New energy grid-connected system nonlinear frequency domain coupling impedance model modeling method and device

Abstract

The invention relates to the technical field of new energy power systems, in particular to a modeling method and device for a nonlinear frequency domain coupling impedance model of a new energy grid-connected system. The method comprises the steps of establishing an impedance model of the single-machine converter, analyzing a frequency sweeping result of the single-machine converter obtained in the impedance model establishing process according to the multi-time scale characteristic of a controller, obtaining an impedance characteristic analysis result of each frequency band, analyzing corresponding parameters of the impedance model under different frequency bands according to the impedance characteristic analysis result of each frequency band, and calculating the impedance model of the single-machine converter according to the parameters. According to the obtained impedance transfer function and parameters of each frequency band, establishing an output impedance mathematical model of the inverter under all working conditions; according to the output impedance mathematical model, the feeder equivalent impedance of the feeder from the tail end to the head end is determined, and the overall output impedance of the multiple parallel feeder lines at the collection cabinet is calculated. According to the method, the problems that a multi-dimensional coupling mechanism is not clear and the equivalent modeling precision needs to be improved in the process of expanding the single machine characteristics to the whole station level in the prior art can be solved.

Description

Technical Field

[0001] This invention relates to the field of new energy power electronic equipment technology, and in particular to a method and apparatus for modeling a nonlinear frequency domain coupling impedance model of a new energy grid-connected system. Background Technology

[0002] With the rapid development of the new energy power generation industry, its grid-connected scale continues to expand, playing an increasingly important role in the energy structure transformation. However, this trend has also brought many challenges to power system stability: on the one hand, the support capacity of the receiving-end AC grid for the power system is gradually weakening, and power electronic power supplies are prone to stability problems such as wideband oscillations and overvoltages; on the other hand, power electronic power supplies need to be controlled by inverters or converters to achieve grid connection. The current loop and phase-locked loop introduced to keep synchronized with the power system may cause high-frequency resonance in the rapid control of the inverter or converter. After coupling with the inductive impedance of the traditional power grid, this high-frequency resonance is very likely to induce harmonic oscillations or subsynchronous oscillations. In order to evaluate the stability of new energy power plants connected to the AC grid, frequency sweep analysis of the grid-connected system of new energy power plants and verification of whether the output characteristics of new energy power plants and grid impedance meet the stability criteria have become necessary measures. Currently, in the field of grid-connected stability analysis of renewable energy power plants, station-level stability analysis requires the construction of a cross-scale analysis system encompassing "single-unit equipment - cluster system". Existing research mainly focuses on renewable energy power generation units, such as the sweep frequency impedance characteristics of converters. However, in the process of extending single-unit characteristics to the entire station level, the multi-dimensional coupling mechanism remains unclear, and the accuracy of equivalent modeling needs to be improved. This significantly restricts the accurate assessment of the stability of renewable energy power plants connected to the AC power grid. Summary of the Invention

[0003] This invention provides a method and apparatus for modeling a nonlinear frequency domain coupling impedance model for a new energy grid-connected system, in order to solve the problem that the existing technology cannot accurately assess the stability of new energy power plants connected to the AC power grid.

[0004] In a first aspect, embodiments of the present invention provide a method and apparatus for modeling a nonlinear frequency domain coupling impedance model of a new energy grid-connected system, comprising: An impedance model of a single-unit converter is established, and based on the multi-time-scale characteristics of the controller, the frequency sweep results of the single-unit converter obtained during the establishment of the impedance model are analyzed to obtain the impedance characteristic analysis results for each frequency band. Based on the impedance characteristic analysis results of each frequency band, the corresponding parameters of the impedance model under different frequency bands are analyzed, and a mathematical model of the inverter's output impedance under all operating conditions is established based on the obtained impedance transfer function and parameters of each frequency band. Based on the output impedance mathematical model, combined with the station's collector line parameters, inverter connection type, and station topology, the feeder equivalent impedance from the end to the beginning of the feeder is determined. Based on the feeder equivalent impedance, the overall output impedance of the new energy station level with multiple feeders connected in parallel at the collection cabinet is calculated.

[0005] Secondly, embodiments of the present invention provide a modeling device for a nonlinear frequency domain coupling impedance model of a new energy grid-connected system, comprising: The modeling module is used to build the impedance model of a single-unit converter. The analysis module is used to analyze the frequency sweep results of the single-unit converter obtained during the establishment of the impedance model based on the multi-time-scale characteristics of the controller, and to obtain the impedance characteristic analysis results of each frequency band. The analysis module is also used to analyze the corresponding parameters of the impedance model under different frequency bands based on the impedance characteristic analysis results of each frequency band, and obtain the impedance transfer function and parameters of each frequency band. The modeling module is also used to establish a mathematical model of the inverter's output impedance under all operating conditions based on the obtained impedance transfer functions and parameters for each frequency band. The calculation module is used to determine the equivalent impedance of the feeder from the end to the beginning based on the output impedance mathematical model, combined with the parameters of the power collection line of the power station, the connection form of the inverter and the topology of the power station, and to calculate the overall output impedance of the new energy power station with multiple feeders connected in parallel at the collection cabinet based on the equivalent impedance of the feeder.

[0006] This invention provides a method and apparatus for modeling a nonlinear frequency domain coupling impedance model for a new energy grid-connected system. By establishing an impedance model for a single-unit converter and analyzing the frequency sweep results of the single-unit converter obtained during the impedance model establishment process based on the multi-time-scale characteristics of the controller, impedance characteristic analysis results for each frequency band are obtained. Based on the impedance characteristic analysis results for each frequency band, the corresponding parameters of the impedance model under different frequency bands are analyzed. Based on the obtained impedance transfer function and parameters for each frequency band, a mathematical model of the inverter's output impedance under all operating conditions is established. Based on the output impedance mathematical model, combined with the parameters of the power collection line at the power station, the connection form of the inverter, and the power station topology, the equivalent impedance of the feeder from the end to the beginning is determined. Based on the equivalent impedance of the feeder, the overall output impedance of the new energy power station level with multiple feeders connected in parallel at the collection cabinet is calculated. This invention establishes a three-level impedance model of "equipment-cluster-site," organically combining single-unit converter modeling, multi-unit parallel equivalent impedance, and site network impedance distribution. This achieves a comprehensive stability analysis from the component level to the system level. Furthermore, based on the multi-timescale characteristics of the controller, the target frequency band is divided into different frequency bands and modeled separately, avoiding coupling interference between different control links at similar frequencies. This significantly improves the impedance modeling accuracy over a wide frequency range, thus solving the problem in existing technologies where the multi-dimensional coupling mechanism is unclear and the equivalent modeling accuracy needs improvement when extending single-unit characteristics to the entire site level. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a flowchart illustrating the implementation of the nonlinear frequency domain coupling impedance modeling method for new energy grid-connected systems provided in this embodiment of the invention. Figure 2 This is a schematic diagram of the modeling process provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the converter frequency sweep circuit topology provided in an embodiment of the present invention; Figure 4 This is a swept-frequency Bode plot (frequency-amplitude Bode plot) of an inverter under a single operating condition provided in an embodiment of the present invention. Figure 5 This is a swept-frequency Bode plot (frequency-phase Bode plot) of an inverter under a single operating condition provided in an embodiment of the present invention. Figure 6 This is a block diagram of a single-machine controller grid-connected system and control block provided in an embodiment of the present invention; Figure 7 (1) is a schematic diagram of the amplitude of the inverter impedance amplitude-frequency characteristic under all operating conditions when the inverter power factor is 1.0, provided in the embodiment of the present invention; Figure 7 (2) is a phase diagram of the inverter impedance amplitude-frequency characteristic under all operating conditions when the inverter power factor is 1.0, provided in the embodiment of the present invention; Figure 7 (3) is a schematic diagram of the amplitude of the inverter impedance amplitude-frequency characteristic under the full operating condition of the inverter corresponding to the active power under the active power provided in the embodiment of the present invention; Figure 7 (4) is a phase diagram of the inverter impedance amplitude-frequency characteristic under the full operating condition of the inverter corresponding to the active power under the maximum capacitive reactive power provided in the embodiment of the present invention; Figure 7 (5) is a schematic diagram of the amplitude of the inverter impedance amplitude-frequency characteristic under the full operating condition of the inverter corresponding to the active power under the active power provided in the embodiment of the present invention. Figure 7 (6) is a phase diagram of the inverter impedance amplitude-frequency characteristic under the full operating condition of the inverter corresponding to the active power under the active power provided in the embodiment of the present invention; Figure 8 This is a schematic diagram of the inverter output structure provided in an embodiment of the present invention; Figure 9 This is a schematic diagram of the output impedance model of the inverter provided in an embodiment of the present invention; Figure 10 This is a schematic diagram of the aggregation of the station-level impedance model provided in the embodiment of the present invention; Figure 11 This is a schematic diagram of the process for determining the equivalent impedance of a feeder from its end to its beginning, provided by an embodiment of the present invention. Figure 12 This is a schematic diagram showing the equivalent aggregation of inverter output impedance and line series impedance provided in an embodiment of the present invention; Figure 13 This is a schematic diagram of a modeling device for a nonlinear frequency domain coupling impedance model of a new energy grid-connected system provided in an embodiment of the present invention. Detailed Implementation

