New energy base rapid frequency supporting capability monitoring method based on pooling station monitoring data

Abstract

The invention relates to the technical field of power system operation and control, and discloses a pooling station monitoring data-based new energy base rapid frequency supporting capability monitoring method, which comprises the following steps of: firstly, automatically identifying a disturbance moment based on frequency and power change characteristics, and carrying out smoothing and exception suppression processing on disturbance data; then combining frequency response mechanisms of wind power, photovoltaic and energy storage equipment to construct an inertia response characteristic and primary frequency modulation characteristic model of the new energy base, and identifying an equivalent inertia constant and an equivalent droop coefficient of the base by using measurement data during disturbance; solving the external equivalent reactance of the new energy base based on the sensitivity relationship between the complex power and the voltage phasor; and finally, calculating the fast frequency support capacity, sustainable support duration and frequency response index of the new energy base in combination with the parameters, thereby realizing online monitoring and evaluation of the fast frequency support capacity.

Description

Technical Field

[0001] This invention belongs to the field of power system operation and control technology, and specifically relates to a method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from collection stations. Background Technology

[0002] As the proportion of new energy sources such as wind power and photovoltaics in the power system continues to increase, the overall inertia level and primary frequency regulation capability of the power grid are declining. Traditional power systems rely on the mechanical rotational inertia of synchronous generators to suppress rapid frequency drops during disturbances and achieve frequency stability through primary frequency regulation. However, under conditions of high proportion of new energy sources, a large number of power electronic grid-connected devices lack physical inertia, and their frequency response depends entirely on control strategies. This makes the power grid more prone to problems such as large frequency drops, rapid changes, and slow recovery processes when encountering disturbances. Therefore, accurately obtaining the actual inertia support and primary frequency regulation capability of new energy bases has become an important technical requirement for maintaining power grid frequency security.

[0003] New energy bases typically adopt a large-scale, centralized topology, such as... Figure 1 As shown in the diagram. In such bases, various resources, including wind power, photovoltaics, electrochemical energy storage, and a small amount of solar thermal and supporting thermal power, are connected to the collection station in a radial manner, and finally connected to the power grid through smart transformers. This centralized topology is suitable for comprehensive monitoring and unified dispatching. Only phasor measurement units (PMUs) need to be deployed at the collection station to collect key operating parameters such as voltage phasors, frequency, active power, reactive power, and phase angle of the entire new energy base in real time. Nevertheless, the power systems of large-scale new energy bases often exhibit characteristics such as long access branches, long electrical distances between nodes, and relatively weak grid structures, resulting in complex and variable operating environments. Since these bases typically lack large conventional power sources and are electrically weakly connected to the main grid, their frequency support mainly relies on the control strategies of the new energy devices themselves, the rapid response capability of electrochemical energy storage, and the inertial support capability of a small number of synchronous units. These conditions lead to significant uncertainty in the dynamic response of new energy bases under frequency disturbances, placing higher demands on the accurate quantification of their rapid frequency support capabilities.

[0004] Current monitoring of renewable energy frequency support capabilities largely relies on SCADA data from power plants or control models provided by equipment manufacturers. SCADA data suffers from low sampling rates and poor time synchronization, making it difficult to reflect millisecond-level frequency changes at the initial stage of disturbances. Methods based on equipment models are hampered by factors such as opaque unit parameters, large variations in operating conditions, and poor model adaptability, failing to accurately describe the dynamic response of renewable energy under real grid disturbances. Furthermore, existing methods struggle to simultaneously obtain the equivalent reactance of renewable energy bases to the external environment and their coupling characteristics with the upstream grid, resulting in a lack of quantitative data consistent with actual grid connection points for support capability assessment.

[0005] In recent years, the deployment of Power Measurement Units (PMUs) at renewable energy aggregation stations has made it possible to acquire high-resolution, highly synchronous dynamic measurement data. This provides a data foundation for directly inverting the inertial constant, droop coefficient, and equivalent reactance of renewable energy bases based on disturbance measurements. However, existing PMU-based analysis methods mostly remain at the level of frequency event recording and simple statistics, lacking a systematic approach encompassing disturbance identification, data cleaning, frequency response modeling, parameter identification, and quantitative assessment of support capabilities. Especially in real-world scenarios with high noise levels, multiple disturbances, and complex unit responses, how to continuously and stably monitor the rapid frequency support capabilities of renewable energy bases using PMU data remains an unsolved technical problem.

[0006] Therefore, this application designs a method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from collection stations to solve the above problems. Summary of the Invention

[0007] The purpose of this invention is to solve the technical problems mentioned in the background section and to provide a method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from data collection stations.

[0008] This invention is implemented as follows:

[0009] The method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from data collection stations includes the following steps:

[0010] Step (1): Configure a phasor measurement unit on the side of the new energy base collection station to synchronously collect the voltage phasor, system frequency, frequency change rate, active power and reactive power measurement data of the collection station bus to form time series monitoring data;

[0011] Step (2): Based on the time series monitoring data, the frequency disturbances during the system operation are automatically identified, and the data is divided into the steady-state period before the disturbance, the period at the moment the disturbance occurs, and the dynamic recovery period after the disturbance.

[0012] Step (3): Based on the frequency response mechanism of wind turbine, photovoltaic power generation unit and electrochemical energy storage device, establish a primary frequency regulation response model and an inertial response model for the new energy base. The model expresses the relationship between output power and frequency deviation and frequency change rate as a piecewise linear relationship with dead zone, distinguishing the rapid power boosting segment of wind power and photovoltaic, the inertial support segment of energy storage and the primary frequency regulation segment.

