Wind power inertia analysis method considering fan inertia support

Abstract

The application discloses a wind power inertia analysis method considering fan inertia support and relates to the technical field of wind power system control, and comprises the following steps: collecting dynamic response data of a target generator set under a preset disturbance working condition, constructing a double-loop dynamic model based on the dynamic response data; coupling the double-loop dynamic model with a preset power grid model in a closed loop to construct a unit-power grid joint dynamic simulation model; obtaining a multi-dimensional inertia response characteristic matrix based on the joint dynamic simulation model; obtaining a contribution value of a virtual inertia control layer to power grid frequency support based on the inertia response characteristic matrix; and dynamically adjusting control parameters of a virtual inertia controller in the target generator set according to the quantitative contribution value. The application simulates the response characteristics of the wind power generator set under different disturbances through the double-loop dynamic model and the unit-power grid joint simulation, analyzes the inertia response characteristics based on real-time data, has the ability of efficiently and accurately adjusting the virtual inertia control layer, and enhances the stability of the power grid.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wind power system control, and more particularly, to a wind power inertia analysis method considering wind turbine inertia support. BACKGROUND

[0002] With the increasing installed capacity of wind power in power systems, wind turbine inertia support has become an important means to ensure the stability of power grid frequency. The so-called wind turbine inertia support refers to the release or absorption of rotor kinetic energy by wind turbine under disturbance conditions to assist the power grid to maintain frequency stability. At present, wind turbines generally use double-fed induction generators or full-power converter structures, and the electrical isolation leads to the weakening of inertia characteristics, so it is urgent to realize virtual inertia response through control strategy.

[0003] In the prior art, wind turbine inertia support is usually achieved by fixed gain virtual inertia control or frequency deviation based feedforward / feedback regulation of MPPT layer output active power. This kind of method mainly evaluates the inertia contribution by overall power response curve, lacks phased quantitative analysis of inertia support stage, primary frequency modulation stage and recovery stage, and does not systematically evaluate and optimize different disturbance amplitudes and durations, resulting in the inability to accurately identify the actual contribution of each stage to the stability of power grid frequency, and the control parameters are difficult to dynamically adjust to adapt to complex disturbance conditions. SUMMARY

[0004] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present application provide a wind power inertia analysis method considering wind turbine inertia support, which acquires wind turbine operation data and power grid frequency dynamic data, constructs a double-loop dynamic model and couples the power grid model to form a simulation system. Under disturbance conditions, the inertia and frequency modulation response of each stage are calculated, the inertia response feature matrix is constructed, and the contribution value of the virtual inertia control layer is inversely calculated based on the matrix, so as to realize dynamic optimization adjustment, thereby solving the problems that the inertia response cannot be quantified in stages and the control parameters are difficult to dynamically adjust in the prior art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme:

[0006] The wind power inertia analysis method considering wind turbine inertia support comprises the following steps: acquiring dynamic response data of a target generator under a preset disturbance condition, and constructing a double-loop dynamic model based on the dynamic response data; coupling the double-loop dynamic model with a preset power grid model to construct a unit-grid joint dynamic simulation model for simulating the interaction between the unit and the power grid; obtaining a multi-dimensional inertia response feature matrix based on the joint dynamic simulation model; obtaining the contribution value of the virtual inertia control layer to the power grid frequency support based on the inertia response feature matrix; and dynamically adjusting the control parameters of the virtual inertia controller in the target generator according to the quantitative contribution value.

[0007] In a preferred embodiment, the joint dynamic simulation model is used to obtain a multi-dimensional inertia response feature matrix, specifically: a disturbance event is set in the dynamic simulation model, dynamic simulation calculation is performed, inertia response values and frequency modulation response values at different time stages in the simulation data are extracted, and a multi-dimensional inertia response feature matrix is constructed based on the response values.

[0008] In a preferred embodiment, the double-loop dynamic model is constructed by: based on the rotor speed and the pitch angle, a maximum power point tracking control layer dynamic model is constructed by a nonlinear adaptive control algorithm as an inner loop; based on the grid frequency and its rate of change, a virtual inertia control layer dynamic model is constructed by using a feedforward-feedback composite control strategy to adjust the active power output of the converter as an outer loop; and the double-loop dynamic model is formed by coupling the inner loop and the outer loop.