[0009] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0010] To assess the stability of new energy power plants connected to the AC power grid, frequency sweep analysis of the grid-connected system of the new energy power plants is required to verify that the output characteristics of the new energy power plants and the grid impedance meet the stability criteria.

[0011] Station-level stability analysis requires the construction of a cross-scale analysis framework encompassing "single-unit equipment to cluster system". Current technologies primarily focus on characterizing the swept-frequency impedance characteristics of new energy power generation units (such as converters). However, when extending single-unit characteristics to the entire station level, significant technical bottlenecks remain, including unclear multi-dimensional coupling mechanisms and insufficient accuracy in equivalent modeling. Specifically, key issues such as harmonic interactions caused by multi-unit parallel coupling, the spatiotemporal coupling relationship between station network impedance distribution characteristics and equipment control strategies, and the dynamic response superposition mechanism under complex operating conditions, lack a standardized solution framework.

[0012] To address the unclear multi-dimensional coupling mechanism when extending single-unit characteristics to the entire site, this invention constructs a three-level impedance coupling model of "equipment-cluster-site," considering the inverter impedance characteristics under multiple operating conditions and the influence of frequency band decoupling control links, thus achieving more accurate impedance modeling. To make the objectives, technical solutions, and advantages of this invention clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0013] Figure 1 The implementation flowchart of a nonlinear frequency domain coupling impedance modeling method for a new energy grid-connected system provided in this embodiment of the invention is detailed below: Step 101: Establish the impedance model of the single-unit converter, and analyze the frequency sweep results of the single-unit converter obtained in the process of establishing the impedance model based on the multi-time scale characteristics of the controller, so as to obtain the impedance characteristic analysis results of each frequency band.

[0014] Impedance modeling of a single-unit converter is a fundamental step in achieving station-level stability analysis. Common methods for single-unit converter impedance modeling fall into two main categories: analytical modeling and experimental modeling, and numerical modeling.

[0015] The steps of the analytical modeling method are as follows: First, write the voltage and current equations, filter equations, and control equations of the converter in the abc coordinate system; then... Alternatively, a dq transformation can be used to transform the abc coordinate system to a synchronous rotating coordinate system; the active power, reactive power, DC voltage, and grid voltage measured at the steady-state operating point can be linearized using small signals; finally, the output value can be obtained using a multi-input multi-output state-space equation.

[0016] When considering the paths for output current and voltage, the output impedance (or input impedance) matrix can be defined: ;in, and They are respectively in dq Voltage small-signal quantities and current small-signal quantities in the coordinate system.

[0017] The steps of experimentation and numerical modeling are as follows: The hardware-in-the-loop simulation frequency sweep test platform is built, and the modeling process is as follows: Figure 2 As shown.

[0018] (1) Only a single frequency disturbance signal is injected into the converter each time. or Single-frequency scanning injection, traversing the target frequency band sequentially according to a preset step size, measuring the system response at each frequency point, and acquiring time-domain data of voltage or current. See also Figure 3 The diagram shows the topology of the frequency sweep circuit for the converter. Figure 3 The image only shows a portion of the converter frequency sweep circuit topology, where module 1 is the disturbance signal. Input location: Module 2 is the impedance voltage or current measurement point.

[0019] The signal injected here is a pure sine wave: , ,in The current scanning frequency, The amplitude is typically 1% to 5% of the nominal voltage or current. The target frequency band can be set as needed; in this embodiment, the target frequency band can be set from 2 Hz to 1 kHz.

[0020] (2) Improve signal quality by using bandpass filters or removing DC components.

[0021] (3) Calculate And extract the amplitude and phase.

[0022] Optional, let These represent the time-domain sequences of the sampled voltage and current, respectively. Considering the injection of a single-frequency sine wave, in steady state, the frequency of these two signals... The component at point can be approximately extracted using the following formula: , ; in It is a discrete spectrum in the index The complex value at that point, i.e., the corresponding frequency. phasors on; Represents the imaginary unit; exponent term It is the basis function used to extract this frequency component.

[0023] In frequency At that point, the complex impedance is defined as: ; The above formula is obtained from It is a plural number, and can be written as: , ; The calculated amplitude is: ; ; The phase is calculated as follows: ; ; (4) Calculate the impedance.

[0024] In frequency At that point, the complex impedance is defined as: ; The impedance amplitude is then: ; The impedance phase angle is: ; (5) Change the injection frequency By repeating steps (1) to (4) above, the impedance amplitude-frequency characteristics and phase-frequency characteristics of the entire target frequency range can be obtained.

[0025] (6) By repeating this process at different frequency points, an impedance Bode plot can be drawn, as follows: Figure 4 and Figure 5 The Bode plot of the inverter under a single operating condition is shown.

[0026] After obtaining the Bode plot of the inverter, the inverter can be analyzed by sweeping the frequency based on the Bode plot in order to obtain the impedance characteristics of each frequency band.

[0027] In one embodiment, based on the multi-time-scale characteristics of the controller, the frequency sweep results of the single-unit converter obtained during the impedance model establishment process are analyzed to obtain impedance characteristic analysis results for each frequency band, which may include: Obtain the frequency sweep results of the single-unit converter obtained during the impedance model establishment process; The frequency bands in the frequency sweep results are divided into low-mid frequency, mid frequency, and mid-high frequency. Based on the multi-timescale characteristics of the controller, the frequency sweep results of the inverter under a single operating condition are analyzed based on different frequency bands to obtain the correlation between the impedance characteristics of each frequency band and the control loop and power operating conditions.

[0028] In this embodiment, see Figure 6The diagram shown illustrates a grid-connected system with a single controller. We consider the single controller's inner current loop control, phase-locked loop (PLL), P / Q loop with pre-filter, and DC voltage outer loop control. Based on the controller's multi-timescale characteristics, the target frequency band is divided into low-to-mid frequency, mid-frequency, and mid-to-high frequency. Low-to-mid frequency can be a few Hz, mid-frequency is a few Hz to tens of Hz, and mid-to-high frequency is tens of Hz to hundreds of Hz. Their corresponding dynamic characteristics are shown in Table 1. As shown in Table 1, the low-frequency model is mainly dominated by the outer loop control, requiring consideration of DC voltage and power control. The mid-frequency band is mainly determined by the dynamics of the PLL, while the high-frequency band is determined by the rapid dynamics of the inner current loop and filters. The bandwidths of different control loops need to be clearly separated to avoid instability caused by mutual coupling at close frequencies.