[0013] Step (4): Substitute the frequency deviation, frequency change rate and active power change obtained from the collection station side into the inertial response model and the primary frequency regulation model, and obtain the equivalent inertial constant and equivalent droop coefficient of the new energy base as a whole through the online parameter identification algorithm, so as to obtain the rapid frequency support capability of the new energy base in the early stage and middle and late stage of the disturbance.

[0014] Step (5): During the disturbance period, based on the relationship between complex power injection and grid admittance matrix, sensitivity analysis is performed on the voltage amplitude and voltage phase angle changes of the substation bus, and the partial derivatives of active and reactive power increments with respect to voltage phasor increments are obtained, thereby obtaining the external equivalent reactance of the new energy base, which is used to characterize the electrical coupling strength between the new energy base and the upper-level grid.

[0015] Step (6): The PMU time series data is subjected to weighted least squares fitting and smoothing. The frequency trajectory and active power trajectory are denoised and outlier removed. The smoothed frequency deviation, frequency change rate and active power change are used as the iterative update inputs for inertia constant, droop coefficient and equivalent reactance.

[0016] Step (7): Based on the updated equivalent inertia constant, equivalent droop coefficient and equivalent reactance, calculate the maximum available fast frequency support capacity, sustainable support duration and characteristic indicators of the frequency response process of the new energy base under a given disturbance condition, and use the results as the monitoring output of the fast frequency support capability of the new energy base.

[0017] Furthermore, steps (1) and 2 specifically include:

[0018] The phasor measurement unit synchronously collects the phasor information of the three-phase voltage and current of the busbar at the collection station with a high sampling rate, and calculates the system frequency and frequency change rate in real time, while recording the active power, reactive power and voltage phase angle.

[0019] The absolute value of the frequency change rate and the active power change rate are compared with preset thresholds. When either quantity exceeds the threshold, a disturbance is determined to have occurred. The sampling time when the threshold is first exceeded is taken as the disturbance start time. Time windows containing several sampling points are extracted before and after the threshold as disturbance analysis intervals.

[0020] By continuously performing threshold detection on subsequent monitoring data through a sliding time window, the system can automatically identify and segment multiple disturbance events.

[0021] Furthermore, the primary frequency regulation response model and rapid support model of the wind turbine and photovoltaic power generation unit in step (3) include:

[0022] Establish a linear droop relationship between the active power output of wind power and photovoltaic power and the system frequency deviation. Keep the output power constant within the preset primary frequency regulation dead zone, and adjust the output power proportionally according to the droop coefficient outside the dead zone.

[0023] When the frequency deviation exceeds the rapid support trigger threshold, the rapid power boost function of wind power or photovoltaic is triggered. Under the premise of not exceeding the unit's available power margin and technical limits, the output active power is increased or decreased in a short period of time to support the rapid frequency recovery.

[0024] The rapid power boosting process uses a piecewise linear relationship to describe the underfrequency support section and the overfrequency reduction section, and sets upper and lower limits to ensure the safe operation of the unit.

[0025] Furthermore, the frequency response model of the electrochemical energy storage device in step (3) includes:

[0026] In the inertial support stage, the change in active power of the energy storage device is established in a proportional relationship with the frequency change rate. When the frequency change rate is large, the power is released or absorbed quickly to simulate the suppression effect of the rotor inertia of the synchronous generator on the frequency change.

[0027] During the primary frequency regulation phase, a linear droop relationship is established between the active power adjustment of the energy storage device and the frequency deviation. There is no response within the primary frequency regulation dead zone, and the power is gradually adjusted according to the set droop coefficient outside the dead zone.

[0028] Based on the state of charge of the energy storage device, the available support capacity and maximum support duration are limited. When the state of charge is below the lower limit or above the upper limit, the inertial and primary frequency regulation support is automatically reduced or withdrawn to ensure the safe operation of the energy storage.

[0029] Furthermore, the process of obtaining the equivalent inertia constant and equivalent droop coefficient of the new energy base in step (4) includes:

[0030] The new energy base is equivalent to a virtual synchronous generator. The input of the virtual machine group is the frequency deviation, frequency change rate and active power change measured at the collection station, and the output is the equivalent inertial constant and equivalent droop coefficient.

[0031] Within the disturbance analysis time window, based on the dynamic equilibrium relationship of the virtual synchronous machine, an algebraic equation is constructed between the equivalent inertial constant and the change in active power and the rate of change in frequency, and an algebraic equation is constructed between the equivalent droop coefficient and the frequency deviation and the change in active power.

[0032] By employing recursive least squares or similar online identification algorithms, the equivalent inertia constant and equivalent droop coefficient are updated at each sampling time to achieve dynamic tracking of the frequency support capability of new energy bases.

[0033] Furthermore, the process of obtaining the equivalent reactance of the new energy base in step (5) includes:

[0034] Taking the substation busbar as the research object, an equivalent multi-busbar network model including the new energy base and its upstream power grid is established, and the voltage and injection current of each busbar are represented by the admittance matrix.

[0035] Within the disturbance time window, the partial derivatives of the increments of active and reactive power of the collecting station bus with respect to the increments of bus voltage angle and voltage amplitude are calculated, and these are used as the sensitivity of power to voltage.

