanism that responds to changes in operating conditions, ensuring that the virtual impedance parameters always match the current operating state. Specifically, this method first collects the voltage and current information of the converter in real time, and then obtains two key dynamic indicators through a specific grid strength assessment algorithm: one is the normalized active power reflecting the converter's own operating state, and the other is the estimated short-circuit ratio reflecting the characteristics of the external grid. These two indicators, representing the internal operating conditions and the external environment respectively, are then used as the dual-dimensional input to the fuzzy logic controller. The fuzzy logic controller has a pre-set fuzzy rule base and, through fuzzification and fuzzy inference processes, comprehensively judges the current power-grid strength composite operating condition and generates adjustment coefficients for adjusting virtual resistance and virtual inductance. Finally, these adjustment coefficients are used to perform defuzzification calculations with pre-set reference virtual resistance and reference virtual inductance to obtain the final virtual resistance and virtual inductance values, which are then implemented in the converter's control loop to correct the voltage reference command, thereby achieving dynamic adaptive adjustment of the virtual impedance.

[0015] Figure 1 This is a flowchart of a converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to an embodiment of the present invention. Figure 2 This is a data flow diagram illustrating a converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to an embodiment of the present invention. Figure 1 and Figure 2According to an embodiment of the present invention, a converter virtual impedance control method adapted to the parallel operation of multiple wind turbines includes the following steps: S100, acquiring the three-phase voltage at the grid connection point from the voltage transformer at the grid connection point, and acquiring the three-phase output current of the converter from the current transformer on the output side of the converter; S200, performing grid strength assessment on the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain normalized active power and estimated short-circuit ratio; S300, performing parameter fuzzification on the normalized active power and estimated short-circuit ratio through a fuzzy logic controller to obtain normalized... S400: Based on the fuzzy rule base, perform fuzzy inference on the normalized active power membership vector and the estimated short-circuit ratio membership vector to obtain the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient; S500: Based on the reference virtual resistance and reference virtual inductance, defuzzify the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient to obtain the virtual resistance value and the virtual inductance value; S600: Based on the virtual resistance value and the virtual inductance value, perform voltage reference command correction.

[0016] Specifically, in step S100, the three-phase voltage at the grid connection point is collected from the voltage transformer at the grid connection point, and the three-phase output current of the converter is collected from the current transformer on the output side of the converter. It is understood that since the operating conditions of the converter and the characteristics of the external power grid both need to be calculated and evaluated using real-time electrical quantities, the three-phase voltage at the grid connection point and the three-phase output current of the converter are the most fundamental and direct physical quantities for calculating active power and estimating grid strength. Therefore, this invention collects the three-phase voltage at the grid connection point from the voltage transformer at the grid connection point and the three-phase output current of the converter from the current transformer on the output side of the converter to obtain the basic data input required for subsequent adaptive control. This provides accurate and real-time raw measurement values ​​for subsequent normalized active power calculation and short-circuit ratio estimation.

[0017] More specifically, in a specific example of the present invention, the acquisition process in step S100 is implemented as follows: First, a voltage transformer is configured at the point of common coupling where the wind turbine converter connects to the power grid, and a current transformer is configured on the AC output line of the converter. The voltage transformer and the current transformer proportionally convert the high-voltage, high-current electrical signals into low-voltage, low-current analog signals suitable for measurement and processing. Next, these analog signals are acquired by the data acquisition unit and conditioned by necessary signal conditioning circuits, such as anti-aliasing filters. Finally, the conditioned analog signals are synchronously sampled and quantized by a high-precision analog-to-digital converter, converting them into discrete digital signal sequences to obtain the instantaneous values ​​of the three-phase voltage at the grid connection point and the three-phase current output by the converter, for subsequent digital signal processing and calculation by the controller.

[0018] Specifically, in step S200, the grid strength assessment is performed on the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain the normalized active power and the estimated short-circuit ratio. It is understood that the optimal parameters of the converter's virtual impedance depend simultaneously on the converter's own operating conditions and the characteristics of the external grid. These two factors, namely output power and grid strength, change dynamically during actual wind farm operation. Therefore, this invention further performs grid strength assessment on the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain the normalized active power and the estimated short-circuit ratio, thereby obtaining two key input indicators characterizing the internal operating conditions and the external environment in real time and quantitatively. This provides a decision-making basis for the subsequent fuzzy logic controller, enabling it to adaptively adjust parameters according to the current complex operating conditions, solving the problem that fixed parameters cannot accommodate different operating states.

[0019] Figure 3 This document describes a flowchart illustrating a method for controlling the virtual impedance of a converter adapted to the parallel operation of multiple wind turbines, based on an embodiment of the present invention. The flowchart describes a grid strength assessment of the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain normalized active power and estimate the short-circuit ratio. Figure 3 As shown, step S200 includes: S210, performing coordinate transformation on the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain the d-axis voltage, q-axis voltage, d-axis current, and q-axis current; S220, calculating the instantaneous active power based on the d-axis voltage, q-axis voltage, d-axis current, and q-axis current, and normalizing the instantaneous active power to obtain the normalized active power; S230, estimating the grid impedance at the grid connection point using the online disturbance observation method to obtain the grid impedance value; S240, dividing the grid reference impedance and the grid impedance value to obtain the estimated short-circuit ratio.

[0020] In step S210, the three-phase voltage at the grid connection point and the three-phase output current of the converter are transformed using coordinates to obtain the d-axis voltage, q-axis voltage, d-axis current, and q-axis current. It is understood that the acquired three-phase voltage at the grid connection point and the three-phase output current of the converter are time-varying AC signals, making instantaneous power calculation and control loop design directly based on these three-phase AC quantities extremely complex. Therefore, in the technical solution of this invention, the three-phase voltage at the grid connection point and the three-phase output current of the converter are further transformed using coordinates to obtain the d-axis voltage, q-axis voltage, d-axis current, and q-axis current, thereby converting the time-varying AC quantities into DC quantities in a synchronous rotating coordinate system. This greatly simplifies the subsequent power calculation and control algorithm implementation and provides a foundation for achieving decoupled control of active and reactive power.