[0029] Table 1

[0030] Based on the dynamic characteristics of the inverter described above, the amplitude-frequency characteristic curve of the inverter under single operating condition impedance sweep is analyzed, and the analysis results are shown in Table 2 below.

[0031] Table 2

[0032] When performing full-condition amplitude-frequency characteristic curve analysis on inverter impedance sweep frequency, 20 operating conditions can be swept based on different inverter power factors. See Table 3 for the inverter impedance sweep operating conditions. Figure 7 The inverter impedance amplitude-frequency characteristics at different power factors are shown in (1)-7 (6). Analysis Figure 7 (1)-7(6) can yield the analysis results shown in Table 4.

[0033] Table 3

[0034] Table 4

[0035] Step 102: Based on the impedance characteristic analysis results of each frequency band, analyze the corresponding parameters of the impedance model under different frequency bands, and establish the mathematical model of the inverter's output impedance under all operating conditions based on the obtained impedance transfer function and parameters of each frequency band.

[0036] When analyzing the corresponding parameters of the impedance model at different frequency bands, refer to Figure 8 The inverter output structure is shown.

[0037] In one embodiment, based on the impedance characteristic analysis results of each frequency band, the corresponding parameters of the impedance model under different frequency bands are analyzed, and based on the obtained impedance transfer function and parameters of each frequency band, a mathematical model of the inverter's output impedance under all operating conditions is established, including: Based on the impedance characteristic analysis results of each frequency band, the corresponding parameters of the impedance model under different frequency bands are analyzed to obtain the impedance transfer function and parameters of each frequency band; based on the impedance transfer function and parameters of each frequency band, a mathematical model of the inverter's output impedance under all operating conditions is established.

[0038] Assuming the bandwidths of the outer loop, PLL, and inner loop are separate, and the low-to-mid frequency, mid-frequency, and mid-to-high frequency bands each occupy different frequency bands, then only the influence of the corresponding sub-model needs to be considered within each frequency band. At the frequency band boundaries, the contributions of each sub-model are equivalent. Therefore, in this embodiment, we divide the bandwidth, determine the corresponding impedance transfer function based on the output impedance characteristics of the inverter in each frequency band, and then combine the bandwidths.

[0039] Optionally, in the low-frequency band, the inverter's output impedance characteristics are primarily dominated by the outer-loop control, mainly including slow control loops such as DC bus voltage control and active or reactive power control. These outer-loop controls typically have low bandwidth, determining the inverter's small-signal power support characteristics within this frequency band. Therefore, in the low-frequency band, the inverter's equivalent output impedance often exhibits a first-order dynamic characteristic with negative impedance. Thus, the equivalent output impedance in the low-frequency band can be approximated by a first-order system, and the low-frequency impedance transfer function is: ;in, This indicates the impedance of a low-frequency inverter. This indicates the first equivalent negative impedance presented by the converter in the low-frequency range. Represents the Laplace variable. This represents the time constant related to the DC-side dynamics or outer loop control of the inverter. molecular The numerator is a constant, and the numerator is negative to reflect the negative impedance characteristic; the denominator is a constant. It is a first-order polynomial dependent on the outer loop time constant. This transfer function in DC (… The impedance at point () is approximately It exhibits a corner point near the outer ring bandwidth frequency.

[0040] Optionally, in the mid-frequency band, the converter output impedance is primarily governed by the dynamics of the phase-locked loop (PLL). For grid-connected inverters based on PLL control, changes in the phase or frequency of the grid voltage under small disturbances will couple to the inverter output current through the PLL, significantly altering the equivalent impedance characteristics. The PLL is essentially a second-order closed-loop system, typically containing a PI regulator. Its dynamics often introduce conjugate pole pairs in the frequency domain, affecting the inverter's impedance amplitude and phase characteristics in that frequency band. This manifests as follows: within the PLL bandwidth, the inverter output impedance may exhibit capacitive and negative impedance characteristics, i.e., the impedance phase is close to -90° and the real part is negative. Therefore, the mid-frequency equivalent impedance can be approximated by a second-order system with relatively low damping. The mid-frequency impedance transfer function is: ;in, This indicates the impedance of a mid-frequency inverter. This represents the second equivalent negative impedance introduced by phase-locked loop control in the mid-frequency band. Reflecting the magnitude of negative impedance, This refers to the zero-point frequency introduced by PLL control. For example, the integral action of PLL PI control introduces a zero-pole pair at the origin. This indicates that the PLL damping ratio may be small, resulting in capacitive phase characteristics. This represents the natural frequency of the phase-locked loop (PLL), which is approximately proportional to the PLL bandwidth.

[0041] At high frequencies, the inverter output impedance is primarily determined by the rapid dynamics of the inner current loop and the output filter. Inverters typically employ an LCL-type output filter, comprising an inverter-side inductor, a parallel capacitor, and a grid-side inductor. The inner current loop performs high-speed control of the inverter-side inductor current, with a bandwidth typically in the hundreds of hertz range, significantly reducing the inverter output impedance within this bandwidth and suppressing filter resonance. Therefore, the impedance characteristics at high frequencies exhibit a high-order dynamic dominated by the physical components of the LCL filter: the filter has an inherent resonant frequency and a phase roll-off introduced by control delay. Thus, the equivalent output impedance at high frequencies can be approximated as a second-order or higher system, including the filter's resonant poles and damping. The high-frequency impedance transfer function is: ;in, This indicates the impedance of a high-frequency inverter. Indicates the filter inductance. Indicates equivalent inductive damping, It can be viewed as a series inductive reactance model approximated at high frequencies. Indicates the resonant angular frequency of the filter. This represents the filter's resonant damping factor. The denominator is... Indicates at the resonant frequency There exists a pair of conjugate poles at a certain point, but due to the current control effect, these conjugate poles have a certain damping. This damping is the reciprocal of the quality factor.

[0042] In summary, by using frequency band splicing, the influence of different control links on impedance can be decoupled while ensuring accuracy.

[0043] When combining bandwidths, the low-frequency model can be connected to the mid-frequency model near the PLL bandwidth to ensure that the impedance amplitudes calculated by the two models are in phase at that frequency. This allows for a smooth transition between the mid-frequency model and the high-frequency model near the inner loop bandwidth. Through this progressive splicing method, the equivalent transfer functions of the three frequency bands are merged into a complete frequency domain model.

[0044] The following verifies this frequency domain model by performing nonlinear fitting on the measured impedance of the inverter under each operating condition from 10% to 90%, and extracting the corresponding parameters. , , , , , , , , , .

[0045] The amplitude can be estimated from the real part of the impedance in the low-frequency range. Because... It should be noted that the effects of PLLs and filters are approximately constant at extremely low frequencies. If the impedance is close to pure resistance and exhibits negative impedance characteristics at the lowest frequency, then the first-order ( (Low frequency) can be used as Initial value. For example, if the impedance amplitude in the low-frequency range is approximately a few ohms at 10% power, then... This value can be used as the initial value; under different working conditions It may increase with increasing power.

[0046] This determines the inflection point frequency of the first-order roll-off at low frequencies. It can be observed at the frequency at which the measured amplitude changes from flat to decreasing, for example, the outer loop control bandwidth can range from a few Hz to tens of Hz. The initial value can be... ,in Let be the estimated inflection point frequency. For example, if the frequency decreases at approximately 10 Hz, then take . 0.016 seconds.