[0009] In a preferred embodiment, the joint dynamic simulation model is used to obtain a multi-dimensional inertia response feature matrix by a linear adaptive control algorithm.

[0010] In a preferred embodiment, the feedforward-feedback composite control strategy is specifically: the grid frequency signal and its rate of change data are obtained; based on the rate of change of the grid frequency, a feedforward power increment is calculated through a feedforward control channel; based on the deviation of the grid frequency signal from the rated frequency, a feedback power increment is calculated through a feedback control channel; the feedforward power increment and the feedback power increment are added to obtain a comprehensive power adjustment amount; and the comprehensive power adjustment amount is input to the active power control of the converter to obtain an updated active power.

[0011] The updated active power is used as the input for the next feedforward and feedback control calculation.

[0012] In a preferred embodiment, the double-loop dynamic model is coupled with a preset grid model in a closed loop to construct a unit-grid joint dynamic simulation model for simulating the interaction between the unit and the grid, specifically: a grid model with frequency deviation as a state variable is constructed; the active power increment output by the double-loop dynamic model is decomposed into a virtual inertia control layer output and an MPPT layer output; a first coupling channel is established to input the active power increment output by the double-loop dynamic model into the power balance equation of the grid model for output coupling; a second coupling channel is established to feed back the frequency deviation output by the grid model to the virtual inertia control layer for input coupling; and the unit-grid joint dynamic simulation model is constructed through the input coupling and the output coupling.

[0013] In a preferred embodiment, the multi-dimensional inertia response feature matrix is obtained based on the joint dynamic simulation model, specifically: a disturbance sequence with different amplitudes and durations is constructed to form a disturbance signal; based on the disturbance signal, the instantaneous change rate and the recovery slope of the power grid frequency curve data of the simulation model output are calculated by using the variable step integral method; the instantaneous change rate and the recovery slope data are divided into inertia support stage data, primary frequency modulation stage data and recovery steady state stage data according to a preset time window, and the inertia response value and the frequency modulation response value of each stage are calculated; the inertia response value and the frequency modulation response value of each stage are normalized according to the disturbance amplitude and the duration to construct the inertia response feature matrix.

[0014] In a preferred embodiment, the control parameters of the virtual inertia controller in the target generator are dynamically adjusted, specifically: the inertia response feature matrix is calculated by stage inversion to obtain initial contribution values of each stage; the initial contribution values of each stage are optimized to generate optimized contribution values; the optimized contribution values are converted into parameter increments of the virtual inertia controller and updated to the virtual inertia control layer for dynamic adjustment.

[0015] In a preferred embodiment, the inertia response feature matrix is calculated by stage inversion to obtain initial contribution values of each stage, specifically: the stages include an inertia support stage, a primary frequency modulation stage and a recovery stage; the rotor kinetic energy change rate, the second derivative of the power grid frequency and the pitch angle change acceleration are collected to construct a three-dimensional phase space; the inertia support ring, the frequency modulation helix and the recovery steady state point are identified as geometric features through the three-dimensional phase space; and the initial contribution values of each stage are calculated according to the geometric features.

[0016] In a preferred embodiment, the initial contribution values of each stage are optimized to generate optimized contribution values, specifically: the stage performance evaluation vectors are calculated according to the initial contribution values of each stage and the corresponding power grid frequency response indicators; the optimization objective function is constructed based on the stage performance evaluation vectors to output comprehensive optimization indicators; the upper limit of the rotor kinetic energy reserve of the wind turbine, the power regulation range of the converter and the allowable fluctuation threshold of the power grid frequency are collected to construct a constraint condition set to output the allowed value range; and the optimized stage contribution values are obtained by using a constraint optimization algorithm based on the initial stage contribution values, the comprehensive optimization indicators and the allowed value range.

[0017] The wind turbine inertia support wind power inertia analysis method has the following technical effects and advantages:

[0018] 1.The present application realizes accurate evaluation of inertia output of each stage under different disturbance amplitude and duration by constructing inertia response feature matrix and using three-dimensional phase space staging inversion algorithm to quantify frequency response contribution of wind turbine in inertia support stage, primary frequency modulation stage and recovery steady stage, solving the problem of overall power curve evaluation in prior art and inability to analyze by stage.

