A broadband impedance scanning method for new energy

Abstract

This application provides a broadband impedance scanning method for new energy sources, comprising: acquiring the grid-connected system status and prior parameters of the injection link; generating a broadband disturbance command based on the status information, controlling the silicon carbide disturbance injection unit to superimpose a small-signal disturbance through the injection transformer; determining a feedforward compensation term based on the prior parameters and the command, and introducing it into the control loop of the energy storage converter unit to counteract the closed-loop suppression effect; acquiring signals and performing baseline subtraction processing to obtain a pure disturbance response; calculating the impedance based on the response and performing verification. This invention, through physical function decoupling and a feedforward adaptive compensation mechanism, overcomes the limitations of traditional measurement bandwidth without changing the main power impedance characteristics. Combined with baseline subtraction and consistency verification mechanisms, it improves the accuracy and robustness against operating condition drift in broadband impedance measurements.

Description

Technical Field

[0001] This application relates to the field of new energy grid-connected impedance measurement technology, and more specifically to a new energy broadband impedance scanning method. Background Technology

[0002] With the continuous increase in the installed capacity of renewable energy, the wide-frequency oscillation problem caused by the large-scale grid connection of renewable energy converters is becoming increasingly prominent. Obtaining the frequency domain impedance characteristics of the grid connection interface is of engineering significance for revealing the oscillation mechanism and supporting system stability assessment and parameter optimization. In field measurements, it is necessary to quickly and with high fidelity acquire impedance response data over a wide frequency range to maintain the safe and stable operation of renewable energy power plants.

[0003] Currently, engineering practices for field impedance measurement mainly rely on disturbance injection technology based on the main converter or external, independent high-power frequency sweeping devices. The former utilizes the existing grid-connected converter at the power station and generates disturbances by superimposing harmonic commands of a predetermined frequency into the output through modification of the control software; the latter requires a large power electronic device with broadband excitation capability connected in parallel or series at the grid connection point. Its rated capacity usually needs to reach a considerable proportion of the main power of the power station. A broadband excitation signal is injected into the grid through an independent control system, and the voltage and current responses at the grid connection point are recorded simultaneously.

[0004] However, when facing the complex and ever-changing operating conditions of new energy grid connection, existing measurement methods still have common problems in terms of effective frequency sweep range, measurement fidelity, and system operating condition consistency. On the one hand, due to the limitations of the main converter's hardware switching frequency and the inherent adversarial suppression effect of the closed-loop control system, the disturbance injection capability in the mid-to-high frequency band is limited, causing the acquired response signal to be easily submerged by grid background harmonics, making it difficult to obtain a high-precision broadband impedance spectrum. On the other hand, during the long-duration frequency sweep process, fluctuations in the operating conditions within the substation and changes in the grid topology can disrupt the consistency of time-series measurement data, causing the final spliced ​​impedance result to lose its physical accuracy. Therefore, there is an urgent need for a method that can effectively improve the accuracy and robustness of broadband disturbance measurement without interfering with the steady-state operation of the system. Summary of the Invention

[0005] The purpose of this invention is to provide a broadband impedance scanning method for new energy sources, in order to solve the aforementioned problems existing in the prior art.

[0006] Technical solution: A broadband impedance scanning method for new energy sources, comprising:

[0007] Obtain the operating status information of the grid-connected system and the pre-calibrated prior parameters of the injection link;

[0008] Based on the grid-connected system's operating status information, a wideband disturbance command is generated, and the silicon carbide disturbance injection unit is controlled to superimpose a small signal disturbance through the injection transformer at the grid connection point according to the wideband disturbance command.

[0009] Based on the prior parameters of the injected link and the broadband disturbance command, the feedforward compensation term is determined and introduced into the control loop of the energy storage converter unit that maintains steady-state grid connection, so as to counteract the closed-loop suppression effect of the energy storage converter unit on small-signal disturbances.

[0010] Collect the grid connection point voltage and current signals before and after the small signal disturbance is superimposed, perform baseline subtraction processing, and obtain the pure disturbance response component;

[0011] The equivalent port impedance of the grid connection point is calculated based on the pure disturbance response components, and the effectiveness and relative stability of the impedance measurement are verified by combining the pre-configured judgment threshold.