[0021] More specifically, in a specific example of the present invention, the coordinate transformation in step S210 is implemented as follows: First, the digital signal of the three-phase voltage at the grid connection point is input into a phase-locked loop (PLL). This PLL is used to track the phase and frequency of the grid voltage in real time and outputs a phase angle synchronized with the grid voltage. Next, the collected three-phase voltage at the grid connection point and the three-phase output current of the converter are subjected to Clarke transformation to convert them from a three-phase stationary coordinate system to a two-phase stationary coordinate system, thus obtaining the corresponding... The voltage and current components of the shaft. Finally, the phase angle output by the phase-locked loop, synchronized with the grid voltage, is used. ,right The voltage and current components of the axes are subjected to Park transformation to convert them from a two-phase stationary coordinate system to a two-phase synchronous rotating coordinate system, thereby accurately obtaining the corresponding d-axis voltage, q-axis voltage, d-axis current and q-axis current.

[0022] In step S220, instantaneous active power is calculated based on d-axis voltage, q-axis voltage, d-axis current, and q-axis current, and then normalized to obtain normalized active power. It is understood that the input to the fuzzy logic controller requires a standardized indicator that clearly characterizes the current load level of the converter. However, the absolute value of instantaneous active power is related to the rated capacity of the converter, which is not conducive to constructing universal fuzzy rules. Therefore, in the technical solution of this invention, instantaneous active power is further calculated based on d-axis voltage, q-axis voltage, d-axis current, and q-axis current, and then normalized to obtain normalized active power. This provides a relative load rate indicator that varies within a specific range and is independent of the specific unit capacity. This provides a precise and standardized input for subsequent fuzzification processing, enabling accurate determination of the current active power status.

[0023] More specifically, in a specific example of the present invention, the calculation and normalization process of instantaneous active power is implemented as follows: First, the d-axis voltage, q-axis voltage, d-axis current, and q-axis current obtained by coordinate transformation in step S210 are substituted into the instantaneous active power calculation formula in the synchronous rotating coordinate system. The instantaneous active power output of the converter is calculated. Subsequently, the preset rated active power value of the converter is read from the controller. Finally, the calculated instantaneous active power Divide by the rated active power value This yields a dimensionless normalized active power. The normalized active power The value of fluctuates between 0 and 1, where 0 corresponds to the unloaded state of the strain gauge and 1 corresponds to the fully loaded state of the strain gauge.

[0024] In step S230, the grid impedance at the grid connection point is estimated using an online disturbance observation method to obtain the grid impedance value. It is understood that grid strength is a key external factor determining system stability, and grid impedance is a direct electrical parameter for assessing grid strength; the grid impedance value changes with grid operating mode or topology. Therefore, in the technical solution of this invention, the grid impedance at the grid connection point is further estimated using an online disturbance observation method to obtain the grid impedance value, thereby acquiring raw data characterizing the external grid properties in real time. This provides an accurate intermediate quantity for subsequent calculation and estimation of the short-circuit ratio, ensuring that fuzzy logic control can simultaneously perceive changes in external grid conditions.

[0025] More specifically, in a specific example of the present invention, the implementation process of the online disturbance observation method in step S230 is as follows: First, in the current control loop of the converter, a periodic disturbance signal of a specific frequency and a small amplitude, such as a high-frequency sinusoidal signal, is superimposed on the q-axis current reference command or the d-axis current reference command. Then, the voltage response component of the corresponding frequency caused by the disturbance signal is synchronously measured at the grid connection point. This requires separating the response of the specific frequency from the fundamental wave and other harmonics through a digital filter or Fourier transform. Finally, based on the amplitude and phase of the injected current disturbance and the amplitude and phase of the measured voltage response, Ohm's law is used to calculate in the frequency domain to identify the equivalent grid impedance value of the grid connection point at the disturbance frequency.

[0026] In step S240, the grid reference impedance and the grid impedance value are divided to obtain the estimated short-circuit ratio. It is understood that the estimated grid impedance value is an absolute electrical parameter, its magnitude depending on the system voltage level, and is not suitable for constructing universal control rules. The short-circuit ratio, on the other hand, is a standardized dimensionless indicator widely used in power systems to characterize grid strength. Therefore, in the technical solution of this invention, the grid reference impedance and the grid impedance value are further divided to obtain the estimated short-circuit ratio, thereby converting the estimated absolute impedance value into a relative and standardized measure of grid strength. This provides a second key input parameter for the subsequent fuzzy logic controller, enabling it to assess the strength of the grid based on a unified standard.

[0027] More specifically, in a specific example of the present invention, the calculation process for estimating the short-circuit ratio is implemented as follows: First, in the parameter storage area of ​​the converter controller, the rated voltage and rated capacity values ​​of the system are preset. Based on these two rated values, the grid reference impedance of the system is calculated and stored as a fixed parameter of the system. Next, the grid impedance value obtained in step S230 is acquired in real time. Finally, the constant grid reference impedance is used as the dividend, and the real-time estimated grid impedance value is used as the divisor to perform a division operation. The result of the operation is the current estimated short-circuit ratio, which is then transmitted to the subsequent fuzzy logic control flow.

[0028] Specifically, in step S300, the normalized active power and estimated short-circuit ratio are parameterized by a fuzzy logic controller to obtain a normalized active power membership vector and an estimated short-circuit ratio membership vector. It is understood that the normalized active power and estimated short-circuit ratio are precise numerical quantities, while subsequent fuzzy inference processes require decision-making based on fuzzy language concepts such as high, low, strong, and weak. Therefore, in the technical solution of this invention, the normalized active power and estimated short-circuit ratio are further parameterized by a fuzzy logic controller to obtain normalized active power membership vector and estimated short-circuit ratio membership vector, thereby converting precise numerical inputs into linguistic information that the fuzzy logic controller can understand and process. This allows for the quantification of the degree of conformity of the current precise operating condition with various preset fuzzy concepts, providing standard input for subsequent fuzzy inference.