[0036] Based on the aforementioned sensitivity, the equivalent reactance parameter is calculated inversely to characterize the impact of the new energy base on the bus voltage and power distribution during frequency support, thus serving as a quantitative indicator of the strength of the power grid and the electrical coupling characteristics.

[0037] Furthermore, the process of performing weighted least squares fitting and smoothing on the PMU time series data in step (6) includes:

[0038] Within each disturbance analysis time window, frequency, rate of change of frequency, and change of active power are regarded as fitting objects with respect to time, and linear or piecewise linear fitting functions are established.

[0039] Weights are assigned to each sampling point within the window, with data points closer to the disturbance center having larger weights and data points farther from the disturbance center having smaller weights, in order to highlight the fitting accuracy for key response stages; a saturated weight function is introduced during the fitting process to automatically reduce the weights or remove outlier data with large residuals, thereby reducing the impact of measurement noise and instantaneous shocks on the parameter estimation results;

[0040] The fitted smooth frequency trajectory, frequency change rate trajectory, and active power change trajectory are used as input data for subsequent iterative updates of the equivalent inertia constant, equivalent droop coefficient, and equivalent reactance.

[0041] Furthermore, the iterative update of the inertia constant, droop coefficient, and equivalent reactance in step (6) adopts the following mechanism:

[0042] Construct a residual vector consisting of frequency deviation, frequency change rate, and active power change, and construct a Jacobian matrix from the partial derivatives of the residual vector with respect to each parameter to be identified.

[0043] Based on Gauss-Newton or Levenberg-Marquardt iterative algorithms, the equivalent inertial constant, equivalent droop coefficient, and equivalent reactance are jointly iteratively corrected using residual vectors and Jacobian matrices.

[0044] When the norm of the residual is less than the preset threshold or the number of iterations reaches the upper limit, the current iteration result is output as the frequency support capability parameter of the new energy base under this disturbance event.

[0045] Furthermore, the monitoring results of the rapid frequency support capability of the new energy base output in step (7) include:

[0046] The equivalent inertial constant reflects the ability of a new energy base to suppress the rate of frequency change in the early stages of disturbance.

[0047] The equivalent droop coefficient reflects the sensitivity of the new energy base to active power regulation of frequency deviation during the primary frequency regulation stage.

[0048] Equivalent reactance parameters reflect the electrical coupling strength between the new energy base and the upstream power grid and the degree of influence on voltage deviation;

[0049] The maximum available fast frequency support capacity and sustainable support duration under a given typical disturbance amplitude are used to assess the ability of new energy bases to undertake inertial support and primary frequency regulation support obligations.

[0050] Dynamic performance indicators such as the expected minimum frequency and recovery time under corresponding disturbance conditions are used to provide early warning and decision-making basis for dispatching agencies and operators.

[0051] Furthermore, this method is used by dispatch control centers or regional power grid control centers to conduct online monitoring of large-scale wind, solar, and energy storage bases connected to the grid, and has the following application modes:

[0052] The monitoring process is automatically invoked under different types of disturbance scenarios, including continuous load change scenarios, single large unit disconnection scenarios, and multiple unit disconnection scenarios, and the rapid frequency support capability of the new energy base under each scenario is classified and evaluated.

[0053] The equivalent inertial constant, equivalent droop coefficient, and equivalent reactance time series obtained from each disturbance event are statistically analyzed to form a long-term evaluation curve of the frequency support capability of the new energy base, providing a basis for the transformation plan, control strategy optimization, and auxiliary service assessment of the new energy base.

[0054] When the monitoring results are below the preset threshold, an early warning is issued to the dispatcher, prompting the need to adjust the output of new energy sources, optimize energy storage control strategies, or activate other frequency support resources, thereby improving the frequency security margin of the power system under the high penetration of new energy sources.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] 1. This invention acquires dynamic measurement data of frequency, phase angle, and power using a high-precision phasor measurement unit at the new energy base's collection station. This allows key characteristics of frequency disturbances to be recorded in a high-timescale, highly synchronized, and highly sensitive manner, providing a realistic and detailed operational basis for the online identification of the new energy base's inertial support and primary frequency regulation capabilities. Compared to existing methods that rely on SCADA or low-sampling data from within the station, this invention can accurately capture millisecond-level frequency change characteristics in the early stages of disturbances, effectively avoiding inertial estimation bias caused by insufficient data resolution, and making the monitoring results closer to the actual operating state.

[0057] 2. This invention establishes an equivalent frequency response model for wind power, photovoltaic, and energy storage devices, and combines this with dynamic measurement information during disturbances to identify the overall equivalent inertia constant, equivalent droop coefficient, and external equivalent reactance of the new energy base online. This enables a quantitative assessment of the rapid frequency support capability of the new energy base without needing to obtain the unit's internal control strategy or rely on manufacturer parameters. This method directly reflects the actual response level of the new energy base during real disturbances and can reveal the changing patterns of frequency support capability under different operating conditions, seasons, and output combinations, providing an empirical basis for the operation management and auxiliary service assessment of new energy bases.

[0058] 3. This invention introduces adaptive data processing mechanisms such as weighted least squares to suppress measurement noise, abnormal data, and multiple disturbances. It also constructs a continuous-time-scale capability assessment result through the recursive update of equivalent parameters, making the monitoring process more stable and robust. Based on the inertia constant, droop coefficient, support capacity, support duration, and frequency minimum point obtained by this invention, the power grid dispatching agency can more accurately grasp the actual frequency support level of the new energy base, rationally allocate frequency regulation resources, improve the frequency safety margin of the system under the condition of high proportion of new energy penetration, and provide a clear direction for the performance improvement and control strategy optimization of the new energy base. Attached Figure Description

[0059] Figure 1 This is the framework for the new energy base of the present invention;

[0060] Figure 2 This is a block diagram illustrating the principle of monitoring the high-frequency support capability of new energy bases according to the present invention.