[0012] For example, the silicon carbide disturbance injection unit is connected to the primary winding of the injection transformer via a second-order low-pass filter;

[0013] Among them, the cutoff frequency of the second-order low-pass filter is greater than the highest scanning frequency of the small signal disturbance, and less than the preset ratio of the switching frequency of the silicon carbide disturbance injection unit.

[0014] The ratio of the power frequency voltage drop generated by the equivalent leakage inductance of the secondary side of the injected transformer under the rated current of the main power channel to the rated voltage at the grid connection point is less than the preset voltage drop ratio threshold.

[0015] For example, the step of determining the feedforward compensation term based on the injected link prior parameters and the broadband disturbance command, and introducing the feedforward compensation term into the control loop of the energy storage converter unit that maintains steady-state grid connection, includes:

[0016] Based on the injection link transfer function in the prior parameters of the injection link and the feedforward channel transfer function of the energy storage converter, a comprehensive calculation is performed to obtain the frequency adaptive equivalent conversion coefficient that includes amplitude matching and phase alignment characteristics.

[0017] Based on the upper limit frequency of the effective feedforward bandwidth of the energy storage converter, the scanning frequency points are divided into a feedforward compensable subset and a residual subset.

[0018] For the frequency points within the feedforward compensable subset, a feedforward compensation term is constructed based on the frequency adaptive equivalent conversion coefficient and the wideband disturbance command, and the feedforward compensation term is subtracted from the original voltage feedforward reference of the energy storage converter unit.

[0019] For example, in the step of generating a broadband disturbance command based on the grid-connected system operating status information, the generated broadband disturbance command includes a frequency sweep mode configuration, which includes a single-frequency point-by-point scanning mode and a multi-frequency parallel scanning mode.

[0020] For example, when the frequency sweep mode is configured as a multi-frequency parallel sweep mode, the step of determining the sweep frequency in the step of generating a wideband disturbance command based on the grid-connected system operating status information includes:

[0021] Constrain the scanning frequency to a preset frequency grid;

[0022] Candidate frequency groups are constructed based on the scanning frequencies on the frequency grid, and the sets of second- and third-order intermodulation products of the frequencies within the group are determined.

[0023] Calculate the frequency difference between the product frequencies in the sets of second- and third-order intermodulation products and any scanning frequency;

[0024] When the frequency difference is greater than the preset intermodulation protection bandwidth, the candidate frequency group is determined to meet the intermodulation safety constraints.

[0025] For example, the steps of acquiring the grid connection point voltage and current signals before and after the superposition of small-signal disturbances, performing baseline subtraction processing, and obtaining the pure disturbance response component include:

[0026] Within the baseline acquisition window without disturbance injection enabled, background frequency domain components are extracted from the grid connection point voltage and current signals based on preset Fourier extraction coefficients.

[0027] Within the injection measurement window after disturbance injection is enabled and transient decay has occurred, the total frequency domain component is extracted from the grid connection point voltage and current signals based on the Fourier extraction coefficients, wherein the time length of the injection measurement window is equal to that of the baseline acquisition window.

[0028] By performing a complex subtraction operation between the total frequency domain component and the background frequency domain component, the pure disturbance response component that eliminates background harmonic interference is obtained.

[0029] For example, before the steps of obtaining the grid-connected system operating status information and the pre-calibrated injection link prior parameters, a step of pre-calibrating the injection link prior parameters is also included, specifically including:

[0030] Under the steady-state operation of the grid-connected system and without disturbance injection, the grid connection point voltage is collected to estimate the broadband background noise power spectrum at each scanning frequency point;

[0031] Under no-load or light-load conditions of the grid-connected system, the silicon carbide disturbance injection unit is controlled to output calibration signals sequentially and synchronously collect the response voltage at the grid connection point in order to extract the injection link transfer function at each scanning frequency point.