[0029] More specifically, in a specific example of the present invention, the normalized active power and the estimated short-circuit ratio are parameterized by a fuzzy logic controller to obtain the normalized active power membership vector and the estimated short-circuit ratio membership vector, including: mapping the normalized active power and the estimated short-circuit ratio into fuzzy language based on a fuzzy language set and a membership mapping function to obtain the normalized active power membership vector and the estimated short-circuit ratio membership vector.

[0030] More specifically, the parameter fuzzification process is implemented based on a pre-defined fuzzy language set and membership mapping functions. For normalized active power, a fuzzy language set is defined, for example, containing three fuzzy subsets: {low, medium, high}. Simultaneously, membership mapping functions are assigned to these three fuzzy subsets, such as trapezoidal functions or trigonometric functions. These functions define the membership degree of the precise normalized active power value from 0 to 1 in the concepts of low, medium, and high. When a real-time normalized active power value is input, it is calculated using these three membership mapping functions to obtain a normalized active power membership vector containing three components. Similarly, for the estimated short-circuit ratio, a fuzzy language set is defined, for example, containing three fuzzy subsets: {weak, medium, strong}, and corresponding membership mapping functions are configured for these three subsets. When a real-time estimated short-circuit ratio is input, it is also mapped to obtain its membership degree in the concepts of weak, medium, and strong, thereby generating an estimated short-circuit ratio membership vector.

[0031] Specifically, in step S400, based on the fuzzy rule base, fuzzy inference is performed on the normalized active power membership vector and the estimated short-circuit ratio membership vector to obtain the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient. It is understood that step S300 only obtains the membership degrees of the normalized active power and the estimated short-circuit ratio on their respective fuzzy language sets, and has not yet formed a specific control decision. The adaptive adjustment of the system requires a core decision-making mechanism that can simulate the experience of engineers and determine the adjustment direction of the virtual impedance parameters based on different operating condition combinations. Therefore, in the technical solution of this invention, fuzzy inference is further performed on the normalized active power membership vector and the estimated short-circuit ratio membership vector based on the fuzzy rule base to obtain the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient. This allows the input operating condition membership information to be converted into a fuzzy control signal representing the output adjustment direction according to a preset expert control strategy. In this way, complex, multi-objective control logic (such as prioritizing stability under low-power weak network conditions and prioritizing loss reduction under high-power strong network conditions) can be solidified into a series of clear rules, enabling intelligent and adaptive adjustment of virtual impedance.

[0032] More specifically, in a specific example of the present invention, the fuzzy inference process in step S400 is implemented as follows: First, a fuzzy rule base containing multiple IF-AND-THEN rules is pre-set in the fuzzy logic controller. The input premise of this fuzzy rule base is the fuzzy language state of normalized active power and estimated short-circuit ratio, and the output conclusion is the fuzzy language setting of virtual resistance adjustment coefficient and virtual inductance adjustment coefficient. For example, the fuzzy rule base may contain the following rules: when the normalized active power is determined to be low and the estimated short-circuit ratio is determined to be weak, the virtual resistance adjustment coefficient is set to large and the virtual inductance adjustment coefficient is set to small to enhance the damping of the system under harsh operating conditions; when the normalized active power is determined to be high and the estimated short-circuit ratio is determined to be strong, the virtual resistance adjustment coefficient is set to small and the virtual inductance adjustment coefficient is set to medium to reduce power loss under full load. This fuzzy rule base can also include other rules, such as: when the normalized active power is determined to be high and the estimated short-circuit ratio is determined to be weak, the virtual resistance adjustment coefficient is set to large and the virtual inductance adjustment coefficient is set to small, in order to simultaneously consider the stability under weak grid conditions and the voltage drop of high power output; when the normalized active power is determined to be low and the estimated short-circuit ratio is determined to be strong, the virtual resistance adjustment coefficient is set to medium and the virtual inductance adjustment coefficient is set to small, in order to balance damping and loss under strong grid and low power conditions; when the normalized active power is determined to be low and the estimated short-circuit ratio is determined to be medium, the virtual resistance adjustment coefficient is set to medium and the virtual inductance adjustment coefficient is set to small, in order to maintain basic damping and reduce voltage drop; when the normalized active power is determined to be medium and the estimated short-circuit ratio is determined to be strong ... When the power is weak, the virtual resistance adjustment factor is set to large and the virtual inductance adjustment factor is set to small to enhance stability under weak grid conditions. When the normalized active power is determined to be medium and the estimated short-circuit ratio is determined to be medium, the virtual resistance adjustment factor and the virtual inductance adjustment factor are both set to medium to strike a balance between stability and losses. When the normalized active power is determined to be medium and the estimated short-circuit ratio is determined to be strong, the virtual resistance adjustment factor is set to small and the virtual inductance adjustment factor is set to medium to improve efficiency while maintaining dynamic performance. When the normalized active power is determined to be high and the estimated short-circuit ratio is determined to be medium, the virtual resistance adjustment factor is set to medium and the virtual inductance adjustment factor is set to small to control voltage drop and reduce losses under medium grid conditions.During real-time operation, the fuzzy inference engine acquires the input normalized active power membership vector and the estimated short-circuit ratio membership vector. For each rule in the fuzzy rule base, it calculates the activation strength of its preconditions. This activation strength is obtained by performing a bitwise AND operation on the two input membership components, for example, taking the minimum value. Then, this activation strength is used to adjust the output fuzzy set corresponding to the rule, for example, large, small, or medium. Finally, the output fuzzy sets activated by all rules are merged to form a total, synthetic fuzzy output result, namely the fuzzy output of the virtual resistance adjustment coefficient and the fuzzy output of the virtual inductance adjustment coefficient to be resolved.