[0047] The depth of negative impedance in the mid-frequency range is affected. It can be estimated based on the amplitude of impedance changes in the PLL's operating frequency band. For example, if a transition or amplitude trough occurs in the real part of the impedance at tens of Hz under a certain operating condition, it can be inferred that... This is close to the negative impedance compensation provided by this frequency band. The initial value can be taken as... Values ​​of the same order of magnitude, or estimates based on the depth of the impedance valley.

[0048] The natural frequency of a PLL is usually close to its bandwidth and can be inferred from the frequencies of significant changes in impedance phase or amplitude. For example, if the phase changes rapidly or shows peaks or valleys around a certain frequency, it is likely close to the natural frequency of the PLL. Get the initial value. .

[0049] The PLL damping ratio affects the peak value of the mid-frequency resonance. If in The presence of significant peaks or troughs in the nearby impedance amplitude indicates low damping. (Small); if the response is smooth, the damping ratio will be larger. An initial value of 0.5~0.7 can be used as a general damping ratio; otherwise, a rough estimate can be made based on the peak amplitude. For example, if the peak amplitude is greater than 3dB, the estimated value should be... Less than 0.5.

[0050] The zero-frequency introduced by PLL control. Typically, the integral action of a PI control introduces a pole at the origin and a zero at a higher frequency for phase correction. If the measured phase is higher than... The appearance of a secondary inflection point or a flattening of the amplitude curve may be due to a zero-point effect. Initial values ​​can be chosen... .

[0051] The impedance of a filter inductor can be estimated from the impedance slope in the high-frequency range. In the high-frequency region, such as above the resonant frequency, the impedance amplitude increases with frequency as an inductive reactance. Based on the impedance amplitude slope in the high-frequency range... Estimate Alternatively, the initial value can be based on the hardware parameter ratings provided by the manufacturer.

[0052] : Filter resonant angular frequency. Estimated based on the location of the impedance amplitude peak at high frequencies. For example, if a resonant peak occurs at approximately 300 Hz under certain operating conditions, then... .

[0053] The filter's resonant damping factor determines the sharpness of the resonance peak. It is estimated from the amplitude and phase characteristics of the resonance peak. A higher peak value indicates a more dramatic phase transition, resulting in lower damping and a higher quality factor. Initial values ​​can be estimated using the quality factor; for example, the peak amplitude is approximately equal to... ,but If the peak value is 20 dB, then Approximately 0.1; if there is no obvious peak value, then a larger damping value should be used, for example... =1.

[0054] The impedance modeling results for a certain type of inverter are shown in Table 5. Table 5

[0055] It can approximately reproduce the Bode plot characteristics of the original inverter in the range of 2 Hz to 1 kHz, including the slope changes of each segment of the amplitude curve and the phase transitions.

[0056] Optionally, the full-condition output impedance mathematical model adopts a fixed structural form to express the impedance characteristics of the converter in each frequency band, and introduces a power-related modulation coefficient to reflect the change in impedance amplitude under different output power. Based on the "three-segment" frequency domain characteristics of the typical converter output impedance, we divide the model into three parts: the equivalent negative impedance in the low-frequency band, the PLL-dominated negative impedance in the mid-frequency band, and the filter inductive impedance in the high-frequency band. Each part has a clear physical meaning and is represented by a parameterized transfer function, which are then superimposed to obtain the overall output impedance model.

[0057] In one embodiment, based on the impedance transfer function and parameters of each frequency band, a mathematical model of the inverter's output impedance under all operating conditions is established, which may include: The impedance model structure for each frequency band is determined based on the impedance transfer function of each frequency band. Determine the power modulation coefficient based on the parameters; Based on the frequency band impedance model structure and power modulation coefficient, the mathematical model of the inverter's output impedance under all operating conditions is obtained.

[0058] Optionally, when determining the impedance model structure for each frequency band based on the impedance transfer function of each band, the impedance parameter table measured under various operating conditions of the inverter is first used to construct... and The functional relationship between output power and output power. Except and Other impedance parameters, such as filter inductance, remain constant under different power levels.

[0059] Assuming within the power variation range and The power changes approximately linearly with the power, and their functional relationship is established using linear interpolation.

[0060] ,here This represents the equivalent impedance value under 100% rated power output conditions. As a percentage of power, in At that time, corresponding to 10% power, ,exist At that time, corresponding to 90% power, .

[0061] Based on For measurement data points between 0.9 and 0.9, according to Table 5, they can be approximately considered as... and about The relationship is linear and passes through the origin. When the output power approaches zero, the negative impedance effect of the inverter output impedance tends to disappear. Therefore, we can write: .

[0062] Similarly, , The equivalent impedance parameter caused by the PLL at rated power is given by the power rating. Scale proportionally. .

[0063] Refer to Table 5 for calculation based on the interpolation algorithm. , , , The specific algorithm is as follows: Enter the first column of Table 5 The second column , fourth column Then sort the data points by the values ​​in the first column, and use the formula... and This generates the Lagrange interpolation formula. Input the first column value to be estimated, perform the interpolation calculation, and return the result, thus obtaining... , , , Substitute this into the original formula to obtain the power variable.

[0064] Optionally, the same model with power-related modulation coefficients can be introduced, so that the model can be characterized by only the different coefficient values ​​under each operating condition.

[0065] Define the normalization coefficients respectively and Characterizing the power effect. Therefore, the above formula can be written as: ; ; but , .

[0066] For linear proportional relationships, the above coefficients and Approximately equal to ,Right now . and Is it following A monotonically increasing function, in When the value is equal to 1, the power dependence can be separated from the structure of the impedance expression using these two coefficients, serving as a modulation factor.

[0067] In summary, the sum of the impedances across all frequency bands constitutes the overall model of the output impedance at the grid-connected inverter port; that is, the mathematical model of the output impedance is: ;in, For the mathematical model of output impedance, see [link to relevant documentation]. Figure 9 The output impedance model of the inverter is shown.

[0068] Step 103: Based on the mathematical model of output impedance, combined with the parameters of the power collection line of the power station, the connection form of the inverter and the topology of the power station, determine the equivalent impedance of the feeder from the end to the beginning, and calculate the overall output impedance of the new energy power station with multiple feeders connected in parallel at the collection cabinet based on the equivalent impedance of the feeder.

[0069] See Figure 10 The diagram shown is a clustered impedance model of the site-level impedance, i.e., the series impedance model of the lines from the inverter to the collection node. It can be seen that each inverter is connected to the collection cabinet node through a collection line of a certain length, which is... The (km) line can be equivalent to a series impedance model. Assuming the collector cabinet is an ideal node with an impedance approximately of 0, it collects the outputs of all inverters after passing through the line. Since the total output impedance of each inverter acts in parallel at the common collector bus, they can be equivalent to a single output impedance.

[0070] In one embodiment, see Figure 11 As shown, based on the output impedance mathematical model, combined with the station collector line parameters, inverter connection type, and station topology, determining the feeder equivalent impedance from the end to the beginning can include the following steps: Step 111: Based on the mathematical model of output impedance and the site topology, determine the total output impedance of the current inverter based on the site collection line parameters between the current inverter access collection cabinet nodes.

[0071] Table 6 shows a schematic diagram of the power generation unit topology within a photovoltaic power station. The length is... A line of (kilometers) can be equivalent to a series impedance model: ,in, Indicates the impedance of the station's collector lines. This indicates the resistance of the station's collector lines. Represents the imaginary unit. Represents angular frequency. This indicates the inductance of the station's collector lines. Table Six

[0072] See Figure 12The diagram shown is an equivalent aggregation of the inverter output impedance and the line series impedance. Here, we assume that all inverters have the same output impedance characteristics. Therefore, the total output impedance of the current inverter is the sum of the current inverter's output impedance and the corresponding line impedance in series.

