Feasible region state space mapping evaluation method for droop and inertia support coefficients of wind farm

Abstract

The application discloses a wind farm droop and inertia support coefficient feasible region state space mapping evaluation method, wherein historical database operation data participating in primary frequency modulation is utilized on the wind farm side to construct a state space mapping linear function relationship formula among the wind farm droop coefficient, the wind farm inertia support coefficient, the wind speed, the power system droop coefficient and the power system inertia coefficient in the process of primary frequency modulation and inertia support, and the final speed of the wind turbine and the maximum power of the wind turbine; real-time wind speed of the wind farm and received power system droop and inertia parameters are measured, the linear function relationship formula among the above variables is utilized, the wind turbine final speed limit and the wind turbine power limit are considered, the wind farm droop and inertia support coefficient feasible region boundary is evaluated online, and the power system operation side is reported; the power system operation side updates the system coefficient and issues each wind farm, and the optimal droop-inertia frequency modulation coefficient feasible region of each wind farm and the power system where the wind farm is located is solved.

Description

Technical Field

[0001] This invention relates to the field of power system operation and control technology, and in particular to a method for evaluating the feasible region state-space mapping of wind farm droop and inertia support coefficient considering the dynamic characteristics of power systems. Background Technology

[0002] Wind farms should assess the feasible region of maximum droop-inertia coefficient under different operating conditions and report it to the system operator before participating in primary frequency regulation. Existing technologies include the method proposed in M. Fakhari Moghaddam Arani and YA-RI Mohamed, "Dynamic droop control for wind turbines participating in primary frequency regulation in microgrids," IEEE Trans. Smart Grid, vol. 9, no. 6, pp. 5742–5751, Nov. 2018, which proposes a method for assessing the feasible region of maximum droop, where the droop coefficient can be dynamically adjusted based on reserve power. In the literature YK Wu, WH Yang, YL Hu, and PQ Dzung, “Frequency regulation at a windfarm using time-varying inertia and droop controls,” IEEE Trans. Ind. Appl., vol. 55, no. 1, pp. 213-224, Jan.-Feb. 2019-, and H. Ye, W. Pei, and Z. Qi, “Analytical modelling of inertial and droop responses from a wind farm for short-term frequency regulation in power systems,” IEEE Trans. Power Syst., vol. 31, no. 5, pp. 3414-3423, Sept. 2016, the droop and inertia coefficients are expressed as analytical functions of the wind turbine rotor speed and wind speed, respectively. The authors evaluated the boundaries of each feasible region under different states.The literature MA González-Cagigal, JA Rosendo-Macías, and A. Gómez-Expósito, “Parameter Estimation of Wind Turbines With PMSM Using Cubature Kalman Filters,” IEEE Trans. Power Syst., vol. 35, no. 3, pp. 1796-1804, May 2020, uses a cubic Kalman filter to evaluate the boundary of the inertial feasible region based on the wind turbine rotor speed. However, the above method only constructs the feasible region for each coefficient independently. In reality, droop and inertial control in wind farms are both achieved by adjusting active power output, and therefore they interact. If the interaction mechanism between droop and inertial coefficients is ignored, it will be difficult to construct an accurate feasible region and achieve optimal PFR support for the wind farm.

[0003] In power systems with high renewable energy penetration, frequency dynamics change when numerous wind farms provide support based on assessment results. This leads to different frequency variations under the same load fluctuations, consequently affecting the capacity of each wind farm. In this context, considering the coupling relationship of primary frequency response capabilities among wind farms is crucial to obtaining the maximum coefficient boundary. Theoretically, the droop-inertia feasible region of a wind farm can be analytically solved at the system operator's end using physical models and centralized optimization methods. For example, the droop-inertia coefficient feasible region of wind power in the system was calculated using the parameter space method based on frequency stability, wind speed, and turbine status. The droop-inertia feasible region of each wind farm was constructed using the minimum frequency and the rate of frequency change, and the optimal coefficient scheme was solved based on turbine rotor stability. However, the centralized assessment in the above method requires the system operator to obtain state measurement data of all wind turbines in the grid, which leads to a heavy communication burden, and real-time turbine status data is difficult to upload from the wind farms.