[0019] 2.The present application realizes online dynamic adjustment of virtual inertia output of wind turbine under different disturbance conditions by combining the inertia contribution value of each stage obtained by inversion with the upper limit of rotor kinetic energy reserve of wind turbine, power regulation range of converter and allowable fluctuation threshold of grid frequency for constraint optimization and dynamic updating to virtual inertia control layer, solving the problem of fixed control parameters and inadaptability to various disturbances in prior art. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 The present application considers the flowchart of wind power inertia analysis method of wind turbine inertia support.

[0021] Figure 2 The present application considers the structure diagram of unit-grid joint dynamic simulation model. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0023] Embodiment 1, Figure 1 The present application considers the wind power inertia analysis method of wind turbine inertia support, including:

[0024] S1, collecting dynamic response data of the target generator set under the preset disturbance condition, and constructing a double-loop dynamic model based on the dynamic response data;

[0025] In this embodiment, dynamic response data of the target generator set under the preset disturbance condition is collected, and a double-loop dynamic model is constructed based on the dynamic response data, specifically:

[0026] The double-loop dynamic model includes a maximum power point tracking control layer and a virtual inertia control layer. Dynamic response data of the target generator set under the preset disturbance condition is collected through a rotor speed sensor, a pitch angle encoder and a power transmitter of the wind turbine, and the collected data includes rotor speed , pitch angle The actual output power of the wind turbine is measured using operating data. Based on rotor speed and pitch angle, a dynamic model of the maximum power point tracking control layer is constructed using a nonlinear adaptive control algorithm. This model serves as the inner loop, and data including real-time grid frequency is collected by the grid synchronization phasor measurement unit. and its rate of change Based on the grid frequency and its rate of change, a feedforward-feedback composite control strategy is adopted to adjust the active power output of the converter, and a virtual inertia control layer dynamic model is constructed as the outer loop. The inner and outer loops are coupled to form a double-loop dynamic model.

[0027] By rotor speed and propeller pitch angle The nonlinear mapping model is constructed using the following formula:

[0028]

[0029] in, air density, Wind turbine blade radius, The wind energy utilization coefficient is the tip speed ratio. and nonlinear functions, Wind speed is measured using an anemometer.

[0030] The dynamic model function of the inner loop maximum power point tracking control layer of the dual-loop dynamic model. Through optimization Obtain, specifically .

[0031] The nonlinear adaptive control algorithm is specifically as follows:

[0032] The power error is obtained by comparing the output power of the maximum power point function with the actual output power of the unit. The power error is input into the control gain update formula to obtain the updated control gain. The specific formula is as follows:

[0033]

[0034] in, for Control gain at any time, For adaptive learning rate ( The convergence speed is determined by the convergence rate. For nonlinear correction terms ( To suppress excessive gain growth and avoid excessive pitch angle adjustment, the pitch angle adjustment is calculated using a nonlinear control law based on the updated control gain.

[0035]

[0036] wherein, is a symbol function, which ensures the direction correctness. A new pitch angle is obtained, the new pitch angle acts on the aerodynamic model of the fan, and the total active power increment of the unit is recalculated , forming a closed loop.

[0037] The feedforward-feedback compound control strategy is specifically:

[0038] The rate of change of the grid frequency is input into the feedforward control module to simulate the inertia characteristics of the synchronous machine and calculate the feedforward power increment:

[0039]

[0040] wherein, is the feedforward gain, which is related to the virtual inertia constant .

[0041] The deviation between the grid frequency signal and the rated frequency is calculated, and the grid frequency deviation is input into the feedback control module to calculate the feedback power increment:

[0042]

[0043] wherein, is the proportional gain, which determines the frequency regulation accuracy, is the rated frequency.

[0044] The feedforward power increment and the feedback power increment are added to obtain the comprehensive power adjustment amount , the comprehensive power adjustment amount is input into the active power control of the converter to obtain the updated active power output , the converter output power affects the grid frequency, and the PMU collects new frequency data and feeds back to the control layer.