[0032] The broadband background noise power spectrum and the injection link transfer function are used together to construct the pre-calibrated injection link prior parameters.

[0033] For example, the steps of performing impedance measurement validity and relative stability verification in conjunction with a pre-configured decision threshold include:

[0034] When dividing the scanning frequency points into multiple frequency groups for batch injection measurement, the same shared pilot frequency is embedded in each frequency group;

[0035] Extract the pilot impedance measured at different frequency groups for the shared pilot frequency;

[0036] Calculate the cross-group relative deviation of the pilot impedance between different frequency groups;

[0037] When the relative deviation between groups exceeds the preset consistency threshold, the system operating condition is determined to have drifted, and the corresponding frequency group after the drift inflection point is marked as the group to be retested based on the time node when the deviation exceeds the standard.

[0038] For example, after calculating the cross-group relative deviation of the pilot impedance between different frequency groups, the method further includes:

[0039] When all cross-group relative deviations do not exceed the consistency threshold, for a given shared pilot frequency, calculate the average value of the pilot impedance measured in multiple frequency groups, and use the average value as the output impedance of the shared pilot frequency in the broadband impedance spectrum at the final grid connection point.

[0040] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer-readable storage medium, and the computer program is configured to execute the above-described new energy broadband impedance scanning method when running.

[0041] Beneficial effects: This invention breaks through the traditional measurement bandwidth limitation by decoupling physical functions and using a feedforward adaptive compensation mechanism without changing the main power impedance characteristics. Combined with baseline subtraction and consistency verification mechanisms, it improves the accuracy and robustness of wideband impedance measurement under operating conditions. Attached Figure Description

[0042] Figure 1 This is a flowchart of the steps of the new energy broadband impedance scanning method of the present invention.

[0043] Figure 2 This is a flowchart illustrating the overall execution of the new energy broadband impedance scanning method of the present invention.

[0044] Figure 3 This is a flowchart illustrating the steps of introducing a feedforward compensation term into the control loop of the energy storage converter unit that maintains steady-state grid connection, as per the present invention.

[0045] Figure 4 This is a flowchart illustrating the steps of obtaining the pure disturbance response component according to the present invention.

[0046] Figure 5 This is a flowchart of the steps for pre-calibrating the prior parameters of the injection link in this invention. Detailed Implementation

[0047] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0048] Example 1: This example provides a broadband impedance scanning method for new energy sources, such as... Figure 1 As shown, the method includes the following steps:

[0049] Optionally, the system can acquire the operating status information of the grid-connected system and the pre-calibrated prior parameters of the injection link.

[0050] In this embodiment, the grid-connected system operating status information refers to real-time data reflecting the current physical characteristics of the new energy power station's grid-connected operation. Specifically, the grid-connected system operating status information includes, but is not limited to, the current control mode of the energy storage converter unit (such as constant voltage and constant frequency mode or grid following mode), the rated voltage amplitude at the grid connection point, and the rated current of the main power channel. The injected link prior parameters refer to calibration data reflecting the physical transmission characteristics of the broadband disturbance signal from the generation end to the PCC at the grid connection point. In this embodiment, the injected link prior parameters are obtained through offline calibration, specifically including the injected link transfer function and background noise power spectrum at each scanning frequency point.

[0051] Specifically, this embodiment constructs a hardware architecture for executing the method. Its overall structure includes a new energy generation terminal, an energy storage converter unit, an injection transformer, a silicon carbide disturbance injection unit, and a grid terminal. The new energy generation terminal is connected to the energy storage converter unit and is used to provide and regulate the main power energy of the system. The silicon carbide disturbance injection unit uses silicon carbide MOSFET power devices as high-frequency switching execution units to achieve small-signal disturbance injection in the mid-to-high frequency band. Silicon carbide MOSFET power devices have low switching losses and high switching speed characteristics. In this embodiment, the switching frequency of the silicon carbide MOSFET is set to 100 kHz. By decoupling the grid-connected operation function from the wideband scanning function, the energy storage converter unit maintains the grid-connected steady-state operating point and provides low-frequency high-power regulation capability, while the silicon carbide disturbance injection unit only undertakes the generation and injection of mid-to-high frequency small-amplitude disturbance signals. This expands the effective measurement frequency band while reducing the need for modification to the original system.