[0004] In contrast, evaluations conducted at the wind farm level exhibit significant advantages such as rapid response, high reliability, and low communication burden. Priyatosh Mahish et al. used aggregated wind speed and power reserves, along with the frequency of each wind farm at the common coupling point, as allocation metrics to iteratively evaluate the droop and inertia feasible regions of each wind farm based on system dynamics. However, in reality, the power system droop-inertia coefficient should be aggregated from multiple wind farms in real time. Setting the power system droop-inertia coefficient as a constant fails to reflect changes in the system's frequency regulation capability. Therefore, it is crucial to propose an iterative evaluation scheme that coordinates between wind farms and system operators, taking into account variations in the primary frequency regulation response capability of wind farms within the system.

[0005] Although the aforementioned coordinated iterative methods do not require centralized modeling and evaluation, they still necessitate obtaining static state parameters from the wind farm. Thus, if the wind farm model is inaccurate or incomplete, the evaluated feasible region will certainly be unreliable. Furthermore, the lengthy solution times and optimization based on large-scale physical models make these methods unsuitable for online applications and tracking wind speed changes. To address these issues, data-driven methods have attracted significant attention because they can utilize historical data to construct equivalent primary frequency response models of wind farms. For example: 1) An ANN model to fit the nonlinear dynamics of the primary frequency response of a wind farm and evaluate the active power of the wind farm under different operating conditions. 2) A deep reinforcement learning method to construct a primary frequency response model of the wind farm and evaluate the droop-inertia feasible region. 3) An improved fuzzy method to evaluate the feasible region of the wind farm droop coefficient based on wind speed changes. However, these methods require large amounts of sampling data and complex training processes, and lack physical interpretability. Due to the lack of historical samples of extreme scenarios, the accuracy of the maximum boundary of the wind farm's feasible region cannot be guaranteed.

[0006] Due to the advantages of global linearization, Koopman operator theory (KOT) has been applied to fit nonlinear dynamic models. Based on KOT, nonlinear dynamic algebraic models in low-dimensional state spaces can be transformed into global linear forms in infinite-dimensional state spaces without loss of accuracy, and linear Koopman operator matrices can be estimated through data-driven training

[29] ,

[30] . In practical applications, increasing the dimension to an appropriate number is sufficient to transform nonlinear models into high-precision linear mathematical models. The global linearity property enables the establishment of accurate equivalent first-order frequency modulation evaluation models over a wide range without the need for historical samples to cover extreme scenarios. Summary of the Invention

[0007] This invention aims to propose a state-space mapping evaluation method for the feasible region of wind farm droop and inertia support coefficients. It utilizes a linear function model that includes the relationship between the final turbine speed, maximum turbine power, wind farm droop coefficient, wind farm inertia support coefficient, wind speed, power system droop coefficient, and power system inertia coefficient during primary frequency regulation and inertia support processes. By measuring the real-time wind speed of the wind farm and receiving the power system droop and inertia parameters, and using the constructed linear function relationship between the above variables, the optimal feasible region of the droop-inertia frequency regulation coefficient for each wind farm is obtained.

[0008] To achieve the above-mentioned objectives, the present invention proposes the following technical solution: This invention proposes a method for evaluating the feasible region state-space mapping of wind farm sag and inertia support coefficient, including: Step 1: Utilizing the operational data from the historical databases of each wind farm in the power system that participated in primary frequency regulation and inertia support processes, offline data samples for each wind farm are obtained through state-space mapping and least squares training, and a state-space mapping linear function relationship is constructed. This step specifically includes the following processing: 1-1. Obtain the data for each wind farm in the power system. G A data sample, the parameters of which include the power during the primary frequency regulation start-up phase of the wind turbine. Final fan speed Wind farm droop coefficient Inertia support coefficient Wind speed Power system droop coefficient and power system inertia coefficient ; 1-2. Based on the data sample, construct the following formula: The input data sample is defined as follows:

[0009] The output data sample is defined as follows:

[0010] in, Representing the i The first wind farm g The power of a single-stage frequency regulation start-up of a wind turbine is measured in a data sample. Representing the i The first wind farm g The final rotational speed of the wind turbine in each data sample Representing the i The first wind farm g The droop coefficient of each data sample wind farm Representing the i The first wind farm g Inertia support coefficient of each data sample Representing the i The first wind farm g Wind speed of each data sample Representing the i The first wind farm g The power system droop coefficient for a data sample Representing the i The first wind farm g The inertia coefficient of a power system in a data sample; 1-3. Arrange the output samples and the upgraded input samples in order to obtain the output sample set and the upgraded input sample set. The upgraded input sample set is defined as follows:

[0011] The output sample set is defined as:

[0012] 1-4. Construct the power generation during the primary frequency regulation start-up phase of the wind turbine through least squares data-driven training. Final fan speed Wind farm droop coefficient Inertia support coefficient Wind speed Power system droop coefficient and power system inertia coefficient Least squares data-driven matrix between The formula is as follows:

[0013] in, represent matrix transpose, represent The pseudo-inverse of the matrix; Construct a linear functional relationship for the state-space mapping, defined as:

[0014] Step 2: Based on the linear function relationship obtained in Step 1, evaluate the feasible region of wind farm droop and inertia coefficient through online collaborative iterative analysis. The specific process is as follows: 2-1. Measure the real-time wind speed at each wind farm. Substitute this into the corresponding linear function relationship of the wind farm; 2-2. Setting the iteration step The power system operator iteratively calculates the power system droop coefficient. and power system inertia coefficient As shown in the following formula;

[0015]

[0016]

[0017]

[0018] in, This represents the droop coefficient provided by the thermal power unit; And and The data is distributed to each wind farm and substituted into the linear function formula of each wind farm. 2-3. On the wind farm side, based on the state-space mapping linear function relationship and the power system droop coefficient... and inertia coefficient Iteratively solve the wind farm droop coefficient and And report it to the power system operator; 2-3-1. Define the permissible final turbine speed limit during the primary frequency regulation process of a wind farm. and the active power limit of the wind turbine ,in, Wind turbine active power limit Defined as:

[0019] Ultimate speed limit of the wind turbine Defined as:

[0020] in, Least squares data-driven linear matrix The first line in Represents a least-squares data-driven linear matrix The second line in represent t +1 iteration update i Inertia support coefficient of a wind farm represent t +1 iteration update i The droop coefficient of a wind farm; 2-3-2. Set initial values ​​for wind farm sag and inertia support coefficients during online evaluation, where the initial value for the wind farm sag coefficient is... Initial value of wind farm inertia support coefficient ; 2-3-3. Solving the active power limit equation of the wind turbine using the bisection method The feasible boundary of the inertia support coefficient is obtained through iterative updates. ; 2-3-4. Solving for the final speed limit of the fan using the bisection method The equation yields the feasible boundary for iteratively updated droop coefficients. 2-3-5 The power system operation side receives the droop coefficient reported by each wind farm. and inertia support coefficient Update the system coefficients, and perform the above iterative process between the wind farm and the power system until convergence; for each wind farm, determine the inertia support coefficient. and droop coefficient Does it meet one of the following convergence conditions?

[0021] in, and The convergence condition is set. If convergence condition one is met, then output the droop coefficient. and inertia support coefficient The feasible region of the optimal droop-inertia frequency regulation coefficient for each wind farm is obtained by solving the problem and reported to the power system operator. If the condition is not met, proceed to step 2-3-3 and continue iteratively performing the wind farm-side assessment. 2-4. Droop coefficients reported by the power system operation side to each wind farm and inertia support coefficient These two coefficients are summed separately to update the system coefficients; the above iterative process is performed between the wind farm and the power system until convergence, and the power system droop coefficient is determined. Does the inertia coefficient satisfy convergence condition two?

[0022] If convergence condition two holds, then the feasible boundary for the output power system droop coefficient is... and inertia coefficient feasible boundary Thus, the feasible region of the optimal droop-inertia frequency regulation coefficient for each wind farm and its power system can be obtained; If the above conditions are not met, proceed to step 2-3 to continue the collaborative iteration between the wind farm and the power system operation side.

[0023] Furthermore, the number of enhanced observation dimensions is set to... , Defined as:

[0024] Defined as:

[0025] in, Representative and input sample Basis vectors of the same dimension, basis vector matrix The value of the middle element is Random values ​​within a range Representing the e The nonlinear enhancement of the observation dimension mapping scalar function in dimension Representing the i The first wind farm g Nonlinear enhanced observation dimension mapping vector function for each data sample.

[0026] Furthermore, the input samples are up-dimensioned in the state space to obtain the up-dimension input samples, defined as:

[0027] in, Representing the i The first wind farm g Nonlinear enhanced observation dimension mapping vector function for each data sample.