[0045] S2, the double-loop dynamic model is coupled with the preset grid model to construct a unit-grid joint dynamic simulation model for simulating the interaction between the unit and the grid;

[0046] In this embodiment, the unit-grid joint dynamic simulation model is constructed, specifically:

[0047] A grid model with frequency deviation as a state variable is constructed, specifically:

[0048]

[0049] wherein, is the equivalent inertia constant of the grid, which reflects the system inertia, is the total active power increment of the unit, is the load disturbance power, is the grid damping coefficient.

[0050] The active power increment output by the double-loop dynamic model is decomposed into a virtual inertia control layer output and an MPPT layer output , a first coupling channel is established, the active power increment output by the double-loop dynamic model is input into a power balance equation of the grid model, and output coupling is performed, specifically as follows:

[0051]

[0052] The unit power change directly affects the grid frequency dynamics.

[0053] A second coupling channel is established, and the frequency deviation output by the grid model is fed back to the virtual inertia control layer, and input coupling is performed, specifically as follows:

[0054]

[0055] The frequency deviation acts on the unit power output through the outer loop control strategy.

[0056] Through the closed-loop connection of the output coupling and the input coupling, the double-loop dynamic model and the grid model are integrated, the unit output and the grid response interact, a unit-grid joint dynamic simulation model is formed, joint dynamic simulation is realized, and a unit power response curve and a grid frequency change curve are output.

[0057] S3, based on the joint dynamic simulation model, a multi-dimensional inertia response feature matrix is obtained;

[0058] In this embodiment, based on the joint dynamic simulation model, a multi-dimensional inertia response feature matrix is obtained, specifically as follows:

[0059] In the dynamic simulation model, a disturbance event is set, in the disturbance working condition, the inertia response value and the frequency modulation response value in each stage are calculated through the simulation model, the inertia response feature matrix is constructed, and specifically as follows:

[0060] In the dynamic simulation model, a disturbance event is set, a power mutation or a cut-in or cut-out working condition that may occur in the grid is simulated, and a power disturbance sequence with different amplitudes and durations is constructed According to the amplitude and duration combination, a plurality of disturbance sequences are generated, and a disturbance signal is formed:

[0061]

[0062] According to the grid load fluctuation range, the minimum amplitude and the maximum amplitude are set, and the amplitude step size is determined, wherein For the expected generated amplitude quantity, the amplitude sequence is obtained Similarly, according to the frequency modulation characteristics of the power grid, the shortest duration and the longest duration are set, and the time step is determined, so as to obtain the duration sequence . The amplitude sequence and the duration sequence are used to form a Cartesian product, and each pair corresponds to a disturbance signal.

[0063] The multi-dimensional inertia response feature matrix is obtained by the linear adaptive control algorithm, specifically:

[0064] The generated disturbance signal is input into the simulation model, and the dynamic curve of the grid frequency with time is solved by the variable step integration method, and the instantaneous change rate and the recovery slope are recorded. The step can be automatically adjusted according to the system frequency change rate. When the frequency changes rapidly, a small step is used to ensure the calculation accuracy, and when the frequency changes slowly, a large step is used to improve the calculation efficiency and ensure the accuracy in the rapid change stage. The instantaneous change rate is:

[0065]

[0066] Wherein, is the current integration step.

[0067] The recovery slope is:

[0068]

[0069] Wherein, , are the start and end times of each stage respectively.

[0070] The instantaneous change rate and the recovery slope data are divided into inertia support stage, primary frequency modulation stage and recovery steady stage according to the preset time window, and the data of the corresponding stage, the inertia support stage represents the time period from the occurrence of the disturbance to the completion of the inertia response, the primary frequency modulation stage represents the time period from the minimum frequency to the frequency close to the steady state, and the recovery steady stage represents that the frequency is basically stable and the system tends to be a new steady state, wherein, is the disturbance occurrence time, is the inertia response end time, is the primary frequency modulation end time, To restore the steady state end time. The inertia response value and the frequency modulation response value of each stage are used to quantify the instantaneous suppression ability of the fan to the grid frequency disturbance in each stage and the frequency restoration ability of the system in each stage, and the inertia response value of each stage is the maximum absolute value of the frequency instantaneous change rate, that is , wherein is The frequency instantaneous change rate of the stage, is the corresponding stage time period, , 1 corresponds to the inertia support stage, 2 corresponds to the primary frequency modulation stage, and 3 corresponds to the steady state restoration stage. The inertia response value reflects the support ability of instantaneous power to frequency.