[0061] Figure 3 This is a simulation model diagram of the present invention;

[0062] Figure 4 This is a diagram showing the monitoring results of the synchronizing machine under sudden load changes according to the present invention;

[0063] Figure 5 The figures show the experimental results under three frequency events of this invention;

[0064] Figure 6 This is an SFR equivalent model diagram of the new energy base of this invention;

[0065] Figure 7 This is a verification diagram of the frequency response model of the present invention.

[0066] Figure 8 This is a Monte Carlo analysis diagram of the present invention. Detailed Implementation

[0067] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0068] The implementation of the present invention will be described in detail below with reference to specific embodiments.

[0069] Example 1: This example provides a rapid frequency support capability monitoring method for new energy bases based on monitoring data from a collection station. This method analyzes the frequency changes and power responses of wind power, photovoltaic, and electrochemical energy storage equipment in the new energy base during disturbances, constructing an equivalent inertial model and an equivalent droop frequency modulation model suitable for grid-connected new energy groups. Furthermore, it achieves real-time monitoring of the frequency support capability of the new energy base based on PMU measurement data. This method does not require access to the site to obtain equipment parameters and does not rely on traditional synchronous machine models. It can directly utilize measurement information such as voltage phase angle, frequency, and power from the collection station side to achieve a complete assessment of the dynamic frequency support level of the new energy base under small and large disturbances.

[0070] After the disturbance occurs, the frequency response of the wind turbine can be expressed by equation (1):

[0071]

[0072] Where P is the active power output of the wind turbine; P0 is the initial power before the disturbance; δ is the primary frequency regulation droop rate; P n f is the rated power; f is the real-time system frequency; f d This is the primary frequency modulation dead zone; f n This is the system's rated frequency. This formula describes the linear active power response of a wind turbine under frequency deviation and is the basis for the overall primary frequency regulation modeling of a new energy base.

[0073] Outside the primary frequency regulation range, the fast frequency support is mainly achieved by the temporary energy storage and speed release of the wind power inverter, and its active power response can be written as equation (2):

[0074]

[0075] Where ΔP is the additional active power change during the rapid support phase; K L1 K H1 These represent the frequency response coefficients in the under-frequency and over-frequency regions, respectively; f L1 f H1 This is the corresponding trigger threshold.

[0076] As a key regulating unit in new energy bases, energy storage systems can provide millisecond-level inertial support and sub-second-level primary frequency regulation support during frequency disturbances.

[0077] The change in active power during its inertial phase can be expressed by equation (3):

[0078]

[0079] Among them, T j P is the inertial time constant of the energy storage power station; N This is the rated power of the energy storage.

[0080] Equation (3) reflects the rate of change of electrochemical energy storage frequency. The inertial equivalent active response under the given conditions. The response of the energy storage system during the primary frequency regulation stage can be approximated by equation (4):

[0081]

[0082] Among them, K f f is the primary frequency regulation coefficient for energy storage; d1 This is a frequency modulation dead zone.

[0083] While there are no synchronous generators within the new energy power station, to describe the overall frequency response characteristics of the system, the synchronous machine inertial model can be used as a reference to construct the equivalent inertial response equation of the new energy base. The synchronous machine inertial response can be described by equations (5) and (6):

[0084]

[0085] p m =K(ω-ω0)(6)

[0086] Among them, M G =2H G For the rotor inertia term of the unit; ω G ω0 is the actual rotational speed angular frequency; ωp is the rated system frequency; m For mechanical power; p G For electromagnetic power; D G is the damping coefficient; K is the speed regulation ratio coefficient. In this embodiment, the model is used to construct the overall equivalent inertial structure of the new energy base, that is, to equate a large number of wind, solar and energy storage devices to a unified "virtual inertial unit".

[0087] Based on the above model, the structured equations for the equivalent inertial constant and equivalent droop characteristics of the new energy base can be further written, as shown in equation (7):

[0088]

[0089] In the formula: H D,h K represents the equivalent inertial constant of the new energy base. D The equivalent droop coefficient; Δω h For the frequency deviation of the bus at the collection station; Δp e The active power change measured at the collection station.

[0090] For the derivation of the equation of motion of the synchronous generator rotor, the above equation can be rewritten as:

[0091]

[0092] In the formula: It can be estimated through measurements from the aggregation station PMU, and:

[0093]

[0094] In the formula: x D,h It is the equivalent reactance of the object being measured; It is the derivative of the active power quota used to adjust the frequency of bus h.

[0095] Equation (8) may have numerical problems in actual calculations. That is, if the denominator α changes sign and crosses zero a few seconds before the disturbance occurs, the equation will have singularity. To avoid this singularity, the following differential equation is used to determine the equivalent inertial constant and the equivalent droop coefficient:

[0096]

[0097] in,

[0098] In the formula: T M T represents the coefficient used in calculating the equivalent inertial constant. D This represents the coefficient used when calculating the equivalent droop coefficient.