[0052] Optionally, the silicon carbide disturbance injection unit is connected to the primary winding of the injection transformer via a second-order low-pass filter; wherein, the cutoff frequency of the second-order low-pass filter is greater than the highest scanning frequency of the small-signal disturbance and less than a preset ratio of the switching frequency of the silicon carbide disturbance injection unit; the ratio of the power frequency voltage drop generated by the equivalent leakage inductance of the secondary side of the injection transformer under the rated current of the main power channel to the rated voltage at the grid connection point is less than a preset voltage drop ratio threshold.

[0053] In this embodiment, the second-order low-pass filter is composed of an inductor and a capacitor, used to suppress harmonic ripple generated by the high-frequency switching operation of the silicon carbide device. The cutoff frequency of the second-order low-pass filter is calculated using the following formula:

[0054] f c ,LC=1 / (2*3.14*sqrt(L f *C f ));

[0055] In the formula, f c LC is the cutoff frequency of the second-order low-pass filter, and L f C is the value of the filter inductance. f This is the value of the filter capacitor.

[0056] Accordingly, in order to ensure that the disturbance signal within the scanning frequency range is not significantly attenuated by the filter, while effectively filtering out high-frequency switching components, the cutoff frequency design must satisfy constraints. For example, let the highest scanning frequency of the small-signal disturbance be f. scan The switching frequency of the silicon carbide perturbation injection unit is fmax. sw For SiC, the preset ratio can be 1 / 5, meaning the design constraint is satisfied:

[0057] f scan ,max <f c LC <f sw ,SiC / 5;

[0058] Specifically, the frequency constraints described above ensure the injection accuracy of the disturbance signal. Furthermore, as a key component in broadband disturbance coupling, the injection transformer's design must balance the injected disturbance power with its impedance impact on the main power path. The equivalent leakage inductance L on the secondary side of the injection transformer... σ s should be as small as possible. Specifically, the negligible condition means that the ratio of the voltage drop generated by the equivalent leakage inductance on the secondary side at the power frequency angular frequency and the rated current of the main power channel to the rated voltage at the grid connection point is less than 1%, that is, it satisfies:

[0059] (w1*L σ ,s*I rated ) / V PCC ,rated<1%;

[0060] In the formula, w1 is the power frequency angular frequency, L σ ,s is the equivalent leakage inductance injected into the secondary side of the transformer, I rated Main power channel rated current, V PCC ,rated is the rated voltage at the grid connection point. When the above formula holds true, the change in the impedance characteristics of the main power channel caused by the injection transformer can be ignored within the engineering accuracy range.

[0061] Optionally, the grid-connected system is a three-phase AC system, and the voltage and current signals at the grid connection point are three-phase signals. In the step of performing baseline subtraction to obtain pure disturbance response components, coordinate transformation is performed on the three-phase signals, and disturbance response components of the direct axis and quadrature axis are extracted in the synchronous rotating coordinate system. Complex division is performed on the voltage and current components in the disturbance response components of the direct axis and quadrature axis to construct a two-dimensional impedance matrix.

[0062] Specifically, when the grid-connected system is applied in a three-phase AC substation, the silicon carbide disturbance injection unit can adopt a three-phase full-bridge voltage source inverter topology. The collected three-phase voltage and current signals at the grid connection point are converted to the dq rotating coordinate system through Parker transformation. The direct-axis (d-axis) and quadrature-axis (q-axis) components reflect the dynamic characteristics of the system in the rotating vector space.

[0063] Based on this, by performing two independent small-signal injection actions on the d-axis and q-axis respectively, the corresponding complex frequency domain response is extracted, thus constructing a 2*2 dimension impedance matrix. The two-dimensional impedance matrix includes direct-axis self-impedance, quadrature-axis self-impedance, and inter-axis mutual impedance, capable of describing the frequency characteristics of a three-phase system over a wide frequency band. During coordinate transformation, a synchronous high-speed acquisition system is used; in this embodiment, the sampling rate is set to 200 kS / s, and the analog-to-digital converter resolution is 16 bits. The time deviation between the voltage and current sampling channels is controlled within 1% of the sampling period, ensuring the phase accuracy of the impedance matrix construction.