[0071] The frequency modulation response value is solved by the recovery slope of the frequency curve of each stage, that is:

[0072]

[0073] , wherein is The grid frequency curve of the stage.

[0074] The inertia response value and the frequency modulation response value of each stage under multiple disturbances are normalized according to the disturbance amplitude and duration to construct an inertia response feature matrix, which is:

[0075] To eliminate the dimensional differences of different amplitude columns and different duration rows, bidirectional normalization (once for rows and once for columns) is adopted to pull the response values of different durations under the same amplitude to , so as to compare the relative strength of the duration dimension, and pull the response values of different amplitudes under the same duration to , so as to compare the relative strength of the amplitude dimension. The two kinds of normalization results are fused by a weight coefficient to obtain two groups of data, which are the normalized inertia response value and the frequency modulation response value , for each stage , the inertia matrix block , the element is , and the frequency modulation matrix block , the element is , the inertia matrix block and the frequency modulation matrix block are horizontally spliced to obtain the inertia response feature matrix:

[0076]

[0077] , wherein and are the relative weights of the inertia and frequency modulation two types of indexes. S4, based on the inertia response feature matrix, the contribution value of the virtual inertia control layer to the support of the grid frequency is obtained;

[0078] In the embodiment, the inertia response characteristic matrix is subjected to stage inversion calculation to obtain initial contribution values of each stage, specifically:

[0079] Each stage includes an inertia support stage, a primary frequency modulation stage, and a recovery stage. The stage inversion calculation is realized by a three-dimensional phase space geometric feature recognition method. For each disturbance condition and stage, the rotor kinetic energy change rate , the second derivative of the grid frequency , and the pitch angle change acceleration are extracted. The rotor kinetic energy change rate (i.e., the change in rotor speed) directly affects the response capability of the wind turbine to the grid frequency disturbance. The change rate of the rotor kinetic energy provides real-time dynamic feedback for the virtual inertia and determines whether the wind turbine can provide sufficient inertia support when the grid frequency fluctuates. The change in grid frequency (i.e., frequency deviation) is the direct target of the frequency support of the wind turbine. The second derivative of the frequency (i.e., the acceleration of the frequency change) reflects the rate of change of the grid frequency and the dynamic characteristics of the frequency disturbance. The acceleration of the pitch angle change effectively reflects the dynamic capability of the wind turbine to adjust power. When the frequency disturbance occurs, the wind turbine needs to quickly adjust the power output. The acceleration of the pitch angle change reflects the sensitivity of the wind turbine to respond to the grid frequency disturbance.

[0080] The rotor kinetic energy change rate , the second derivative of the grid frequency , and the pitch angle change acceleration are combined to form a three-dimensional phase space . Each time step corresponds to a three-dimensional point. The trajectory is formed by time series, and the processed trajectory is .

[0081] For the inertia support stage, the inertia support ring features, maximum radial deviation and ring trajectory area , are identified. The specific steps are as follows: selecting the time period of the inertia support stage , and extracting the three-dimensional phase space trajectory points in the time period .

[0082] The trajectory centroid of the trajectory points is calculated as follows:

[0083]

[0084] wherein is the number of time steps.

[0085] The radial distance of each trajectory point to the trajectory centroid is calculated as follows:

[0086]

[0087] Find the maximum radial distance, that is, the point farthest from the centroid among all trajectory points:

[0088]

[0089] Calculate the projected area of the inertia support stage trajectory by the convex hull algorithm to obtain the ring area:

[0090]

[0091] The contribution of the inertia support stage mainly reflects the ability of the system to stabilize the frequency change through inertia support. The maximum radial deviation of the inertia support ring and the ring trajectory area measure the strength of inertia support. The initial contribution value of the inertia support stage The formula is:

[0092]

[0093] Reflects the inhibitory effect of inertia support on frequency change. The larger the value, the greater the impact of inertia support on frequency.