[0099] As can be seen from equation (9), to calculate the equivalent inertia constant of a new energy base, it is necessary to first know its internal equivalent reactance. The internal reactance of a single device such as a synchronous generator is easy to measure [a reference needs to be added, as it is not common knowledge]. However, a new energy base contains a variety of resources and its electrical network topology, making it difficult to measure its equivalent reactance. Therefore, this paper proposes a method for estimating the equivalent reactance of a new energy base.

[0100] First, consider injecting complex power into a power grid with n buses. Using the power flow equations, the complex power can be expressed as:

[0101]

[0102] In the formula: These are the active and reactive power injections for each bus; It is the vector of voltages for each bus; This is the admittance matrix of the network; represents element-wise multiplication, and * represents the conjugate of a complex number.

[0103] Rewriting equation (12), omitting the time parameter for easier subsequent derivation, we write the h-th element of p and q as:

[0104]

[0105] In the formula: p h and q h It is the h-th element of p and q; G h,k and B h,k yes The real and imaginary parts of the (h, k) elements v k It is the voltage amplitude at bus k; θ h,k =θ h -θ k θ h and θ k These are the voltage phase angles of bus h and k, respectively. Differentiating equation (13), and rewriting the active and reactive power injection deviations of bus h as the sum of two components:

[0106]

[0107] In the formula: dp h ′ and dq h ′ represents the active and reactive power quotas related to changes in the phase angle of the bus voltage; dp h "and dq h "" refers to the quota of active and reactive power related to the change in the amplitude of the bus voltage.

[0108] Where dq″ h and dp″ h Approximately:

[0109]

[0110] In the formula: It is B ′ Element; It is B ″ The elements; ω is the vector of bus frequencies; This represents the normalized transient rate of change of the bus voltage.

[0111] After obtaining the equivalent reactance, the parametric decomposition relationships described by equations (16)-(20) can be obtained:

[0112]

[0113] In the formula: It is the h-th element of ρ.

[0114] Again B″ h It can be obtained from the following formula:

[0115] In the formula: B h " is the equivalent internal susceptance of the equipment at bus h, B" D,h The equivalent reactance of the subnetwork is given by equations (16) and (17).

[0116]

[0117] Due to the equivalent reactance x D,h It is susceptance B D,h The reciprocal of, therefore:

[0118]

[0119] The final equivalent reactance is then estimated using an adaptive weighting method, employing equations (21) and (22):

[0120]

[0121] In the formula: γ(α) is determined by formula (21); y It is a very small positive threshold, which helps to reduce the influence of noise and improve the accuracy of γ(y); y The value range can generally be selected from

[10] . -7 10 -5 ]; T x The value range can generally be selected from

[10] . -2 10 -1 ].

[0122] In summary, the overall framework for the monitoring method of high-frequency support capability of new energy bases can be obtained. Figure 2 As shown in the figure;

[0123] To improve the robustness of the parameter estimation process, the time series data acquired by the PMU are smoothed. Let the PMU data be D = {(t n ,y n )}, n=1,2,3....,t n It is the nth time window, y nIt is the value of the nth data point, and N is the window length of the time series.

[0124] The optimization objective function for each time window is shown below:

[0125]

[0126] In the formula: w i It is the weight of each data point; y i It is the numerical value of each data point; f(t) i ;θ) is the fitted function; θ is the parameter vector of the fitted function.

[0127] By using a linear fitting structure Y=kT+b, we can obtain equations (24)-(26):

[0128] Y = kT + b(24)

[0129] In the formula: k can be obtained by weighted least squares method; k and b are θ in the above formula; T is the vector of t; Y is the vector of y.

[0130] k=(T T WT) -1 T T WY(25)

[0131] b = Y - kT(26)

[0132] Then, the smoothed data sequence Y is obtained. WLS The parameters are iteratively updated by calculating the smoothing residual R and the Jacobian matrix J, as shown in the following formula, until the smoothing residual is less than a specified value, at which point the iteration stops and the optimal parameters are output.

[0133] Finally, we obtain the recursive formulas (27)-(28) that can be used to update the parameters:

[0134] θ k+1 =θ k -(J T J+λE) -1 J T R(27)

[0135] R = YY WLS (28)

[0136] Based on the above modeling, measurement, and parameter identification process, this implementation method can complete the real-time estimation of the frequency support capability of new energy bases within several sampling periods after a disturbance occurs. This method automatically integrates the responses of various equipment such as wind power, photovoltaics, and energy storage into a unified equivalent model, without relying on internal equipment parameters or plant-level modeling. It can be directly used for scheduling zoning, online monitoring, and control strategy optimization.

[0137] Example 2: This example is based on the scheme of Example 1. Through case analysis, it verifies the performance and accuracy of the proposed high-frequency support capability monitoring method for new energy bases. A simulation model is constructed using a new energy base in a certain region of Xinjiang as a blueprint. The simplified power grid of this region is shown below. Figure 3 As shown, the internal bus voltage level of the new energy base is 220kV, and the voltage of bus 1 and 2 at the grid connection point of the substation is 750kV. Some parameters of the simulation model are shown in Table 1. All simulation results were obtained using the power system analysis software tool DOME.

[0138] Table 1 Key parameters of the simulation model

[0139]

[0140] The new energy base includes nine 220kV busbars, which are connected to a 750kV power grid via transformers for external transmission. The base comprises three wind power clusters, one photovoltaic cluster, a supporting 700MW independent energy storage power station, and one weakly connected thermal power station. The wind turbines and photovoltaic units employ grid-connected control with droop control, while the energy storage power station uses grid-connected VSG control.