[0064] Example 2 further details the overall execution flow of the new energy broadband impedance scanning method, such as... Figure 2 As shown, in one possible implementation, the following steps are included:

[0065] Optionally, the system can acquire the operating status information of the grid-connected system and the pre-calibrated prior parameters of the injection link.

[0066] Accordingly, the grid-connected system operation status information specifically includes the real-time control parameters of the energy storage converter units within the substation, the rated voltage reference of the grid connection point, and the grid synchronization frequency. The control parameters indicate whether the energy storage converter units are currently in constant voltage and frequency mode or grid-following mode, providing boundary conditions for the generation of subsequent disturbance commands. The pre-calibrated injection link prior parameters are physical characteristic data obtained through offline testing, specifically including the injection link transfer function and background noise power spectrum for each scanning frequency point. The injection link transfer function describes the complex gain relationship between the disturbance signal from the output side of the silicon carbide disturbance injection unit to the injection side of the grid connection point, including amplitude attenuation and phase shift. The background noise power spectrum characterizes the environmental noise distribution characteristics of the grid connection point under static operating conditions, assisting in subsequent signal quality assessment.

[0067] Optionally, a broadband disturbance command is generated based on the grid-connected system operating status information. After the feedforward compensation term takes effect, the silicon carbide disturbance injection unit is controlled to superimpose a small signal disturbance through the injection transformer at the grid connection point according to the broadband disturbance command.

[0068] Specifically, the system calculates the disturbance amplitude that conforms to the small-signal assumption based on the rated voltage at the grid connection point and a preset small disturbance ratio coefficient. The wideband disturbance command includes a predefined scan frequency sequence, the initial phase angle of each scan frequency point, and the corresponding command amplitude. Upon receiving the command, the silicon carbide disturbance injection unit controls its internal switching devices to perform pulse width modulation, generating a disturbance current containing the target frequency component. This disturbance current is superimposed on the grid connection point through an injection transformer physically connected in series in the grid connection circuit. For example, in single-frequency point-by-point scanning mode, the system controls the silicon carbide disturbance injection unit to inject sine waves of individual frequencies sequentially. In multi-frequency parallel scanning mode, the system combines multiple scan frequencies into a composite disturbance signal for simultaneous injection, improving scanning efficiency.

[0069] Optionally, a feedforward compensation term is determined based on the prior parameters of the injection link and the broadband disturbance command. The feedforward compensation term is then introduced into the control loop of the energy storage converter unit that maintains steady-state grid connection, so as to counteract the closed-loop suppression effect of the energy storage converter unit on the small signal disturbance to be injected.

[0070] Furthermore, because the closed-loop control system of the energy storage converter treats the injected disturbance as grid fluctuation and performs suppression operations, the actual response component at the grid connection point is weakened. To eliminate this adverse effect, this embodiment utilizes the injection link transfer function in the injection link prior parameters to map the broadband disturbance command into an equivalent disturbance component on the voltage sampling side of the energy storage converter, thus determining the feedforward compensation term. The feedforward compensation term is then superimposed onto the feedforward control channel of the energy storage converter in a negative injection manner. Through this type of feedforward cancellation mechanism, the control strategy of the energy storage converter no longer responds to the injected disturbance signal, ensuring that the disturbance signal can be superimposed at the grid connection point as designed.

[0071] Optionally, the grid connection point voltage and current signals before and after the small-signal disturbance superposition are collected, and baseline subtraction processing is performed to obtain the pure disturbance response component.