[0033] Specifically, in step S500, based on the reference virtual resistance and reference virtual inductance, the virtual resistance adjustment coefficient and virtual inductance adjustment coefficient are defuzzified to obtain the virtual resistance value and virtual inductance value. It is understood that the virtual resistance adjustment coefficient and virtual inductance adjustment coefficient obtained from the fuzzy inference process are synthesized fuzzy sets, representing a fuzzy language output characterizing the adjustment tendency, rather than precise values ​​that the controller can directly execute. Therefore, in the technical solution of this invention, the virtual resistance adjustment coefficient and virtual inductance adjustment coefficient are further defuzzified based on the reference virtual resistance and reference virtual inductance to obtain the virtual resistance value and virtual inductance value, thereby converting the fuzzy control decision into a precise final parameter that can be used for subsequent control loop calculations. This provides real-time, quantified virtual impedance parameters for subsequent virtual impedance drop calculations, ensuring the final realization of adaptive adjustment.

[0034] Figure 4 This document describes a flowchart illustrating the process of defuzzifying the virtual resistance and virtual inductance adjustment coefficients to obtain virtual resistance and virtual inductance values, based on a reference virtual resistance and reference virtual inductance, as part of a converter virtual impedance control method adapted to multi-fan parallel operation according to an embodiment of the present invention. Figure 4 As shown, step S500 includes: S510, defuzzifying the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient using the centroid method to obtain virtual resistance adjustment coefficient data and virtual inductance adjustment coefficient data; S520, calculating the virtual resistance value using the virtual resistance calculation formula; S530, calculating the virtual inductance value using the virtual inductance calculation formula.

[0035] In step S510, the virtual resistance adjustment coefficient and virtual inductance adjustment coefficient are defuzzified using the centroid method to obtain virtual resistance adjustment coefficient data and virtual inductance adjustment coefficient data. It is understood that the virtual resistance adjustment coefficient and virtual inductance adjustment coefficient obtained from fuzzy inference are fuzzy sets synthesized after the combined effect of all activated rules, and not precise values ​​that can be directly used for numerical calculation. Therefore, in the technical solution of this invention, the virtual resistance adjustment coefficient and virtual inductance adjustment coefficient are further defuzzified using the centroid method to obtain virtual resistance adjustment coefficient data and virtual inductance adjustment coefficient data, thereby converting the fuzzy, qualitative control conclusion into clear, quantitative adjustment coefficient values. This provides a unique, precise value representing the comprehensive intent of all inference rules for subsequent virtual impedance parameter calculation.

[0036] More specifically, in a specific example of the present invention, the implementation process of step S510, centroid method defuzzification, is as follows: First, the fuzzy output of the virtual resistance adjustment coefficient synthesized in the previous step is represented as a specific function graph in the output universe of discourse. The geometric center of the area covered by this function graph is calculated using the centroid method. In the digital controller, this process first discretizes the output universe of discourse, then calculates the sum of the products of all discrete point positions and their corresponding membership degrees, and then divides it by the sum of the membership degrees of all discrete points, i.e., performs a weighted average calculation. The result of this calculation is determined as the unique and accurate virtual resistance adjustment coefficient data. Subsequently, the same centroid method calculation process is performed on the fuzzy output of the virtual inductance adjustment coefficient to obtain accurate virtual inductance adjustment coefficient data. For example, when the system is in the aforementioned high-power, weak grid condition, the result of fuzzy inference will cause the fuzzy output of the virtual resistance adjustment coefficient to be biased towards a large fuzzy subset in the universe of discourse. At this time, the geometric center calculated by the centroid method will correspond to a large value, such as 1.8. Conversely, under high-power, strong grid conditions, the fuzzy output will be biased towards a small fuzzy subset, and the centroid method calculation result will correspond to a smaller value, such as 0.4. This method ensures that the final output adjustment coefficient data can accurately reflect the comprehensive decision-making intent of fuzzy inference.

[0037] Step S520 calculates the virtual resistance value using the virtual resistance calculation formula, specifically: the virtual resistance value is calculated using the following virtual resistance calculation formula: ; in, This is the virtual resistance adjustment coefficient data. Reference virtual resistance, This refers to the virtual resistance value. It is understood that the virtual resistance adjustment coefficient data obtained through defuzzification is only an intermediate coefficient characterizing the adjustment range, and not the final virtual resistance parameter that can be directly used for control loop calculation. Therefore, in the technical solution of this invention, the virtual resistance value is further calculated using a virtual resistance calculation formula. Specifically, this formula multiplies the virtual resistance adjustment coefficient data with a preset reference virtual resistance, thereby applying the adaptive adjustment decision output by the fuzzy logic controller to the system's nominal reference parameter. This generates a precise virtual resistance value that dynamically adjusts with the current operating conditions, for use in subsequent virtual voltage drop calculations and voltage reference command corrections. For example, a reference virtual resistance under nominal operating conditions can be preset. ,like When the system is operating under high power and weak grid conditions, the virtual resistance adjustment coefficient data obtained from the previous centroid method will be... The value could be 1.8, in which case the virtual resistance value calculated in this step is... Then it is This is an increased damping value to cope with weak grids. Conversely, in high-power, strong grid conditions, It could be 0.4, then the calculated value is... for This is a damping value reduced to decrease losses. This method enables the virtual resistance parameter to be adaptively generated according to the operating conditions.