[0073] Optionally, the output impedance of the inverter can be adopted from the mathematical model of the inverter's output impedance under all operating conditions obtained in step 102. After fitting, use This indicates that, based on the mathematical model of output impedance and the site topology, and considering the site collector line parameters between the current inverter access collector nodes, the total output impedance of the current inverter is determined, which may include: according to Determine the total output impedance of the current inverter; this indicates that the total output impedance frequency response of each inverter is composed of its own control characteristics superimposed with a fixed line impedance.

[0074] in, This represents the total output impedance of the current inverter. This represents the fitted value of the mathematical model of output impedance. Indicates the impedance of the station's collector lines. This indicates the resistance of the station's collector lines. Represents the imaginary unit. Represents angular frequency. This indicates the inductance of the station's collector lines.

[0075] Step 112: Based on the total output impedance of the current inverter, calculate the parallel equivalent impedance of the N grid-connected inverters connected in parallel within each box-type transformer.

[0076] Since N grid-connected inverters are connected in parallel within each box-type transformer, the output impedance of the box-type transformer can be obtained by calculating the parallel equivalent impedance of the N inverters after determining the total output impedance of each inverter.

[0077] In one embodiment, calculating the parallel equivalent impedance of N grid-connected inverters connected in parallel within each box-type transformer, based on the total output impedance of the current inverter, may include: Calculate the parallel equivalent impedance of N grid-connected inverters connected in parallel within each box-type transformer; in, Indicates the parallel equivalent impedance. Indicates the current inverter number. This indicates the total number of inverters in each box-type transformer.

[0078] Because the control strategies of each inverter are consistent with the line parameters, therefore This can be simplified to: This equivalent impedance is used to describe the impedance relationship between the low-voltage side busbar of the same box-type transformer and its internal inverter group, laying the foundation for subsequent impedance coupling with the box-type transformer step-up transformer.

[0079] Step 113: Calculate the output impedance of the box-type transformer on the medium-voltage side based on the parallel equivalent impedance and the leakage impedance of the box-type step-up transformer.

[0080] The equivalent impedance of the inverter group obtained in step 112 is connected in series with the leakage impedance of the box-type step-up transformer to obtain the output impedance of the box-type transformer on the medium voltage side.

[0081] Based on the parallel equivalent impedance and the leakage impedance of the box-type step-up transformer, calculate the output impedance of the box-type transformer on the medium-voltage side, including: Calculate the output impedance of the box-type transformer on the medium-voltage side; in, This indicates the output impedance of the transformer substation. This represents the leakage impedance of a box-type step-up transformer, and its value can be determined based on the transformer's equivalent impedance. The type parameters are determined by converting them to the intermediate pressure side; if a three-phase symmetrical model is used... Denote as a diagonal 3×3 matrix block : ; in, , , These represent the equivalent self-impedances of phases A, B, and C output from the box-type transformer, respectively. They include the sum of the equivalent impedance of the inverter group within the box-type transformer after being referred to the medium-voltage side and the transformer leakage reactance. Because a three-phase symmetrical model is used, the impedances of each phase are symmetrical and inter-phase coupling is negligible; therefore, the matrix is ​​diagonal, with off-diagonal elements equal to 0. Generally, diagonal elements represent the self-impedance of each phase, while non-zero off-diagonal elements represent mutual impedance coupling between different phases.

[0082] This step achieves impedance coupling between the equipment level (equivalent impedance of the inverter group) and the primary side electrical equipment (box-type step-up transformer), obtaining the comprehensive impedance characteristics at the output of the box-type transformer, laying the foundation for subsequent impedance modeling of the substation network.

[0083] Step 114: Based on the station topology and the output impedance of the box-type transformer, calculate the equivalent impedance of the feeder from the end box-type transformer upstream along the feeder until it reaches the beginning of the feeder.

[0084] Based on the station topology, the line impedance is calculated segment by segment from the end of the box-type transformer to the collection cabinet node, and the output impedance of each box-type transformer along the line is connected in parallel at the corresponding node. The equivalent impedance is calculated step by step towards the collection cabinet node.

[0085] Optionally, for adjacent nodes , The cable impedance is If node The equivalent impedance at the far end is Then by node To the node The branch impedance viewed from the direction of travel is the sum of the series impedances of each segment: ; At the node At this point, the output impedance of the transformer at this node will be... With slave node Downstream path impedance By connecting in parallel, we obtain the equivalent impedance of the nodes: ; in, This represents the equivalent impedance of the feeder. Indicates from adjacent nodes Towards Branch impedance viewed from the direction of travel. Indicates adjacent nodes , The cable impedance between them Represents a node Equivalent impedance at the far end.

[0086] The impedance of each section of the line and the transformer is gradually merged upstream from the end of the feeder until the equivalent impedance of the connection node between the feeder head and the busbar is obtained. .

[0087] This recursive algorithm can be solved in one step using a matrix form combined with the node-branch correlation matrix, or it can be automated by writing scripts in frequency domain simulation software.

[0088] The above yields the equivalent impedance of a feeder. The process is entirely based on the series and parallel circuit rules, belonging to the impedance aggregation method under linear systems, and does not consider nonlinear factors such as multi-machine coupling and frequency crossing. Therefore, in this embodiment, based on the calculation of the feeder's equivalent impedance, combined with the line equation and topological constraints, a physical information neural network is used for optimization to obtain a more accurate feeder equivalent impedance. Furthermore, when multiple feeders in the station are connected in parallel to the bus, the equivalent impedances of each feeder are further connected in parallel at the bus to obtain the overall output impedance of the entire station.

[0089] In one embodiment, calculating the overall output impedance of a new energy power station with multiple feeders connected in parallel at the collection cabinet based on the feeder equivalent impedance may include: The feeder equivalent impedance, line equations, and station topology constraints are integrated into Physical-Informed Neural Networks (PINNs). The PINNs are trained with frequency as input and equivalent complex impedance at the bus as output to obtain a nonlinear mapping model from frequency to impedance. The total loss function used in training the PINNs is a weighted sum of frequency domain data fitting loss, topology physical consistency loss, and output stability and regularization loss. An optimized frequency impedance coupling model is output using a nonlinear mapping model. Based on the frequency impedance coupling model, the overall output impedance of the new energy power station with multiple feeders connected in parallel at the collection cabinet is calculated.

[0090] Optionally, when training the physical information neural network, the network structure and computational process can partially reflect the connection relationships of the actual collector network, and circuit laws can be incorporated as constraints for the loss term during training. These circuit laws can include Kirchhoff's laws, component impedance formulas, etc.

[0091] The PINN model employs a modular design, corresponding to the topology of actual collector circuits. The input is frequency. The output is the equivalent complex impedance at the busbar. This equivalent complex impedance differs from the traditional impedance model because frequency is incorporated; therefore, it can be called a frequency-impedance coupling model. This indicates that, in order to handle complex numbers, the network output impedance can be configured with two channels: one for the real part and one for the imaginary part.

[0092] To train the PINN model described above, a total loss function with multiple components was designed to simultaneously minimize measurement error and physical constraints. The total loss function is a weighted sum of the frequency domain data fitting loss, the topological physical consistency loss, and the output stability and regularization loss.

[0093] The constrained model for fitting loss to frequency domain data outputs a fitted Bode curve based on the measured values. For each frequency point, the difference between the predicted impedance and the actual impedance is calculated. Mean square error is used to represent both the real and imaginary parts.