[0072] Accordingly, the discrete sequences of voltage and current at the grid connection point are acquired in real time using a synchronous sampling device. To extract the weak disturbance response from the original signal containing the fundamental frequency and complex harmonics, this embodiment performs baseline subtraction. Specifically, the system extracts the background component within the baseline acquisition window without injection and the total component within the measurement window after injection is enabled. The calculation formula for the pure disturbance response component is as follows:

[0073] V p (jwk)=V total (jwk)-V base (jwk);

[0074] In the formula, V p (jwk) represents the extracted pure disturbance voltage response component, V total (jwk) represents the frequency domain component of the total voltage extracted within the measurement window, V base (jwk) represents the background voltage frequency domain component extracted within the baseline acquisition window, where j is the imaginary unit and wk is the angular frequency of the k-th scanning frequency point. By subtracting the background component from the total component, the inherent steady-state interference of the grid-connected system can be eliminated, and response data caused solely by injected disturbances can be obtained.

[0075] Optionally, the equivalent port impedance of the grid connection point is calculated based on the pure disturbance response components, and the effectiveness and relative stability of the impedance measurement are verified by combining a pre-configured judgment threshold.

[0076] Furthermore, the system performs a complex division operation between the pure disturbance voltage response component and the corresponding pure disturbance current response component to obtain the equivalent port impedance at the grid connection point. The calculation formula is as follows:

[0077] Z PCC (jwk)=V p (jwk) / I p(jwk);

[0078] In the formula, Z PCC (jwk) is the equivalent port impedance at the grid connection point, V p (jwk) is the pure disturbance voltage response component, I p (jwk) represents the pure disturbance current response component. After obtaining the calculation results, the system executes self-test logic. Accordingly, validity verification is performed, i.e., determining whether the extracted voltage response amplitude is greater than the preset minimum measurable threshold to ensure that the signal is not drowned out by background noise. Further, relative stability verification is performed, evaluating the consistency of the measurement by comparing the deviation rate of two measurements before and after at the same frequency point. When the deviation rate is less than the preset relative stability threshold (e.g., 5%), the impedance measurement result at that frequency point is determined to be valid and stable. The system summarizes the results of all frequency points and generates a broadband impedance spectrum.

[0079] Example 3 further details the steps of determining the feedforward compensation term based on the injected link prior parameters and broadband disturbance commands, and introducing the feedforward compensation term into the control loop of the energy storage converter unit that maintains steady-state grid connection. For example... Figure 3 As shown, in one possible implementation, the following steps are included:

[0080] Optionally, based on the injection link transfer function in the prior parameters of the injection link and the feedforward channel transfer function of the energy storage converter obtained from the grid-connected system operation status information, a comprehensive conversion is performed to obtain the frequency adaptive equivalent conversion coefficient containing amplitude matching and phase alignment characteristics.

[0081] Specifically, the injection link transfer function characterizes the frequency response characteristics from the output reference terminal of the silicon carbide disturbance injection unit to the actual equivalent disturbance formed at the grid connection point. It comprehensively reflects the frequency attenuation and phase shift laws of the injection transformer turns ratio, magnetizing impedance, leakage inductance, winding resistance, and filter. The feedforward channel transfer function characterizes the transformation characteristics from the grid connection point voltage sampling to the feedforward reference quantity of the energy storage converter unit. During wideband scanning, the amplitude-frequency and phase-frequency characteristics of the injection link shift with frequency changes, and a fixed scaling factor cannot reflect the disturbance transmission relationship in the mid-to-high frequency band. To calculate the degree of perception of the grid connection point disturbance by the feedforward channel of the energy storage converter unit, the system performs complex multiplication operations on the above two transfer functions at the corresponding scanning frequency points to calculate the frequency adaptive equivalent conversion coefficient. The specific calculation formula is as follows:

[0082] κ(w k )=H ff (j*w k )*H T (j*w k );

[0083] In the formula, κ(w)k ) is at the scanning angular frequency w k Frequency adaptive equivalent conversion factor at H ff (j*w k ) is the feedforward channel at the scanning angular frequency w k Complex transfer gain at H T (j*w k ) is the injection link at the scanning angular frequency w k The complex transfer gain at point j, where j is the imaginary unit.