[0038] Step S530 calculates the virtual inductance value using the virtual inductance calculation formula, specifically: the virtual inductance value is calculated using the following virtual inductance calculation formula: ; in, This is the virtual inductance adjustment coefficient data. As a reference virtual inductor, This refers to the virtual inductance value. It is understood that the virtual inductance adjustment coefficient data obtained through defuzzification is merely an intermediate coefficient representing the adjustment range, rather than the final virtual inductance parameter that can be directly used for control loop calculation. Therefore, in the technical solution of this invention, the virtual inductance value is further calculated using a virtual inductance calculation formula. Specifically, this formula involves multiplying the virtual inductance adjustment coefficient data with a preset reference virtual inductance, thereby applying the adaptive adjustment decision output by the fuzzy logic controller to the system's nominal reference parameter. This generates a precise virtual inductance value that dynamically adjusts with the current operating conditions, which, together with the virtual resistance value obtained in the previous step, provides complete parameter input for subsequent virtual voltage drop calculations and voltage reference command corrections. Continuing the previous example, a reference virtual inductance can be preset. ,like When the system operates under high power and weak grid conditions, the goal of fuzzy inference is to reduce the additional reactance voltage drop to ensure power delivery capability. In this case, the virtual inductance regulation coefficient data obtained by the centroid method... It might be a small value, such as 0.4, then the virtual inductance value calculated in this step... Then it is Conversely, under high-power, high-voltage grid conditions, It could be a medium value, such as 1.0, then the calculated... for Thus, the product generated in this step... Compared with the previous step Together, they form a virtual impedance parameter pair adapted to the current operating conditions.

[0039] Specifically, in step S600, voltage reference command correction is performed based on virtual resistance and virtual inductance values. It is understood that the virtual resistance and virtual inductance values ​​calculated in the previous step are the final execution parameters of the adaptive control of this invention. These parameters need to be applied to the converter's control loop to produce an actual control effect. Therefore, in the technical solution of this invention, voltage reference command correction is further performed based on virtual resistance and virtual inductance values ​​to convert the dynamically generated impedance parameters into a correction amount for the original control command of the converter. This enables the converter's final control behavior to equivalently exhibit virtual impedance characteristics at the output that are adapted to the current operating conditions, thereby achieving the ultimate goal of adaptive stability control.

[0040] Figure 5 This is a flowchart illustrating voltage reference command correction based on virtual resistance and virtual inductance values ​​in a converter virtual impedance control method adapted to parallel operation of multiple wind turbines according to an embodiment of the present invention. Figure 5 As shown, step S600 includes: S610, calculating the d-axis virtual voltage drop and q-axis virtual voltage drop based on the virtual resistance value, virtual inductance value, and real-time d-axis current and q-axis current; S620, filtering out the d-axis virtual voltage drop and q-axis virtual voltage drop from the original voltage reference command of the outer loop controller to obtain the final voltage reference command; S630, inputting the final voltage reference command into the converter current inner loop controller to drive the pulse width modulation module to generate a drive signal for adjusting the virtual impedance of the converter.

[0041] In step S610, the d-axis virtual voltage drop and q-axis virtual voltage drop are calculated based on the virtual resistance value, virtual inductance value, real-time d-axis current, and q-axis current. Specifically, this includes calculating the d-axis virtual voltage drop and q-axis virtual voltage drop using the following virtual voltage drop calculation formula: ; ; in, For the d-axis virtual pressure drop, This is a virtual resistance value. For real-time d-axis current, This is a virtual inductance value. This represents the real-time q-axis current. The angular frequency of the power grid. This refers to the virtual voltage drop along the q-axis. It's understood that the virtual resistance and inductance values ​​generated in the previous step are parameters characterizing the desired impedance characteristics, and their control effect needs to be achieved by calculating and compensating for a corresponding voltage drop in the converter control loop. Therefore, in the technical solution of this invention, the virtual voltage drop along the d-axis and q-axis is further calculated based on the virtual resistance and inductance values, as well as the real-time d-axis and q-axis currents. This allows for the real-time calculation of the voltage components that need correction on the d-axis and q-axis according to the virtual impedance electrical model in the synchronous rotating coordinate system. Specifically, this calculation includes: subtracting the grid angular frequency and the product of the virtual inductance and the real-time q-axis current from the product of the virtual resistance value and the real-time d-axis current to obtain the virtual voltage drop along the d-axis; and adding the product of the virtual resistance value and the real-time q-axis current to obtain the virtual voltage drop along the q-axis. For example, when the system is operating under high power and weak grid conditions, the previous steps will generate a larger... like and a smaller like Then these parameters can be used in conjunction with the measured height. and The above formula is used to calculate a component with strong damping characteristics. Large contribution and low reactance voltage drop Small contribution and Conversely, under high-power, high-voltage grid conditions, It will be smaller than , It will be moderate This new pair of parameters will then be used to calculate another set of voltage drop values, which represent lower damping and lower losses. This provides a precise, dynamically changing voltage drop compensation signal for subsequent voltage reference command correction steps, enabling the control system to accurately achieve the desired adaptive impedance characteristics.

[0042] In step S620, the d-axis virtual voltage drop and q-axis virtual voltage drop are filtered out from the original voltage reference command of the outer loop controller to obtain the final voltage reference command. It is understood that the d-axis virtual voltage drop and q-axis virtual voltage drop calculated in the previous step are only the voltage correction amounts that need to be compensated, while the original voltage reference command generated by the conventional power outer loop controller of the converter does not include this adaptive impedance characteristic. Therefore, in the technical solution of this invention, the d-axis virtual voltage drop and q-axis virtual voltage drop are further filtered out from the original voltage reference command of the outer loop controller to obtain the final voltage reference command, thereby applying the dynamically calculated voltage drop compensation amount to the core control command of the converter. This allows the voltage command ultimately fed into the inner current loop to change, forcing the converter's output behavior to externally equivalently exhibit virtual resistance and virtual inductance values ​​that match the current operating conditions, i.e., normalized active power and estimated short-circuit ratio, thereby achieving a closed loop for the entire adaptive control.