[0028] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention has the advantages of not relying on the accuracy of static model parameters, making full use of the frequency regulation capability of wind farms through a collaborative iterative framework, avoiding the computational burden of centralized solution on the system side, reducing the communication burden of traditional feasible domain transmission, and considering the mutual influence between the two parameters. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the overall process of the state-space mapping evaluation method for the feasible region of wind farm sag and inertia support coefficient of the present invention. Figure 2 Here are the detailed flowcharts: (a) Detailed flowchart of step 1, (b) Detailed flowchart of step 2. Detailed Implementation

[0030] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0031] The present invention provides a method for evaluating the feasible region state-space mapping of wind farm sag and inertia support coefficient, the method comprising the following steps: Step 1: Utilize the operational data from the historical database of each wind farm in the power system, which involves the primary frequency regulation and inertia support processes, to offline acquire data for each wind farm (as shown in the first step) through state-space mapping and least squares training. i (Taking a wind farm as an example) Power during the primary frequency regulation and inertia support process of the wind turbine in the primary frequency regulation start-up phase Final fan speed Wind farm droop coefficient Inertia support coefficient Wind speed Power system droop coefficient and power system inertia coefficient The state-space mapping linear function relationship model between them; this step specifically includes the following processing: Step 1-1: Obtain the data for each wind farm in the power system. G Data samples ( G ≧1000), with the first iTaking a wind farm as an example, the parameters of the data sample include the power during the primary frequency regulation startup phase of the wind turbine. Final fan speed Wind farm droop coefficient Inertia support coefficient Wind speed Power system droop coefficient and power system inertia coefficient ; Steps 1-2, with the first i Taking a wind farm as an example, input and output samples are constructed based on the data samples, as shown in the following formula: The input data sample is defined as follows: (1) The output data sample is defined as follows: (2) in, Representing the i The first wind farm g The power of a single-stage frequency regulation start-up of a wind turbine is measured in a data sample. Representing the i The first wind farm g The final rotational speed of the wind turbine in each data sample Representing the i The first wind farm g The droop coefficient of each data sample wind farm Representing the i The first wind farm g Inertia support coefficient of each data sample Representing the i The first wind farm g Wind speed of each data sample Representing the i The first wind farm g The power system droop coefficient for a data sample Representing the i The first wind farm g The inertia coefficient of a power system in a data sample; Steps 1-3: Upgrade the state space of the input samples to obtain the upgraded input samples, defined as: (3) in, Representing the i The first wind farm g Nonlinear enhanced observation dimension mapping vector function for each data sample; Assume the number of augmented observation dimensions is ,So Defined as: (4) Defined as: (5) in, Representative and input sample Basis vectors of the same dimension, basis vector matrix The value of the middle element is Random values ​​within a range Representing the e The nonlinear enhancement of the observation dimension is mapped to a scalar function; Steps 1-4: Arrange the output samples and the upgraded input samples in order to obtain the output sample set and the upgraded input sample set. The upgraded input sample set is defined as follows: (6) The output sample set after dimensionality increase is defined as: (7) Steps 1-5: Construct the power generation during the primary frequency regulation startup phase of the wind turbine through least squares data-driven training. Final fan speed Wind farm droop coefficient Inertia support coefficient Wind speed Power system droop coefficient and power system inertia coefficient Least squares data-driven matrix between The formula is as follows: (8) in, represent matrix transpose, represent The pseudo-inverse of the matrix; Construct a linear functional relationship for the state-space mapping, defined as: (9) Step 2: Based on the linear function relationship (8) obtained in Step 1, evaluate the feasible region of wind farm droop and inertia coefficient online in a collaborative iterative manner. The specific execution process is as follows: Step 2-1: Measure the real-time wind speed at each wind farm. Substitute this into the corresponding linear function relationship of the wind farm; Step 2-2: Set the iteration step The power system operator calculates the next iteration step according to formulas (10)-(14). The power system droop coefficient below and power system inertia coefficient , (10) (11) (12) (13) in, This represents the droop factor provided by thermal power units; and the power system droop factor. and power system inertia coefficient The data is distributed to each wind farm and substituted into the linear function formula of each wind farm. Steps 2-3: On the wind farm side, based on the linear function relationship and the power system droop coefficient... and power system inertia coefficient Solve for the droop coefficient of the wind farm and And report it to the power system operator; Step 2-3-1: Set the iteration step Define the permissible final turbine speed limit during the primary frequency regulation process of a wind farm. and the active power limit of the wind turbine ,in, Wind turbine active power limit Defined as: (14) Ultimate speed limit of the wind turbine Defined as: (15) in, Least squares data-driven linear matrix The first line in Represents a least-squares data-driven linear matrix The second line in represent t +1 iteration update i Inertia support coefficient of a wind farm represent t +1 iteration update i The droop coefficient of a wind farm; Step 2-3-2: Set the initial values ​​for wind farm sag and inertia support coefficients during the online evaluation process, where the initial value for the wind farm sag coefficient is... Initial value of wind farm inertia support coefficient ; Step 2-3-3: Solve the active power limit equation of the wind turbine using the bisection method. The feasible boundary of the inertia support coefficient is obtained through iterative updates. ; Steps 2-3-4: Use the bisection method to solve for the final speed limit of the fan. The equation yields the feasible boundary for iteratively updated droop coefficients. Steps 2-3-5: For each wind farm, determine the inertia support coefficient. and droop coefficient Does it meet the following convergence conditions: (16) in, and The convergence condition is set. If the condition in formula (17) is met, then the droop coefficient will be output. and inertia support coefficient This serves as the feasible region for wind farm sag and inertia support coefficients; and is reported to the power system operator. If the above conditions are not met, then the iteration step... Proceed to step 2-4-3 to continue iterating and perform wind field side assessment; Steps 2-4: The power system operation side reports the droop coefficient to each wind farm. and inertia support coefficient These two coefficients are summed separately to update the system coefficients; the above iterative process is performed between the wind farm and the power system until convergence, and the power system droop coefficient is determined. and inertia coefficient Does it meet the convergence condition in formula (17)? (17) If the condition in formula (18) holds, then the feasible boundary for the output power system droop coefficient is... and inertia coefficient feasible boundary This allows us to solve for the feasible region of the optimal droop-inertia frequency regulation coefficient for each wind farm and its associated power system. If the above conditions are not met, then let the iteration step... Then proceed to steps 2-3 to continue the collaborative iteration between the wind farm and the power system operation side. , , , These are four fixed values ​​set according to needs.