[0141] In the following experiments, the synchronous generator model is represented by a fourth-order (two-shaft) model, both equipped with a turbine governor (TG) and an automatic voltage regulator (AVR). It is also assumed that a static var compensator (SVC) is installed on bus 2 of the power grid.

[0142] First, data measured at six locations on the busbar where the Synchronous Generator (SG) is located in the new energy base were used to verify the accuracy of the proposed method. The accurate equivalent inertia constant of the SG is known to be 6.2s, the governor droop coefficient is 14.5, and the internal reactance is 0.25pu. The fault setting is that the grid load increases by 3% in 1 second. The experimental results of the SG frequency change, reactance, inertia constant, and droop coefficient are as follows: Figure 4 As shown.

[0143] Figure 4 (a) shows the frequency change at bus 6 where the synchronous machine is located. Due to a sudden increase in the main grid load, the frequency suddenly drops, and then recovers to a quasi-steady state under the action of inertia and primary frequency regulation. Figure 4 In (b), the blue dashed line represents the actual reactance value of the synchronous machine, 0.25 pu, while the red solid line represents the reactance of the synchronous machine calculated by the method in this paper. After undergoing rapid transient changes, the estimated result converges to 0.26, with an accuracy of 96%. Figure 4In (c), the blue curve represents the actual value of the SG inertial constant, which is 6.2s, while the red solid line represents the calculated SG inertial constant, which eventually converges to 6.28s, with an accuracy of 98.7%. Figure 4 In (d), the blue dashed line represents the actual value of the sag coefficient of the synchronizer, which is 14.5, and the red solid line represents the calculated sag coefficient of the synchronizer, which is approximately 14.2 after stabilization, with an accuracy rate of 98%.

[0144] In summary, the method proposed in this paper can effectively calculate the reactance of the synchronous machine and accurately estimate its inertia constant and droop coefficient.

[0145] Experiments will be conducted under different scenarios to verify the effectiveness and adaptability of the high-frequency support capability monitoring method for this new energy base.

[0146] (1) Different Typical Frequency Disturbance Events: Frequency disturbance events are divided into large disturbances and small disturbances. The equivalent inertia constant and equivalent droop coefficient of the new energy base are calculated using data from the busbar 3 collection station when different frequency disturbance events occur. The three frequency events set in the experiment are shown below:

[0147] S1: The system experiences a sudden load change incident, i.e., the load connected to the power grid suddenly decreases by 2.4%;

[0148] S2: WG1 wind farm inside the new energy base is disconnected from the grid;

[0149] S3: The small disturbance event in the system is the change in wind turbine power caused by wind speed fluctuations. The Weibull distribution model is used to simulate random wind speed.

[0150] The experimental results obtained from the simulation are as follows Figure 5 As shown in Table 2.

[0151] Table 2 Experimental results of x, H, and K in different scenarios

[0152]

[0153]

[0154] Depend on Figure 5As shown in Table 3, in the S1 high-frequency event scenario, the equivalent reactance of the new energy base calculated using the proposed method changes during the frequency dynamic process. After the frequency recovers to a steady state, it converges to a fixed value of 0.015, corresponding to an equivalent inertial constant of 16.8s and an equivalent droop coefficient of 11.3. In the S2 low-frequency event scenario, the equivalent reactance of the new energy base also changes, eventually converging to 0.013, with an equivalent inertial constant of 17s and an equivalent droop coefficient of 11.5. In the S3 frequency fluctuation scenario, all parameters of the new energy base fluctuate, with the equivalent reactance approximately 0.009, the equivalent inertial constant approximately 16.4s, and the equivalent droop coefficient approximately 12.2. The results for the three scenarios are basically consistent, indicating that the fast-frequency support capability monitoring method proposed in this paper is effective and can adapt to various scenarios. However, under small disturbance conditions, the results measured by the proposed method fluctuate more significantly compared to those under large disturbance conditions, and its accuracy is slightly lower than that under large disturbance conditions.

[0155] Validation of the frequency response model of the new energy base: The new energy base is equivalent to a synchronous power source model, and the constructed frequency response (SFR) model of the new energy base is as follows: Figure 6 As shown, the equivalent inertia constant and equivalent droop coefficient of the new energy base obtained from the simulation experiment above are substituted into the SFR model to demonstrate the adaptability of the proposed method. Simultaneously, for the simulation system constructed in this section, time-domain simulation is used to calculate the system's frequency characteristics and related indicators of the new energy base. By comparing the degree of agreement between the frequency dynamic characteristics obtained from the SFR model and the time-domain simulation under the same load surge disturbance, the adaptability and accuracy of the calculated equivalent inertia constant and equivalent droop coefficient of the new energy base can be reflected. The constructed SFR model is as follows... Figure 6 As shown in the figure, the frequency curve after the simulation experiment is as follows: Figure 7 As shown.

[0156] Depend on Figure 7 It can be seen that the frequency curves obtained from the time-domain simulation and the SFR model are basically the same, which shows that the equivalent inertia constant and equivalent droop coefficient of the new energy base obtained by the method proposed in this paper are accurate and effective.

[0157] The performance of the high-frequency support capability monitoring method for new energy bases proposed in this embodiment under the influence of measurement noise was demonstrated through 200 Monte Carlo simulation experiments. The experimental results are as follows: Figure 8 As shown.