[0094] For topological physical consistency loss, although the model structure has embedded physical relationships, the model output may exhibit subtle physical inconsistencies in the early stages of training or under data noise. Therefore, additional constraints are used to further enhance topological physical consistency, including voltage and current conservation constraints, component parameter formula constraints, and output stability and regularization.

[0095] (1) Voltage and current conservation constraints: For each network node, the residual of its current balance equation is introduced as a loss.

[0096] (2) Component parameter formula constraints: The DC side voltage of the inverter is introduced as a constraint condition.

[0097] (3) Output stability and regularization: To avoid severe oscillations or undesirable peaks in the model, regularization constraints are added.

[0098] Physical rationality constraints are soft constraints imposed on known physical properties. The equivalent impedance of most power supply systems should be passive, meaning it has a non-negative real part and generates no energy. This embodiment penalizes cases where the real part is negative; for example, at high-frequency limits, the inverter's output impedance should not oscillate unbounded. These constraints can be transformed into losses using inequalities.

[0099] In summary, the total loss function is a weighted sum of the components, and the weights of each component can be set as needed.

[0100] After training the PINN model, the equivalent impedance of the feeder, the line equation, etc. are input into the trained PINN model, and the optimized frequency impedance coupling model is output.

[0101] In one embodiment, the aggregation cabinet node is considered as an ideal reference point, and the equivalent impedance of M feeders connected in parallel on it is... Overall output impedance at the station level The parallel formula is obtained by training PINN (Physical Information Neural Network): ; Further, .

[0102] in, By filling the diagonal blocks of the node admittance matrix according to the three-phase symmetry principle and combining them with the mutual admittance of each cable segment, a complete three-phase frequency domain Y-bus matrix can be formed, which can be used for subsequent resonant mode analysis, impedance-Nyquist criterion verification, and digital real-time simulation.

[0103] Optional, calculate the overall output impedance When constructing the admittance matrix, after obtaining the overall output impedance, it can be combined with the admittances of each branch in the station network to construct the three-phase node admittance matrix (Y-bus matrix). Under the assumption of three-phase symmetry, the Y-bus matrix has a block diagonal / block sparse structure, and the admittances between nodes are... Constant diagonal block representation. For any two nodes... and ,and The interconnected admittance blocks of the branches between them in the Y matrix are: ; in, For nodes and The complex admittance scalar of the branches, such as the symmetric sequence admittance of the circuit. for Identity matrix. Similarly, nodes. Diagonally positioned self-guided absorber blocks Equal to node The sum of the admittances of all connected branches. Therefore, the entire three-phase Y-bus node admittance matrix can be considered as a sub-matrix block of size 3: the elements of the diagonal sub-matrix block are the sum of the admittances of each phase branch, and the elements of the off-diagonal sub-matrix block are the mutual admittances between phases. Since the three phases are symmetrical, the mutual admittance values ​​are equal, which are expressed as scalar admittances multiplied by the identity matrix. This three-phase Y-bus matrix can be used for subsequent system resonant mode analysis and impedance stability criterion calculation.

[0104] This invention establishes an impedance model for a single-unit converter and analyzes the frequency sweep results of the single-unit converter obtained during the impedance model establishment process based on the multi-time-scale characteristics of the controller, obtaining impedance characteristic analysis results for each frequency band. Based on the impedance characteristic analysis results for each frequency band, the corresponding parameters of the impedance model under different frequency bands are analyzed, and a mathematical model of the inverter's output impedance under all operating conditions is established based on the obtained impedance transfer function and parameters for each frequency band. Based on the output impedance mathematical model, combined with the parameters of the power collection lines in the power station, the connection form of the inverter, and the power station topology, the equivalent impedance of the feeder from the end to the beginning is determined, and based on the equivalent impedance of the feeder, the overall output impedance of the new energy power station at the collection cabinet with multiple feeders connected in parallel is calculated. This invention, through a three-level impedance model of "equipment-cluster-power station," organically combines single-unit converter modeling, multi-unit parallel equivalent impedance, and power station network impedance distribution, achieving a comprehensive stability analysis from the component level to the system level. Furthermore, based on the multi-time-scale characteristics of the controller, the target frequency band is divided into low-frequency, mid-frequency, and high-frequency bands and modeled separately. This avoids coupling interference between different control links at similar frequencies, significantly improves the impedance modeling accuracy over a wide frequency range, and solves the problem in the existing technology where the multi-dimensional coupling mechanism is not clear and the equivalent modeling accuracy needs to be improved when extending the single-machine characteristics to the whole station level.

[0105] In this embodiment of the invention, the use of PINN to determine the equivalent impedance of the feeder solves the core problem of unclear coupling mechanism when expanding from a single unit to the entire station in the traditional method, thereby improving the calculation accuracy of the equivalent impedance of the feeder.

[0106] The embodiments of the present invention also introduce a power-related modulation coefficient to construct an inverter impedance model covering 10% to 90% of the rated power under all operating conditions. This model accurately reflects the impact of load changes on impedance characteristics, such as deep impedance under light load and high-frequency resonance risk under heavy load, thus solving the limitation of traditional models that are only applicable to a single operating condition.

[0107] This invention also proposes a sweep frequency test platform and data processing method based on hardware-in-the-loop simulation. By combining measured data with model parameters, it can be directly applied to the stability verification of new energy power plant planning and design stages, such as Nyquist criterion analysis and resonant mode identification, providing quantitative indicators for on-site commissioning.

[0108] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0109] The following are device embodiments of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0110] Figure 13 The diagram shows a structural schematic of a nonlinear frequency domain coupling impedance modeling device for a new energy grid-connected system according to an embodiment of the present invention. For ease of explanation, only the parts related to the embodiment of the present invention are shown, and are described in detail below: like Figure 13 As shown, a modeling device 13 for a nonlinear frequency domain coupling impedance model of a new energy grid-connected system includes: a modeling module 131, an analysis module 132, and a calculation module 133.

[0111] Modeling module 131 is used to establish the impedance model of a single-unit converter; Analysis module 132 is used to analyze the frequency sweep results of the single-unit converter obtained during the impedance model establishment process based on the multi-time-scale characteristics of the controller, and to obtain the impedance characteristic analysis results of each frequency band. Analysis module 132 is also used to analyze the corresponding parameters of the impedance model under different frequency bands based on the impedance characteristic analysis results of each frequency band, and obtain the impedance transfer function and parameters of each frequency band. Modeling module 131 is also used to establish a mathematical model of the inverter's output impedance under all operating conditions based on the obtained impedance transfer functions and parameters for each frequency band. The calculation module 133 is used to determine the equivalent impedance of the feeder from the end to the beginning based on the output impedance mathematical model, combined with the parameters of the power collection line of the station, the connection form of the inverter and the station topology, and to calculate the overall output impedance of the new energy station level with multiple feeders connected in parallel at the collection cabinet based on the equivalent impedance of the feeder.

[0112] In one possible implementation, the analysis module 132 analyzes the frequency sweep results of the single-unit converter obtained during the impedance model establishment process based on the multi-time-scale characteristics of the controller. When obtaining the impedance characteristic analysis results for each frequency band, it is used for: Obtain the frequency sweep results of the single-unit converter obtained during the impedance model establishment process; The frequency bands in the frequency sweep results are divided into low-mid frequency, mid frequency, and mid-high frequency. Based on the multi-timescale characteristics of the controller, the frequency sweep results of the inverter under a single operating condition are analyzed based on different frequency bands to obtain the correlation between the impedance characteristics of each frequency band and the control loop and power operating conditions.