[0084] In some optional implementations, when the scanning frequency is at the power frequency or a lower frequency band, the injection link transfer function can be approximated as the pure turns ratio attenuation of the injection transformer, and the feedforward channel transfer function can be approximated as the pure proportional gain. In this case, the system can use a simplified formula to calculate the equivalent conversion coefficients. The specific simplified calculation formula is as follows:

[0085] κ0=K ff / n T ;

[0086] In the formula, κ0 is the simplified equivalent conversion coefficient for the low-frequency band, and K ff For the feedforward channel, n is the pure proportional gain in the low-frequency range. T This refers to the primary / secondary turns ratio injected into the transformer.

[0087] Optionally, based on the upper limit frequency of the effective feedforward bandwidth of the energy storage converter obtained from the grid-connected system operation status information, the scanning frequency points are divided into a feedforward compensable subset and a residual subset, wherein the frequency points in the residual subset exceed the control bandwidth of the energy storage converter and therefore do not perform feedforward compensation.

[0088] Specifically, the upper limit frequency of the effective feedforward bandwidth refers to the highest cutoff frequency at which the feedforward control loop of the energy storage converter can effectively track and suppress grid voltage fluctuations. Optionally, the upper limit frequency of the effective feedforward bandwidth can be determined by injecting a known step or sinusoidal frequency sweep excitation into the feedforward channel of the energy storage converter, and by observing the amplitude-frequency response curve of the feedforward output, extracting the frequency value corresponding to the amplitude decay to -3dB as the upper limit frequency of the effective feedforward bandwidth.

[0089] Correspondingly, when the injected disturbance frequency is higher than this frequency, the control loop is limited by the hardware sampling rate and computation delay, and cannot produce a substantial closed-loop regulation effect on the disturbance. The system extracts each current scanning frequency point and compares each scanning frequency point with the upper limit frequency of the effective feedforward bandwidth. When the scanning frequency point is less than or equal to the upper limit frequency of the effective feedforward bandwidth, the scanning frequency point is assigned to the feedforward compensable subset. When the scanning frequency point is greater than the upper limit frequency of the effective feedforward bandwidth, the scanning frequency point is assigned to the residual subset.

[0090] Furthermore, for frequency components assigned to the residual subset, the system does not perform decoupling compensation operations at the control level. Since their frequencies exceed the control bandwidth, the adverse effects exerted by the energy storage converter on them are negligible. The remaining weak interference components will be eliminated through subsequent baseline subtraction, providing a computational safety net during the impedance data extraction stage.

[0091] Optionally, for frequency points within the feedforward compensable subset, a feedforward compensation term is constructed based on frequency adaptive equivalent conversion coefficients and wideband disturbance commands, and the feedforward compensation term is subtracted from the original voltage feedforward reference quantity generated in real time from the control loop of the energy storage converter unit.

[0092] In this embodiment, for each scanning frequency point within the feedforward compensable subset, the system extracts the disturbance amplitude and initial phase angle set in the broadband disturbance command, and performs calculations by combining the amplitude and phase angle of the frequency adaptive equivalent conversion coefficient. Specifically, the system multiplies the disturbance amplitude set in the command with the amplitude of the conversion coefficient, and adds the initial phase angle set in the command with the phase angle of the conversion coefficient to construct the feedforward compensation term for the corresponding frequency point. This feedforward compensation term is equivalent in the time-domain physical sense to the disturbance component that the energy storage converter's feedforward channel will sense. Furthermore, within the control execution cycle of the energy storage converter, the system subtracts the feedforward compensation term from the original voltage feedforward reference value generated by the control system.

[0093] Introducing the compensated voltage feedforward reference quantity V ff_prime (t)=V ff0 (t)-V comp (t);

[0094] In the formula, V ff0 (t) represents the original voltage feedforward reference value, V comp (t) represents the feedforward compensation term constructed for the feedforward compensable subset, where t is a time variable.

[0095] By performing the above subtraction operation in the voltage feedforward reference, the system cancels the sensing component of the disturbance signal by the feedforward channel, so that the energy storage converter will not judge the measurement excitation issued by the silicon carbide disturbance injection unit as a grid distortion that needs to be suppressed within the scanning frequency band. This avoids the mutual conflict between the equipment's own control loop and the impedance measurement mechanism, and ensures that small signal disturbances can establish an excitation amplitude that meets expectations at the grid connection point.