[0158] Figure 8 The trajectories of the equivalent inertia constant and equivalent droop coefficient of the new energy base obtained from the experiment are shown, where μ and σ represent the mean and standard deviation, respectively. Figure 8In (a), the average value of the equivalent inertial constant μ obtained from the experiment is about 17.2s. Considering the equivalent inertial constant range of 3 times the standard deviation, the range of the equivalent inertial constant is [15.23, 19.17]. The equivalent inertial constant of the new energy base in 200 experiments is within this range. Figure 8 The average value μ of the equivalent droop coefficient obtained in experiment (b) is approximately 10.9. Considering a standard deviation of 3, the range of the equivalent droop coefficient is [12.26, 9.54]. The equivalent droop coefficient of the new energy base in 200 experiments is within this range. Figure 8 (c) shows the frequency at the corresponding collection station in 200 experiments. As can be seen from the figure, the lowest frequency in the experiment varies in the range of [0.9966, 0.9963].

[0159] In summary, random fluctuations cause some errors in the measured equivalent inertia constant and equivalent droop coefficient, but the errors are all within an acceptable range, indicating that the method proposed in this paper has strong robustness.

[0160] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from data collection stations, characterized in that, Includes the following steps: Step (1): Configure a phasor measurement unit on the side of the new energy base collection station to synchronously collect the voltage phasor, system frequency, frequency change rate, active power and reactive power measurement data of the collection station bus to form time series monitoring data; Step (2): Based on the time series monitoring data, the frequency disturbances during the system operation are automatically identified, and the data is divided into the steady-state period before the disturbance, the period at the moment the disturbance occurs, and the dynamic recovery period after the disturbance. Step (3): Based on the frequency response mechanism of wind turbine, photovoltaic power generation unit and electrochemical energy storage device, establish a primary frequency regulation response model and an inertial response model for the new energy base. The model expresses the relationship between output power and frequency deviation and frequency change rate as a piecewise linear relationship with dead zone, distinguishing the rapid power boosting segment of wind power and photovoltaic, the inertial support segment of energy storage and the primary frequency regulation segment. Step (4): Substitute the frequency deviation, frequency change rate and active power change obtained from the collection station side into the inertial response model and the primary frequency regulation model, and obtain the equivalent inertial constant and equivalent droop coefficient of the new energy base as a whole through the online parameter identification algorithm, so as to obtain the rapid frequency support capability of the new energy base in the early stage and middle and late stage of the disturbance. Step (5): During the disturbance period, based on the relationship between complex power injection and grid admittance matrix, sensitivity analysis is performed on the voltage amplitude and voltage phase angle changes of the substation bus, and the partial derivatives of active and reactive power increments with respect to voltage phasor increments are obtained, thereby obtaining the external equivalent reactance of the new energy base, which is used to characterize the electrical coupling strength between the new energy base and the upper-level grid. Step (6): The PMU time series data is subjected to weighted least squares fitting and smoothing. The frequency trajectory and active power trajectory are denoised and outlier removed. The smoothed frequency deviation, frequency change rate and active power change are used as the iterative update inputs for inertia constant, droop coefficient and equivalent reactance. Step (6): Based on the updated equivalent inertia constant, equivalent droop coefficient and equivalent reactance, calculate the maximum available fast frequency support capacity, sustainable support duration and characteristic indicators of the frequency response process of the new energy base under a given disturbance condition, and use the results as the monitoring output of the fast frequency support capability of the new energy base.

2. The method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from a data collection station as described in claim 1, characterized in that, Steps (1) and (2) specifically include: The phasor measurement unit synchronously collects phasor information of the three-phase voltage and current of the busbar at the collection station with a high sampling rate, and calculates the system frequency and frequency change rate in real time, while recording active power, reactive power and voltage phase angle. The absolute value of the frequency change rate and the active power change rate are compared with preset thresholds. When either quantity exceeds the threshold, a disturbance is determined to have occurred. The sampling time when the threshold is first exceeded is taken as the disturbance start time. Time windows containing several sampling points are extracted before and after the threshold as disturbance analysis intervals. By continuously performing threshold detection on subsequent monitoring data through a sliding time window, the system can automatically identify and segment multiple disturbance events.

3. The method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from a data collection station as described in claim 1, characterized in that, The primary frequency regulation response model and rapid support model of the wind turbine and photovoltaic power generation unit in step (3) include: Establish a linear droop relationship between the active power output of wind power and photovoltaic power and the system frequency deviation. Keep the output power constant within the preset primary frequency regulation dead zone, and adjust the output power proportionally according to the droop coefficient outside the dead zone. When the frequency deviation exceeds the rapid support trigger threshold, the rapid power boost function of wind power or photovoltaic is triggered. Under the premise of not exceeding the unit's available power margin and technical limits, the output active power is increased or decreased in a short period of time to support the rapid frequency recovery. The rapid power boosting process uses a piecewise linear relationship to describe the underfrequency support section and the overfrequency reduction section, and sets upper and lower limits to ensure the safe operation of the unit.

4. The method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from a data collection station as described in claim 1, characterized in that, The frequency response model of the electrochemical energy storage device in step (3) includes: In the inertial support stage, the change in active power of the energy storage device is established in a proportional relationship with the frequency change rate. When the frequency change rate is large, the power is released or absorbed quickly to simulate the suppression effect of the rotor inertia of the synchronous generator on the frequency change. During the primary frequency regulation phase, a linear droop relationship is established between the active power adjustment of the energy storage device and the frequency deviation. There is no response within the primary frequency regulation dead zone, and the power is gradually adjusted according to the set droop coefficient outside the dead zone. Based on the state of charge of the energy storage device, the available support capacity and maximum support duration are limited. When the state of charge is below the lower limit or above the upper limit, the inertial and primary frequency regulation support is automatically reduced or withdrawn to ensure the safe operation of the energy storage.