[0113] In one possible implementation, when the analysis module 132 parses the corresponding parameters of the impedance model under different frequency bands based on the impedance characteristic analysis results of each frequency band, and obtains the impedance transfer function of each frequency band, it is used for: The impedance transfer function in the low-frequency band is: ;in, This indicates the impedance of a low-frequency inverter. This represents the first equivalent negative impedance presented by the converter in the low-frequency range. Represents the Laplace variable. This represents the time constant related to the DC-side dynamics or outer loop control of the inverter. The impedance transfer function in the mid-frequency band is: ;in, This indicates the impedance of a mid-frequency inverter. This represents the second equivalent negative impedance introduced by phase-locked loop control in the mid-frequency band. This indicates the high-frequency correction introduced by the controller's zero point. Indicates the damping ratio. This indicates the dominant natural frequency of the phase-locked loop; The impedance transfer function in the high-frequency band is: ;in, This indicates the impedance of a high-frequency inverter. Represents the equivalent inductive impedance. Indicates equivalent inductive damping, Indicates the resonant frequency. This indicates the damping at the conjugate pole.

[0114] In one possible implementation, when the modeling module 131 establishes a mathematical model of the inverter's output impedance under all operating conditions based on the impedance transfer function and parameters of each frequency band, it is used for: The impedance model structure for each frequency band is determined based on the impedance transfer function of each frequency band. Determine the power modulation coefficient based on the parameters; Based on the frequency band impedance model structure and power modulation coefficient, the mathematical model of the inverter's output impedance under all operating conditions is obtained.

[0115] In one possible implementation, the mathematical model for the output impedance is as follows: ;in, This represents the mathematical model of output impedance.

[0116] In one possible implementation, when the calculation module 133 determines the equivalent impedance of the feeder from the end to the beginning based on the output impedance mathematical model, combined with the parameters of the station's collector lines, the inverter's connection type, and the station's topology, it is used for: Based on the mathematical model of output impedance and the site topology, the total output impedance of the current inverter is determined according to the site collection line parameters between the current inverter access collection cabinet nodes. Based on the total output impedance of the current inverter, calculate the parallel equivalent impedance of N grid-connected inverters connected in parallel within each box-type transformer; Calculate the output impedance of the box-type transformer on the medium-voltage side based on the parallel equivalent impedance and the leakage impedance of the box-type step-up transformer. Based on the station topology and the output impedance of the box-type transformer, the equivalent impedance of the feeder is calculated by recursively pushing upstream from the end box-type transformer along the feeder until it reaches the beginning of the feeder.

[0117] In one possible implementation, when the calculation module 133 determines the total output impedance of the current inverter based on the output impedance mathematical model and the site topology, and based on the site collector line parameters between the current inverter access collector nodes, it is used for: according to Determine the total output impedance of the current inverter; in, This represents the total output impedance of the current inverter. This represents the fitted value of the mathematical model of output impedance. Indicates the impedance of the station's collector lines. This indicates the resistance of the station's collector lines. Represents the imaginary unit. Represents angular frequency. Indicates the inductance of the station's collector lines; Calculation module 133 calculates the parallel equivalent impedance of N grid-connected inverters connected in parallel within each box-type transformer based on the total output impedance of the current inverter. This is used for: Calculate the parallel equivalent impedance of N grid-connected inverters connected in parallel within each box-type transformer; in, Indicates the parallel equivalent impedance. Indicates the current inverter number. This indicates the total number of inverters in each box-type transformer; Calculation module 133 calculates the output impedance of the box-type transformer presented to the outside world on the medium-voltage side based on the parallel equivalent impedance and the leakage impedance of the box-type step-up transformer. This is used for: Calculate the output impedance of the box-type transformer on the medium-voltage side; in, This indicates the output impedance of the transformer substation. This indicates the leakage impedance of the box-type step-up transformer; Calculation module 133, based on the station topology and the output impedance of the transformer substation, calculates the equivalent impedance of the feeder recursively upstream from the end transformer substation along the feeder until it reaches the beginning of the feeder. This is used for: according to Determine the node To the node Branch impedance viewed from the direction; where, Indicates from adjacent nodes Towards Branch impedance viewed from the direction of travel. Indicates adjacent nodes , The cable impedance between them Represents a node The equivalent impedance at the far end; according to Determine the node Equivalent impedance; where, Indicates the equivalent impedance of the feeder; Calculate the feeder equivalent impedance by recursively working upstream from the end box-type transformer along the feeder until the feeder head is reached.

[0118] In one possible implementation, when the calculation module 133 calculates the overall output impedance of the new energy power station at the collection cabinet based on the equivalent impedance of the feeders, it is used for: The feeder equivalent impedance, line equation, and station topology constraints are integrated into the physical information neural network. The physical information neural network is trained with frequency as input and equivalent complex impedance at the bus as output to obtain a nonlinear mapping model from frequency to impedance. When training the physical information neural network, the total loss function is the sum of the frequency domain data fitting loss, topology physical consistency loss, output stability and regularization loss, and physical rationality constraints. An optimized frequency impedance coupling model is output using a nonlinear mapping model. Based on the frequency impedance coupling model, the overall output impedance of the new energy power station with multiple feeders connected in parallel at the collection cabinet is calculated.

[0119] The above embodiments establish an impedance model for a single-unit converter through a modeling module. The analysis module then analyzes the frequency sweep results of the single-unit converter obtained during the impedance model establishment process, based on the controller's multi-timescale characteristics, to obtain impedance characteristic analysis results for each frequency band. Based on these results, the corresponding parameters of the impedance model under different frequency bands are analyzed. The modeling module then establishes a mathematical model of the inverter's output impedance under all operating conditions based on the obtained impedance transfer functions and parameters for each frequency band. The calculation module, based on the output impedance mathematical model and considering the station's collector line parameters, inverter connection type, and station topology, determines the equivalent impedance of the feeder from the end to the beginning of the feeder. Based on this equivalent impedance, the overall output impedance of the new energy station at the collection cabinet, where multiple feeders are connected in parallel, is calculated. This embodiment of the invention organically combines single-unit converter modeling, multi-unit parallel equivalent impedance, and station network impedance distribution through a three-level impedance model of "equipment-cluster-station," achieving a comprehensive stability analysis from the component level to the system level. Furthermore, based on the multi-time-scale characteristics of the controller, the target frequency band is divided into low-frequency, mid-frequency, and high-frequency bands and modeled separately. This avoids coupling interference between different control links at similar frequencies, significantly improves the impedance modeling accuracy over a wide frequency range, and solves the problem in the existing technology where the multi-dimensional coupling mechanism is not clear and the equivalent modeling accuracy needs to be improved when extending the single-machine characteristics to the whole station level.

[0120] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A modeling method of a nonlinear frequency domain coupled impedance model of a new energy grid-connected system, characterized in that, include: An impedance model of a single-unit converter is established, and based on the multi-time-scale characteristics of the controller, the frequency sweep results of the single-unit converter obtained during the establishment of the impedance model are analyzed to obtain the impedance characteristic analysis results of each frequency band. Based on the impedance characteristic analysis results of each frequency band, the corresponding parameters of the impedance model under different frequency bands are analyzed, and based on the obtained impedance transfer function and parameters of each frequency band, a mathematical model of the inverter's output impedance under all operating conditions is established. Based on the output impedance mathematical model, combined with the station's collector line parameters, inverter connection type, and station topology, the feeder equivalent impedance from the end to the beginning of the feeder is determined. Based on the feeder equivalent impedance, the overall output impedance of the new energy station level with multiple feeders connected in parallel at the collection cabinet is calculated.