[0043] More specifically, in a specific example of the present invention, filtering out the d-axis virtual voltage drop and q-axis virtual voltage drop from the original voltage reference command of the outer loop controller to obtain the final voltage reference command includes: subtracting the d-axis voltage command from the d-axis virtual voltage drop in the original voltage reference command to obtain the corrected d-axis voltage command; subtracting the q-axis voltage command from the q-axis virtual voltage drop in the original voltage reference command to obtain the corrected q-axis voltage command; and combining the corrected d-axis voltage command and the corrected q-axis voltage command into the final voltage reference command.

[0044] Accordingly, the d-axis voltage command in the original voltage reference command is subtracted from the d-axis virtual voltage drop to obtain the corrected d-axis voltage command. It is understood that the original d-axis voltage command generated by the converter's outer loop controller is an ideal command that does not include adaptive impedance characteristics, while the d-axis virtual voltage drop calculated in the previous step precisely represents the equivalent d-axis voltage correction required under the current operating conditions (e.g., high-power weak grid conditions). Therefore, in the technical solution of this invention, the d-axis voltage command in the original voltage reference command is further subtracted from the d-axis virtual voltage drop to apply the adaptively calculated d-axis voltage drop component to the d-axis control command channel. This generates a corrected d-axis voltage command, which, after being sent to the inner current loop, forces the actual output of the converter to exhibit impedance characteristics determined by dynamically adjusted impedance in the d-axis direction.

[0045] More specifically, in a specific example of the present invention, the process of filtering out the d-axis virtual voltage drop is performed in real time in the digital controller of the converter. Within a control cycle, the controller first obtains the original d-axis voltage command from the output of the power outer loop (e.g., DC voltage control loop or active power control loop). Simultaneously, the controller obtains the d-axis virtual voltage drop value calculated based on the currently adaptively generated virtual resistance value, virtual inductance value, and measured d-axis and q-axis currents from the virtual voltage drop calculation step. Subsequently, the arithmetic logic unit within the controller performs a subtraction operation, subtracting the d-axis virtual voltage drop from the original d-axis voltage command. The result of this subtraction operation is determined as the corrected d-axis voltage command and is immediately transmitted to the d-axis input of the current inner loop controller as one of the target reference voltage components for d-axis current tracking within this control cycle.

[0046] Accordingly, the q-axis voltage command in the original voltage reference command is subtracted from the q-axis virtual voltage drop to obtain the corrected q-axis voltage command. It is understood that the original q-axis voltage command generated by the converter's outer-loop controller, such as the reactive power or AC voltage controller, is also an ideal command that does not include adaptive impedance characteristics, while the q-axis virtual voltage drop calculated in the previous step precisely represents the equivalent q-axis voltage correction required under the current operating conditions. Therefore, in the technical solution of this invention, the q-axis voltage command in the original voltage reference command is further subtracted from the q-axis virtual voltage drop to apply the adaptively calculated q-axis voltage drop component to the q-axis control command channel. This generates a corrected q-axis voltage command, which, in conjunction with the corrected d-axis voltage command obtained in the previous step, forces the actual output of the converter to also exhibit impedance characteristics determined by dynamically adjusted virtual resistance and virtual inductance values ​​in the q-axis direction.

[0047] More specifically, in a particular example of the invention, the process of filtering out the q-axis virtual voltage drop is performed in real time, synchronously with the d-axis correction, in the digital controller of the converter. Within a control cycle, the controller first obtains the original q-axis voltage command from the output of the power outer loop, such as the reactive power control loop or the AC voltage control loop. Simultaneously, the controller obtains the q-axis virtual voltage drop value calculated based on the currently adaptively generated virtual resistance value, virtual inductance value, and the measured d-axis and q-axis currents from the virtual voltage drop calculation step. Subsequently, the arithmetic logic unit within the controller performs a subtraction operation, subtracting the q-axis virtual voltage drop from the original q-axis voltage command. The result of this subtraction is determined as the corrected q-axis voltage command and is immediately transmitted to the q-axis input of the current inner loop controller as one of the target reference voltage components for q-axis current tracking within that control cycle.

[0048] Accordingly, the corrected d-axis voltage command and the corrected q-axis voltage command are combined into the final voltage reference command. It is understood that the corrected d-axis voltage command and the corrected q-axis voltage command are the simultaneous input quantities required by the converter current inner loop controller in the synchronous rotating coordinate system, and together they constitute a unique voltage reference vector. Therefore, in the technical solution of this invention, the corrected d-axis voltage command and the corrected q-axis voltage command are further combined into the final voltage reference command to form a complete dq coordinate system control command that already includes adaptive virtual impedance correction. This ensures that the subsequent current inner loop controller receives a set of accurate and complete d-axis and q-axis control targets in each control cycle, enabling it to accurately track the corrected voltage reference.

[0049] More specifically, in a specific example of the present invention, the synthesis process of the corrected d-axis voltage command and the corrected q-axis voltage command is executed in the digital controller of the converter. Within the same control cycle, the controller combines the value of the corrected d-axis voltage command with the value of the corrected q-axis voltage command as a data pair or vector. This command data pair containing both d-axis and q-axis components is defined as the final voltage reference command. This final voltage reference command is then transmitted to the converter current inner loop controller, wherein the corrected d-axis voltage command is sent to the d-axis control channel input of the current inner loop, and the corrected q-axis voltage command is sent to the q-axis control channel input of the current inner loop.