5. The method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from a data collection station as described in claim 1, characterized in that, The process of obtaining the equivalent inertia constant and equivalent droop coefficient of the new energy base in step (4) includes: The new energy base is equivalent to a virtual synchronous generator. The input of the virtual machine group is the frequency deviation, frequency change rate and active power change measured at the collection station, and the output is the equivalent inertial constant and equivalent droop coefficient. Within the disturbance analysis time window, based on the dynamic equilibrium relationship of the virtual synchronous machine, an algebraic equation is constructed between the equivalent inertial constant and the change in active power and the rate of change in frequency, and an algebraic equation is constructed between the equivalent droop coefficient and the frequency deviation and the change in active power. By employing recursive least squares or similar online identification algorithms, the equivalent inertia constant and equivalent droop coefficient are updated at each sampling time to achieve dynamic tracking of the frequency support capability of new energy bases.

6. The method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from a data collection station as described in claim 1, characterized in that, The process of obtaining the equivalent reactance of the new energy base in step (5) includes: Taking the substation busbar as the research object, an equivalent multi-busbar network model including the new energy base and its upstream power grid is established, and the voltage and injection current of each busbar are represented by the admittance matrix. Within the disturbance time window, the partial derivatives of the increments of active and reactive power of the collecting station bus with respect to the increments of bus voltage angle and voltage amplitude are calculated, and these are used as the sensitivity of power to voltage. Based on the aforementioned sensitivity, the equivalent reactance parameter is calculated inversely to characterize the impact of the new energy base on the bus voltage and power distribution during frequency support, thus serving as a quantitative indicator of the strength of the power grid and the electrical coupling characteristics.

7. The method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from a data collection station as described in claim 1, characterized in that, The process of performing weighted least squares fitting and smoothing on the PMU time series data in step (6) includes: Within each disturbance analysis time window, frequency, rate of change of frequency, and change of active power are regarded as fitting objects with respect to time, and linear or piecewise linear fitting functions are established. Weights are assigned to each sampling point within the window, with data points closer to the disturbance center having larger weights and data points farther from the disturbance center having smaller weights, in order to highlight the fitting accuracy for key response stages; a saturated weight function is introduced during the fitting process to automatically reduce the weights or remove outlier data with large residuals, thereby reducing the impact of measurement noise and instantaneous shocks on the parameter estimation results; The fitted smooth frequency trajectory, frequency change rate trajectory, and active power change trajectory are used as input data for subsequent iterative updates of the equivalent inertia constant, equivalent droop coefficient, and equivalent reactance.

8. The method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from a data collection station as described in claim 1, characterized in that, The iterative update of the inertial constant, droop coefficient, and equivalent reactance in step (6) adopts the following mechanism: Construct a residual vector consisting of frequency deviation, frequency change rate, and active power change, and construct a Jacobian matrix from the partial derivatives of the residual vector with respect to each parameter to be identified. Based on Gauss-Newton or Levenberg-Marquardt iterative algorithms, the equivalent inertial constant, equivalent droop coefficient, and equivalent reactance are jointly iteratively corrected using residual vectors and Jacobian matrices. When the norm of the residual is less than the preset threshold or the number of iterations reaches the upper limit, the current iteration result is output as the frequency support capability parameter of the new energy base under this disturbance event.

9. The method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from a data collection station as described in claim 1, characterized in that, The monitoring results of the rapid frequency support capability of the new energy base output in step (7) include: The equivalent inertial constant reflects the ability of a new energy base to suppress the rate of frequency change in the early stages of disturbance. The equivalent droop coefficient reflects the sensitivity of the new energy base to active power regulation of frequency deviation during the primary frequency regulation stage. Equivalent reactance parameters reflect the electrical coupling strength between the new energy base and the upstream power grid and the degree of influence on voltage deviation; The maximum available fast frequency support capacity and sustainable support duration under a given typical disturbance amplitude are used to assess the ability of new energy bases to undertake inertial support and primary frequency regulation support obligations. Dynamic performance indicators such as the expected minimum frequency and recovery time under corresponding disturbance conditions are used to provide early warning and decision-making basis for dispatching agencies and operators.

10. The method for monitoring the rapid frequency support capability of new energy bases based on monitoring data from a data collection station according to claim 1, characterized in that, This method is used by dispatch control centers or regional power grid control centers to conduct online monitoring of large-scale wind, solar, and energy storage bases connected to the grid, and has the following application modes: The monitoring process is automatically invoked under different types of disturbance scenarios, including continuous load change scenarios, single large unit disconnection scenarios, and multiple unit disconnection scenarios, and the rapid frequency support capability of the new energy base under each scenario is classified and evaluated. The equivalent inertial constant, equivalent droop coefficient, and equivalent reactance time series obtained from each disturbance event are statistically analyzed to form a long-term evaluation curve of the frequency support capability of the new energy base, providing a basis for the transformation plan, control strategy optimization, and auxiliary service assessment of the new energy base. When the monitoring results are below the preset threshold, an early warning is issued to the dispatcher, prompting the need to adjust the output of new energy sources, optimize energy storage control strategies, or activate other frequency support resources, thereby improving the frequency security margin of the power system under the high penetration of new energy sources.

Images