[0094] For the primary frequency modulation stage, identify the characteristics of the frequency modulation spiral: the frequency modulation spiral radius and the pitch, specifically:

[0095] Select the time period of the primary frequency modulation stage Extract the three-dimensional phase space trajectory points in the time period Spiral fitting of the trajectory of the trajectory points forms a frequency modulation spiral. Calculate the average radius of the frequency modulation spiral in the plane:

[0096]

[0097] Where represents the projection of the trajectory point on the plane.

[0098] Calculate the pitch of the frequency modulation spiral:

[0099]

[0100] Where , is the number of trajectory turns, which can be detected by Fourier analysis or trajectory projection.

[0101] The frequency modulation spiral radius and pitch are important characteristics of the frequency modulation stage, reflecting the system's ability to recover frequency. The initial contribution value of the primary frequency modulation stage The formula is:

[0102]

[0103] wherein, and are the maximum radius and pitch of the frequency-modulated helix among all perturbations.

[0104] For the recovery steady stage, a recovery steady point is identified, specifically:

[0105] The time period of the recovery steady stage is selected , a recovery steady point is identified, and the trajectory mean of the recovery steady stage, i.e., the average position of the steady stage, is represented, with the formula being:

[0106]

[0107] The standard deviation of the data points of the recovery steady stage to the steady point is calculated, representing the fluctuation of the steady stage. The formula is:

[0108]

[0109] The contribution value of the recovery steady stage mainly depends on the speed and stability of the system recovering to the steady state, which is closely related to the steady point and the standard deviation The initial contribution value of the recovery steady stage The formula for calculating is:

[0110]

[0111] Reflects the stability of the recovery steady stage. The smaller the standard deviation, the more stable the system recovers, and the greater the contribution value.

[0112] The initial contribution value of each stage is constrained and optimized to generate an optimized contribution value, specifically:

[0113] According to the initial contribution value of each stage and the corresponding power grid frequency response index, a stage performance evaluation vector is calculated:

[0114]

[0115] wherein, represents element-wise multiplication, respectively represent the power grid frequency response indexes (such as frequency deviation, recovery time, etc.) corresponding to the inertia support, primary frequency modulation, and recovery steady stage, which can be obtained through a simulation model.

[0116] According to the stage performance evaluation vector, an optimization objective function is constructed:

[0117]

[0118] wherein, Performance evaluation components representing inertia support, frequency modulation and recovery to steady state phase respectively, Weight coefficients of each phase, representing the relative importance of each phase to the optimization objective function. The weight coefficients should satisfy: .

[0119] By optimizing the objective function, the contribution values of different phases can be balanced, so that the wind turbine can maximize the contribution to the grid frequency regulation and minimize the grid frequency fluctuation.

[0120] Collect the upper limit of wind turbine rotor kinetic energy reserve, converter power regulation range and grid frequency allowable fluctuation threshold to construct the constraint condition set:

[0121]

[0122] Rotor kinetic energy is the key to the inertia support that wind turbines can provide. Its upper limit is determined by the design parameters of wind turbines (such as rotor speed, rotor inertia moment, etc.). The manufacturer of wind turbines usually provides the rotor inertia moment of each unit and the maximum speed , using these data, the upper limit of rotor kinetic energy can be calculated, specifically:

[0123]

[0124] The converter regulation range of wind turbines is mainly determined by the design of the unit and the requirements of the grid. The power regulation range of the converter is usually an important control parameter of the wind turbine, which directly affects the upper and lower limits of active power output. This data can be found in the controller settings of the unit. It can be read through the remote monitoring system or PLC (Programmable Logic Controller) system of the unit. It should satisfy:

[0125]

[0126] where, is the current active power output, is the minimum power that the converter can output (may be negative, indicating that it can absorb grid power), is the maximum power that the converter can output.

[0127] Similarly, the grid frequency allowable fluctuation threshold can be found in the relevant standards in the grid operation manual or power dispatching regulations, which should satisfy:

[0128]

[0129] where, is the difference between the grid frequency and the rated frequency, is the maximum value of the allowable fluctuation of the grid frequency.