2. The modeling method of the nonlinear frequency domain coupled impedance model of the grid-connected system of new energy according to claim 1, characterized in that, Based on the multi-time-scale characteristics of the controller, the frequency sweep results of the single-unit converter obtained during the establishment of the impedance model are analyzed to obtain the impedance characteristic analysis results for each frequency band, including: Obtain the frequency sweep results of the single-unit converter obtained during the process of establishing the impedance model; The frequency bands in the frequency sweep results are divided into low-mid frequency, mid frequency, and mid-high frequency. Based on the multi-timescale characteristics of the controller, the frequency sweep results of the inverter under a single operating condition are analyzed based on different frequency bands to obtain the correlation between the impedance characteristics of each frequency band and the control loop and power operating conditions.

3. The modeling method of claim 2, wherein, Based on the impedance characteristic analysis results of each frequency band, the corresponding parameters of the impedance model under different frequency bands are analyzed. Based on the obtained impedance transfer function and parameters for each frequency band, a mathematical model of the inverter's output impedance under all operating conditions is established, including: Based on the impedance characteristic analysis results of each frequency band, the corresponding parameters of the impedance model under different frequency bands are analyzed to obtain the impedance transfer function and parameters of each frequency band. Based on the impedance transfer function of each frequency band and the parameters, a mathematical model of the inverter's output impedance under all operating conditions is established.

4. The modeling method of claim 3, wherein, Based on the impedance characteristic analysis results of each frequency band, the corresponding parameters of the impedance model under different frequency bands are analyzed to obtain the impedance transfer function of each frequency band, including: The impedance transfer function in the low-frequency band is: ； wherein, Zlow represents the impedance of the low frequency inverter, Zlow,eff represents a first equivalent negative impedance presented by the converter at low frequencies, s represents the Laplace variable, τDC represents a time constant related to the dynamic or outer loop control of the DC side of the inverter; The impedance transfer function in the mid-frequency band is: ； wherein, Zmid represents the impedance of the mid-frequency inverter, ZmidPLL represents the second equivalent negative impedance introduced by the phase-locked loop control in the mid-frequency band, Zmidcorr represents the high-frequency correction introduced by the controller zero, ζ represents the damping ratio, ωn represents the natural frequency dominated by the phase-locked loop. The impedance transfer function in the high-frequency band is: ； wherein, represents the impedance of the high frequency band inverter, represents the equivalent inductive impedance, represents the equivalent inductive damping, represents the resonance frequency, represents the damping at the conjugate pole.

5. The modeling method of claim 3 or 4, wherein, Based on the impedance transfer function of each frequency band and the parameters, a mathematical model of the inverter's output impedance under all operating conditions is established, including: Based on the impedance transfer function of each frequency band, determine the structure of the frequency band impedance model; Based on the parameters, determine the power modulation coefficient; Based on the frequency band impedance model structure and the power modulation coefficient, the mathematical model of the inverter's output impedance under all operating conditions is obtained.

6. The modeling method for nonlinear frequency domain coupling impedance model of new energy grid-connected system according to claim 5, characterized in that, The mathematical model for the output impedance is as follows: ； in, This represents the mathematical model of output impedance.

7. The modeling method for nonlinear frequency domain coupling impedance model of new energy grid-connected system according to claim 5, characterized in that, Based on the aforementioned output impedance mathematical model, and considering the station's collector line parameters, inverter connection type, and station topology, the equivalent impedance of the feeder from its end to its beginning is determined, including: Based on the output impedance mathematical model and the site topology, the total output impedance of the current inverter is determined according to the site collection line parameters between the current inverter access collection cabinet nodes. Based on the total output impedance of the current inverter, calculate the parallel equivalent impedance of N grid-connected inverters connected in parallel within each box-type transformer; Based on the parallel equivalent impedance and the leakage impedance of the box-type step-up transformer, calculate the output impedance of the box-type transformer on the medium voltage side. Based on the station topology and the output impedance of the transformer box, the equivalent impedance of the feeder is calculated by recursively moving upstream from the end transformer box along the feeder until it reaches the beginning of the feeder.

8. The modeling method for nonlinear frequency domain coupling impedance model of new energy grid-connected system according to claim 7, characterized in that, Based on the aforementioned output impedance mathematical model and the site topology, and considering the site collector line parameters between the current inverter access collector nodes, the total output impedance of the current inverter is determined, including: according to Determine the total output impedance of the current inverter; in, This represents the total output impedance of the current inverter. This represents the fitted value of the mathematical model of the output impedance. Indicates the impedance of the station's collector lines. This indicates the resistance of the station's collector lines. Represents the imaginary unit. Represents angular frequency. Indicates the inductance of the station's collector lines; Based on the total output impedance of the current inverter, calculate the parallel equivalent impedance of the N grid-connected inverters connected in parallel within each box-type transformer, including: Calculate the parallel equivalent impedance of N grid-connected inverters connected in parallel within each box-type transformer; in, Indicates the parallel equivalent impedance. Indicates the current inverter number. This indicates the total number of inverters in each box-type transformer; Based on the parallel equivalent impedance and the leakage impedance of the box-type step-up transformer, calculate the output impedance of the box-type transformer on the medium-voltage side, including: Calculate the output impedance of the box-type transformer on the medium-voltage side; in, This indicates the output impedance of the transformer substation. This indicates the leakage impedance of the box-type step-up transformer; Based on the aforementioned station topology and the output impedance of the transformer substation, the equivalent impedance of the feeder is calculated by recursively moving upstream from the end transformer substation to the beginning of the feeder, including: according to Determine the node To the node Branch impedance viewed from the direction; where, Indicates from adjacent nodes Towards Branch impedance viewed from the direction of travel. Indicates adjacent nodes , The cable impedance between them Represents a node The equivalent impedance at the far end; according to Determine the node Equivalent impedance; where, Indicates the equivalent impedance of the feeder; Calculate the feeder equivalent impedance by recursively working upstream from the end box-type transformer along the feeder until the feeder head is reached.

9. The modeling method for nonlinear frequency domain coupling impedance model of new energy grid-connected system according to claim 8, characterized in that, Based on the equivalent impedance of the feeders, calculate the overall output impedance of the new energy power station at the collection cabinet where multiple feeders are connected in parallel, including: The equivalent impedance of the feeder, the line equation, and the station topology constraints are integrated into the physical information neural network. The physical information neural network is trained with frequency as input and equivalent complex impedance at the bus as output to obtain a nonlinear mapping model from frequency to impedance. When training the physical information neural network, the total loss function is the sum of the frequency domain data fitting loss, the topology physical consistency loss, the output stability and regularization loss, and the weighted loss of physical rationality constraints. Using the aforementioned nonlinear mapping model, an optimized frequency impedance coupling model is output. Based on the frequency impedance coupling model, the overall output impedance of the new energy power station level with multiple feeders connected in parallel at the collection cabinet is calculated.

10. A modeling device for a nonlinear frequency domain coupling impedance model of a new energy grid-connected system, characterized in that, include: The modeling module is used to build the impedance model of a single-unit converter. The analysis module is used to analyze the frequency sweep results of the single-unit converter obtained during the establishment of the impedance model based on the multi-time-scale characteristics of the controller, and to obtain the impedance characteristic analysis results of each frequency band. The analysis module is also used to analyze the corresponding parameters of the impedance model under different frequency bands based on the impedance characteristic analysis results of each frequency band, and obtain the impedance transfer function and parameters of each frequency band. The modeling module is also used to establish a mathematical model of the inverter's output impedance under all operating conditions based on the obtained impedance transfer functions and parameters for each frequency band. The calculation module is used to determine the equivalent impedance of the feeder from the end to the beginning based on the output impedance mathematical model, combined with the parameters of the power collection line of the power station, the connection form of the inverter and the topology of the power station, and to calculate the overall output impedance of the new energy power station with multiple feeders connected in parallel at the collection cabinet based on the equivalent impedance of the feeder.

Images