[0050] In step S630, the final voltage reference command is input to the converter current inner loop controller to drive the pulse width modulation module to generate a drive signal for adjusting the converter's virtual impedance. It is understood that the final voltage reference command is a final digital control target that includes adaptive virtual impedance correction. This final digital control target must be converted into the physical switching action of the converter power switches to truly achieve the desired impedance characteristics at the converter output. Therefore, in the technical solution of this invention, the final voltage reference command is further input to the converter current inner loop controller to drive the pulse width modulation module to generate a drive signal for adjusting the converter's virtual impedance, thereby converting the adaptively corrected voltage command into a high-frequency switching gate drive pulse. This enables the converter's power semiconductor devices to accurately switch on and off according to the command containing adaptive virtual impedance information, thus actually exhibiting a virtual impedance characteristic at the converter's AC output that dynamically adjusts according to the operating conditions, i.e., normalized active power and estimated short-circuit ratio.

[0051] More specifically, in a specific example of the present invention, the generation process of the drive signal is implemented as follows: First, the final voltage reference command, i.e., the corrected d-axis voltage command and the corrected q-axis voltage command, is used as the reference input of the converter current inner loop controller. The current inner loop controller compares this reference input with the measured d-axis current and q-axis current, and calculates the d-axis and q-axis modulated voltage signals required to achieve the voltage target through a proportional-integral regulator. Next, the d-axis and q-axis modulated voltage signals are sent to the pulse width modulation module, which performs an inverse coordinate transformation to convert the dq-axis modulated signals into a three-phase modulated waveform, and compares it with a high-frequency triangular carrier wave to obtain the corresponding comparison result. Finally, the comparison result generates multiple high and low level pulse width modulation signals, which are the drive signals. After being isolated and amplified by the gate drive circuit, the drive signals are sent to the gates of each power switch in the converter power bridge to control their high-frequency conduction and turn-off, thereby synthesizing an AC voltage waveform equivalent to the final voltage reference command at the converter output.

[0052] In summary, the converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to the embodiments of the present invention is explained. Instead of using a fixed virtual impedance value, it acquires two key dynamic indicators in real time: the normalized active power characterizing the converter's own operating condition and the estimated short-circuit ratio characterizing the external power grid. These two dynamic indicators are used as inputs to a fuzzy logic controller. A fuzzy inference mechanism is used to comprehensively determine the current complex operating condition, thereby dynamically and online generating adjustment coefficients for virtual resistance and virtual inductance. Based on these adjustment coefficients, the final virtual impedance value is adjusted in real time, enabling it to automatically adapt to power fluctuations and changes in power grid strength, ensuring that the wind turbines maintain good damping characteristics and power output performance across a wide operating range.

[0053] The present invention also provides a converter virtual impedance control system adapted to the parallel operation of multiple wind turbines.

[0054] Figure 6 This is a block diagram of a converter virtual impedance control system adapted to the parallel operation of multiple wind turbines according to an embodiment of the present invention. Figure 6As shown, a converter virtual impedance control system 100 adapted to the parallel operation of multiple wind turbines according to an embodiment of the present invention includes: a voltage and current acquisition module 110, used to acquire the three-phase voltage at the grid connection point from the voltage transformer at the grid connection point, and to acquire the three-phase output current of the converter from the current transformer on the output side of the converter; a grid strength assessment module 120, used to perform grid strength assessment on the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain normalized active power and estimated short-circuit ratio; and a parameter fuzzification module 130, used to perform parameter fuzzification on the normalized active power and estimated short-circuit ratio through a fuzzy logic controller to obtain... The system comprises: a normalized active power membership vector and an estimated short-circuit ratio membership vector; a fuzzy inference module 140, used to perform fuzzy inference on the normalized active power membership vector and the estimated short-circuit ratio membership vector based on a fuzzy rule base to obtain virtual resistance adjustment coefficients and virtual inductance adjustment coefficients; a defuzzification module 150, used to defuzzify the virtual resistance adjustment coefficients and virtual inductance adjustment coefficients based on a reference virtual resistance and a reference virtual inductance to obtain virtual resistance values ​​and virtual inductance values; and a reference command correction module 160, used to perform voltage reference command correction based on the virtual resistance values ​​and virtual inductance values.

[0055] The specific implementation method of the converter virtual impedance control system adapted to the parallel operation of multiple wind turbines provided in this embodiment of the invention can be found in the converter virtual impedance control method adapted to the parallel operation of multiple wind turbines provided in this embodiment of the invention, and will not be repeated here.

[0056] The converter virtual impedance control system 100 adapted to the parallel operation of multiple wind turbines according to embodiments of the present invention can be implemented in various computing devices, such as the main controller of a wind turbine generator, a wind power converter controller, or a dedicated industrial computer. In one possible implementation, the converter virtual impedance control system 100 adapted to the parallel operation of multiple wind turbines according to embodiments of the present invention can be integrated into the computing device as a software module and / or a hardware module. For example, the converter virtual impedance control system 100 adapted to the parallel operation of multiple wind turbines can be a software module in the control firmware of the computing device, or it can be a dedicated control algorithm program developed for the computing device; of course, the converter virtual impedance control system 100 adapted to the parallel operation of multiple wind turbines can also be one of many hardware modules of the computing device, for example, implemented as dedicated digital signal processor logic or field-programmable gate array circuit.

[0057] Those skilled in the art will understand that the above embodiments are specific implementations of the present invention, and in practical applications, various changes can be made in form and detail without departing from the spirit and scope of the present invention.

Claims

1. A method for controlling the virtual impedance of a converter adapted to the parallel operation of multiple wind turbines, characterized in that, include: The three-phase voltage at the grid connection point is collected from the voltage transformer at the grid connection point, and the three-phase current at the output of the converter is collected from the current transformer on the output side of the converter. A grid strength assessment is performed on the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain the normalized active power and estimate the short-circuit ratio; The normalized active power and the estimated short-circuit ratio are fuzzified by a fuzzy logic controller to obtain the normalized active power membership vector and the estimated short-circuit ratio membership vector. Based on the fuzzy rule base, fuzzy reasoning is performed on the normalized active power membership vector and the estimated short-circuit ratio membership vector to obtain the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient. Based on the reference virtual resistance and reference virtual inductance, the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient are defuzzified to obtain the virtual resistance value and the virtual inductance value. Voltage reference command correction is performed based on virtual resistance and virtual inductance values.