[0130] solving the optimal staging contribution value, i.e. the optimized contribution value, using a constraint optimization algorithm , to maximize the frequency regulation performance of the wind turbine and ensure the stability of the power grid.

[0131] S5, according to the quantified contribution value, dynamically adjusting the control parameters of the virtual inertia controller in the target generator set;

[0132] In this embodiment, according to the quantified contribution value, dynamically adjusting the control parameters of the virtual inertia controller in the target generator set, specifically:

[0133] performing staging inversion calculation on the inertia response characteristic matrix to obtain the initial contribution value of each stage;

[0134] performing constraint optimization on the initial contribution value of each stage to generate the optimized contribution value;

[0135] convert the optimized contribution value into a virtual inertia controller parameter increment, and update it to the virtual inertia control layer for dynamic adjustment.

[0136] Among them, the optimized contribution value is converted into a virtual inertia controller parameter increment, and is updated to the virtual inertia control layer for dynamic adjustment, specifically:

[0137] The virtual inertia control layer is mainly used to adjust the active power output of the wind turbine to support the frequency stability of the power grid. The controller parameters of the virtual inertia control layer include the virtual inertia coefficient , the virtual inertia control gain and the power regulation gain These parameters control the dynamic response speed and amplitude of the virtual inertia, which can affect the recovery process of the grid frequency, and its formula expression is

[0138]

[0139] Among them, is the virtual inertia output power, is the power grid power error, is the grid frequency deviation.

[0140] By analyzing the contribution value of different stages , it can be deduced how to convert these contribution values into virtual inertia control parameters. In order to maintain the stability of dynamic adjustment, proportional mapping is needed, as follows:

[0141]

[0142]

[0143] Among them, and is a mapping function to convert the optimized contribution value into a control parameter increment.

[0144] For example, the contribution value can be mapped to the control parameter increment by a linear relationship:

[0145]

[0146]

[0147] wherein, and are the maximum possible values of the virtual inertia control layer.

[0148] The converted controller parameter increment and are added to the current control parameter value, and if the parameter value exceeds its maximum range, saturation limit should be performed to ensure that the parameter does not exceed a reasonable range. These updated virtual inertia control parameters are fed back to the virtual inertia control layer model for real-time control to dynamically adjust the active power output of the wind turbine generator and help the power grid maintain stability.

[0149] The above formulas are all dimensionless numerical calculations, and the formulas are obtained by software simulation of a large amount of data to obtain a formula of the latest real situation. The preset parameters in the formula are set by a person skilled in the art according to the actual situation.

[0150] The above embodiments can be realized wholly or partially by software, hardware, firmware or any combination thereof. When realized by software, the above embodiments can be realized wholly or partially in the form of a computer program product.

[0151] Those skilled in the art can realize that the modules and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solutions. A person skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0152] In addition, the functional modules in each embodiment of the present application can be integrated in one processing module, or each module can exist physically independently, or two or more modules can be integrated in one module.