[0032] In summary, this invention proposes a collaborative iterative evaluation framework between wind farms and the power system operator. The wind farm utilizes historical operational data from primary frequency regulation (PMG) to construct a linear state-space mapping relationship between the final turbine speed, maximum turbine power, wind farm droop coefficient, wind farm inertia support coefficient, wind speed, power system droop coefficient, and power system inertia coefficient during PMG and inertia support processes. Real-time wind speed is measured at the wind farm, and the power system droop and inertia parameters are received. Using the constructed linear relationship between these variables, considering the final turbine speed and power limits, the feasible region boundaries of the wind farm droop and inertia support coefficients are evaluated online and reported to the power system operator. The power system operator sums the two coefficients reported by each wind farm, updates the system coefficients, and distributes them to each wind farm. The above iterative process is executed between the wind farm and the power system until convergence, solving for the optimal feasible region of the droop-inertia frequency regulation coefficients for each wind farm.

[0033] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should also be considered to fall within the scope of protection of the present invention.

Claims

1. A method for mapping the state space of the feasible region of droop and inertia support coefficients of a wind farm, characterized in that, include: Step 1: Utilizing the operational data from the historical databases of each wind farm in the power system that participated in primary frequency regulation and inertia support processes, offline data samples for each wind farm are obtained through state-space mapping and least squares training, and a state-space mapping linear function relationship is constructed. This step specifically includes the following processing: Step 1-1: Obtain the data for each wind farm in the power system. G A data sample, the parameters of which include the power during the primary frequency regulation start-up phase of the wind turbine. Final fan speed Wind farm droop coefficient Inertia support coefficient Wind speed Power system droop coefficient and power system inertia coefficient ; Steps 1-2: Construct the following formula based on the data sample: The input data sample is defined as follows: The output data sample is defined as follows: in, Representing the i The first wind farm g The power of a single-stage frequency regulation start-up of a wind turbine is measured in a data sample. Representing the i The first wind farm g The final rotational speed of the wind turbine in each data sample Representing the i The first wind farm g The droop coefficient of each data sample wind farm Representing the i The first wind farm g Inertia support coefficients for each data sample Representing the i The first wind farm g Wind speed of each data sample Representing the i The first wind farm g The power system droop coefficient for a data sample Representing the i The first wind farm g The inertia coefficient of a power system in a data sample; Steps 1-3: Arrange the output samples and the upgraded input samples in order to obtain the output sample set and the upgraded input sample set. The upgraded input sample set is defined as follows: The output sample set is defined as: Steps 1-4: Construct the power generation during the primary frequency regulation startup phase of the wind turbine through least squares data-driven training. Final fan speed Wind farm droop coefficient Inertia support coefficient Wind speed Power system droop coefficient and power system inertia coefficient Least squares data-driven matrix between The formula is as follows: in, represent matrix transpose, represent The pseudo-inverse of the matrix; Construct a linear functional relationship for the state-space mapping, defined as: Step 2: Based on the linear function relationship obtained in Step 1, conduct online collaborative iterative evaluation of the wind farm. The feasible region for droop and inertia coefficients is determined as follows: Step 2-1: Measure the real-time wind speed at each wind farm. Substitute this into the corresponding linear function relationship of the wind farm; Step 2-2: Set the iteration step The power system operator iteratively calculates the power system droop coefficient. and