2. The converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to claim 1, characterized in that, A grid strength assessment is performed on the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain normalized active power and estimate the short-circuit ratio, including: The three-phase voltage at the grid connection point and the three-phase current at the converter output are transformed by coordinates to obtain the d-axis voltage, q-axis voltage, d-axis current and q-axis current; Instantaneous active power is calculated based on d-axis voltage, q-axis voltage, d-axis current and q-axis current, and then normalized to obtain normalized active power. The grid impedance value is obtained by estimating the grid impedance at the grid connection point using the online disturbance observation method. The short-circuit ratio is estimated by dividing the grid reference impedance by the grid impedance value.

3. The converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to claim 1, characterized in that, The normalized active power and estimated short-circuit ratio are fuzzified using a fuzzy logic controller to obtain the normalized active power membership vector and the estimated short-circuit ratio membership vector, including: Based on the fuzzy language set and membership mapping function, normalized active power and estimated short-circuit ratio are mapped into fuzzy language to obtain the normalized active power membership vector and the estimated short-circuit ratio membership vector.

4. The converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to claim 1, characterized in that, Based on a fuzzy rule base, fuzzy inference is performed on the normalized active power membership vector and the estimated short-circuit ratio membership vector to obtain the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient, including: When the normalized active power is determined to be low and the estimated short-circuit ratio is determined to be weak, the virtual resistance adjustment coefficient is set to large and the virtual inductance adjustment coefficient is set to small. When the normalized active power is determined to be high and the estimated short-circuit ratio is determined to be strong, the virtual resistance adjustment coefficient is set to small and the virtual inductance adjustment coefficient is set to medium.

5. The converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to claim 1, characterized in that, Based on the reference virtual resistance and reference virtual inductance, the virtual resistance adjustment coefficient and virtual inductance adjustment coefficient are defuzzified to obtain the virtual resistance value and virtual inductance value, including: The virtual resistance adjustment coefficient and virtual inductance adjustment coefficient are defuzzified using the centroid method to obtain the virtual resistance adjustment coefficient data and virtual inductance adjustment coefficient data.

6. The converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to claim 5, characterized in that, Based on the reference virtual resistance and reference virtual inductance, the virtual resistance adjustment coefficient and virtual inductance adjustment coefficient are defuzzified to obtain the virtual resistance value and virtual inductance value, and the process also includes: Calculate the virtual resistance value using the following formula: ； in, This is the virtual resistance adjustment coefficient data. As a reference virtual resistor, This is a virtual resistance value; The virtual inductance value is calculated using the following formula: ； in, This is the virtual inductance adjustment coefficient data. As a reference virtual inductor, This is the virtual inductance value.

7. The converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to claim 1, characterized in that, Voltage reference command correction based on virtual resistance and virtual inductance values ​​includes: Based on the virtual resistance value, virtual inductance value, and real-time d-axis current and q-axis current, calculate the d-axis virtual voltage drop and q-axis virtual voltage drop; The d-axis virtual voltage drop and q-axis virtual voltage drop are filtered out from the original voltage reference command of the outer loop controller to obtain the final voltage reference command; The final voltage reference command is input to the converter current inner loop controller to drive the pulse width modulation module to generate a drive signal for adjusting the converter's virtual impedance.

8. The converter virtual impedance control method adapted to the parallel operation of multiple wind turbines according to claim 7, characterized in that, Based on the virtual resistance value, virtual inductance value, and real-time d-axis and q-axis currents, calculate the d-axis virtual voltage drop and q-axis virtual voltage drop, including: The virtual pressure drop along the d-axis and the virtual pressure drop along the q-axis are calculated using the following formulas: ； ； in, For the d-axis virtual pressure drop, This is a virtual resistance value. For real-time d-axis current, This is a virtual inductance value. This represents the real-time q-axis current. The angular frequency of the power grid. This represents the virtual voltage drop along the q-axis.

9. The converter virtual impedance control method for adapting to parallel operation of multiple wind turbines according to claim 7, characterized in that, The final voltage reference command is obtained by filtering out the d-axis virtual voltage drop and q-axis virtual voltage drop from the raw voltage reference command of the outer loop controller, including: Subtract the d-axis voltage command from the original voltage reference command to obtain the corrected d-axis voltage command; Subtract the q-axis voltage command from the q-axis virtual voltage drop in the original voltage reference command to obtain the corrected q-axis voltage command; The corrected d-axis voltage command and the corrected q-axis voltage command are combined into the final voltage reference command.

10. A converter virtual impedance control system adapted to the parallel operation of multiple wind turbines, characterized in that, include: The voltage and current acquisition module is used to acquire the three-phase voltage at the grid connection point from the voltage transformer at the grid connection point, and to acquire the three-phase current output from the current transformer on the output side of the converter. The grid strength assessment module is used to assess the grid strength of the three-phase voltage at the grid connection point and the three-phase output current of the converter to obtain normalized active power and estimate the short-circuit ratio. The parameter fuzzification module is used to fuzzify the normalized active power and the estimated short-circuit ratio through a fuzzy logic controller to obtain the normalized active power membership vector and the estimated short-circuit ratio membership vector. The fuzzy inference module is used to perform fuzzy inference on the normalized active power membership vector and the estimated short-circuit ratio membership vector based on the fuzzy rule base to obtain the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient. The defuzzification module is used to defuzzify the virtual resistance adjustment coefficient and the virtual inductance adjustment coefficient based on the reference virtual resistance and the reference virtual inductance to obtain the virtual resistance value and the virtual inductance value. The reference instruction correction module is used to correct the voltage reference instruction based on the virtual resistance value and the virtual inductance value.

Images