[0153] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0154] Finally, the above merely provides the preferred embodiments of the present application, and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A wind power inertia analysis method considering the inertia support of a wind turbine, characterized in that, The method comprises the following steps: Collecting dynamic response data of a target generator set under a preset disturbance working condition, and constructing a double-loop dynamic model based on the dynamic response data, specifically: constructing a maximum power point tracking control layer dynamic model as an inner loop based on rotor speed and pitch angle through a nonlinear adaptive control algorithm; constructing a virtual inertia control layer dynamic model as an outer loop based on grid frequency and its change rate by using a feedforward-feedback compound control strategy to adjust the active power output of a converter; and coupling the inner loop and the outer loop to form the double-loop dynamic model; Coupling the double-loop dynamic model with a preset grid model to construct a generator-grid joint dynamic simulation model for simulating the interaction between the generator set and the grid; Based on the joint dynamic simulation model, a multi-dimensional inertia response feature matrix is obtained; Based on the inertia response feature matrix, a contribution value of the virtual inertia control layer to the support of the grid frequency is obtained; According to the quantitative contribution value, the control parameters of the virtual inertia controller in the target generator set are dynamically adjusted. 2.The method of claim 1, wherein, The method comprises the following steps: In the dynamic simulation model, a disturbance event is set, dynamic simulation calculation is performed, the inertia response value and the frequency modulation response value at different time stages in the simulation data are extracted, and a multi-dimensional inertia response feature matrix is constructed based on the response values. 3.The wind power inertia analysis method of considering the inertial support of a wind turbine according to claim 2, characterized in that, The method comprises the following steps: 4.The wind power inertia analysis method of considering the inertial support of a wind turbine according to claim 3, wherein, The method comprises the following steps: The feedforward-feedback compound control strategy comprises the following steps: The grid frequency signal and the change rate data thereof are obtained; Based on the change rate of the grid frequency, a feedforward power increment is calculated through a feedforward control channel; Based on the deviation of the grid frequency signal from the rated frequency, a feedback power increment is calculated through a feedback control channel; The feedforward power increment and the feedback power increment are added to obtain a comprehensive power adjustment amount; The comprehensive power adjustment amount is input into the active power control of the converter to obtain updated active power; 5. The wind power inertia analysis method considering the inertia support of a wind turbine according to claim 4, characterized in that, The updated active power is input into the next feedforward and feedback control calculation. The method comprises the following steps: The method comprises the following steps: A grid model taking the frequency deviation as a state variable is constructed; The active power increment output by the double-loop dynamic model is decomposed into the output of the virtual inertia control layer and the output of the MPPT layer; A first coupling channel is established to input the active power increment output by the double-loop dynamic model into the power balance equation of the grid model for output coupling; 6. The wind power inertia analysis method considering the inertia support of a wind turbine according to claim 5, characterized in that, A second coupling channel is established to feed back the frequency deviation output by the grid model to the virtual inertia control layer for input coupling; Through the input coupling and the output coupling, the generator-grid joint dynamic simulation model is constructed. The method comprises the following steps: A disturbance signal is formed by constructing disturbance sequences with different amplitudes and durations; Based on the disturbance signal, the instantaneous change rate and the recovery slope of the grid frequency curve data output by the simulation model are calculated by using a variable step integral method; The transient change rate and the recovery slope data are divided into inertia support stage data, primary frequency modulation stage data and recovery steady stage data according to a preset time window, and inertia response values and frequency modulation response values of each stage are calculated; The inertia response values and the frequency modulation response values of each stage are normalized according to disturbance amplitude and duration to construct an inertia response feature matrix.

7. The wind power inertia analysis method considering the inertia support of a wind turbine according to claim 6, characterized in that, The control parameters of the virtual inertia controller in the dynamic adjustment target generator set are adjusted, specifically: The inertia response feature matrix is calculated by stage inversion to obtain initial contribution values of each stage; The initial contribution values of each stage are optimized to generate optimized contribution values; The optimized contribution values are converted into virtual inertia controller parameter increments and updated to the virtual inertia control layer for dynamic adjustment.

8. The wind power inertia analysis method considering the inertia support of a wind turbine according to claim 7, characterized in that, The inertia response feature matrix is calculated by stage inversion to obtain initial contribution values of each stage, specifically: The stages include inertia support stage, primary frequency modulation stage and recovery stage; Rotor kinetic energy change rate, power grid frequency second derivative and pitch angle change acceleration are collected to construct a three-dimensional phase space; The inertia support ring, frequency modulation helix and recovery steady point are identified as geometric features through the three-dimensional phase space; The initial contribution values of each stage are calculated according to the geometric features. 9.The wind power inertia analysis method of considering the inertial support of a wind turbine according to claim 8, wherein, The initial contribution values of each stage are optimized to generate optimized contribution values, specifically: According to the initial contribution values of each stage and the corresponding power grid frequency response index, a stage performance evaluation vector is calculated; Based on the stage performance evaluation vector, an optimization objective function is constructed, and a comprehensive optimization index is output; The upper limit of rotor kinetic energy reserve of the wind turbine, the power regulation range of the converter and the power grid frequency fluctuation threshold are collected to construct a constraint condition set, and the allowed value range is output; Based on the initial stage contribution value, the comprehensive optimization index and the allowed value range, a constraint optimization algorithm is used to obtain the optimized stage contribution value.

Images