power system inertia coefficient As shown in the following formula; in, This represents the droop coefficient provided by the thermal power unit; And and The data is distributed to each wind farm and substituted into the linear function formula of each wind farm. Steps 2-3: On the wind farm side, based on the state-space mapping linear function relationship and the power system droop coefficient... and inertia coefficient Iteratively solve the wind farm droop coefficient and And report it to the power system operator; Step 2-3-1: Define the permissible final turbine speed limits during the primary frequency regulation process of the wind farm. and the active power limit of the wind turbine Among them, the active power limit of the wind turbine Defined as: Ultimate speed limit of the wind turbine Defined as: in, Least squares data-driven linear matrix The first line in Represents a least-squares data-driven linear matrix The second line in represent t +1 iteration update i Inertia support coefficient of a wind farm represent t +1 iteration update i individual wind farm droop coefficient; Step 2-3-2: Set the initial values ​​for wind farm sag and inertia support coefficients during the online evaluation process, where the initial value for the wind farm sag coefficient is... Initial value of wind farm inertia support coefficient ; Step 2-3-3: Solve the active power limit equation of the wind turbine using the bisection method. The feasible boundary of the inertia support coefficient is obtained through iterative updates. ; Steps 2-3-4: Use the bisection method to solve for the final speed limit of the fan. The equation yields the feasible boundary for iteratively updated droop coefficients. ; Steps 2-3-5: The power system operation side receives the droop coefficient reported by each wind farm. and inertia support coefficient Update the system coefficients, and perform the above iterative process between the wind farm and the power system until convergence; for each wind farm, determine the inertia support coefficient. and droop coefficient Does it meet one of the following convergence conditions? in, and The convergence condition is set. If convergence condition one is met, then output the droop coefficient. and inertia support coefficient The feasible region of the optimal droop-inertia frequency regulation coefficient for each wind farm is obtained by solving the problem and reported to the power system operator. If the condition is not met, proceed to step 2-3-3 and continue iteratively performing the wind farm-side assessment. Steps 2-4: The power system operation side reports the droop coefficient to each wind farm. and inertia support coefficient These two coefficients are summed separately to update the system coefficients; the above iterative process is performed between the wind farm and the power system until convergence, and the power system droop coefficient is determined. and inertia coefficient Does it meet convergence condition two? If convergence condition two holds, then the feasible boundary for the output power system droop coefficient is... and inertia coefficient feasible boundary Thus, the feasible region of the optimal droop-inertia frequency regulation coefficient for each wind farm and its power system can be obtained; If the above conditions are not met, proceed to step 2-3 to continue the collaborative iteration between the wind farm and the power system operation side.

2. The method for evaluating the feasible region state-space mapping of wind farm sag and inertia support coefficient according to claim 1, characterized in that, Set the number of enhanced observation dimensions to , Defined as: Defined as: in, Representative and input sample Basis vectors of the same dimension, basis vector matrix The value of the middle element is Random values ​​within a range Representing the e The nonlinear enhancement of the observation dimension mapping scalar function in dimension Representing the i The first wind farm g Nonlinear enhanced observation dimension mapping vector function for each data sample.

3. The method for evaluating the feasible region state-space mapping of wind farm sag and inertia support coefficient according to claim 1, characterized in that, The input sample is up-dimensioned through the state space to obtain the up-dimension input sample, defined as: in, Representing the i The first wind farm g Nonlinear enhanced observation dimension mapping vector function for each data sample.

Images