[0096] Example 4 further details the parameter adaptive adjustment control process in single-frequency point-by-point scanning mode. In one possible implementation, it includes the following steps:

[0097] Optionally, in the step of generating a broadband disturbance command based on the grid-connected system operating status information, the generated broadband disturbance command includes a frequency sweep mode configuration, which includes a single-frequency point-by-point scanning mode and a multi-frequency parallel scanning mode.

[0098] Accordingly, the frequency sweep mode configuration is used to indicate the signal generation and injection logic adopted by the system in the current measurement task. Specifically, the single-frequency point-by-point scan mode refers to controlling the silicon carbide perturbation injection unit to output a sinusoidal perturbation of a single frequency within a set scan hold time, and then switching to the next frequency to be measured after the current frequency point measurement is completed. The multi-frequency parallel scan mode refers to simultaneously injecting a composite perturbation signal containing multiple frequency components within the same hold time. This embodiment focuses on describing the closed-loop control mechanism in the single-frequency point-by-point scan mode. Since the background harmonics of different frequencies differ greatly, using a fixed amplitude for single-frequency injection will result in insufficient signal-to-noise ratio at some frequencies and redundant amplitude at others. Therefore, adaptive adjustment logic needs to be introduced.

[0099] In one possible implementation, when the frequency sweep mode is configured as a single-frequency point-by-point scanning mode, after acquiring the grid connection point voltage and current signals before and after the small signal disturbance superposition, performing baseline subtraction processing, and obtaining the pure disturbance response component, the method further includes: calculating the measured signal-to-noise ratio of the current frequency point based on the pure disturbance response component and the root mean square of the background noise in the frequency band adjacent to the injection frequency extracted from the grid connection point voltage and current signals.

[0100] Furthermore, to evaluate the measurement quality of the current disturbance signal at the grid connection point, the system extracts the amplitude of the pure disturbance response component at the current frequency as a signal power characterization quantity. Simultaneously, the system extracts a preset bandwidth range centered on the current injection frequency from the spectrum of the acquired grid connection point voltage and current signals as a neighboring frequency band. By excluding the center resolution cell where the current injection frequency is located, the root mean square (RMS) of the spectral amplitude within this neighboring frequency band is calculated to obtain the root mean square (RMS) of the background noise. The RMS of the background noise is used to characterize the local noise floor level near the current frequency point.

[0101] Measured signal-to-noise ratio (SNR) meas =V p_mag / σ noise ;

[0102] In the formula, V p_mag Let σ be the amplitude of the pure disturbance voltage response component. noise The signal-to-noise ratio (SNR) is the root mean square of the background noise. By combining the noise floor levels of adjacent frequency bands to calculate the SNR, the problem of inaccurate estimation of the global broadband noise of the power grid can be avoided, providing accurate local signal quality feedback.

[0103] Optionally, the measured signal-to-noise ratio can be compared with a preset target signal-to-noise ratio.

[0104] Accordingly, the target signal-to-noise ratio (SNR) is a pre-set scalar threshold within the system, used to define the minimum signal quality limit required for the impedance calculation results to achieve engineering validity. The system compares the calculated measured SNR with this target SNR to determine whether the initial injection measurement result at the current frequency is reliable.

[0105] Optionally, when the measured signal-to-noise ratio is lower than the target signal-to-noise ratio, the disturbance amplitude in the single-frequency point-by-point scanning mode is increased within the range not exceeding the preset small signal constraint upper limit, and the wideband disturbance command is updated according to the increased disturbance amplitude to trigger the re-injection and measurement of the current frequency point; when the measured signal-to-noise ratio is not lower than the target signal-to-noise ratio, the measurement of the current frequency point is determined to be valid, the current measurement result is maintained, and the scanning of subsequent frequency points continues.

[0106] Specifically, the small-signal constraint upper limit refers to the maximum perturbation amplitude boundary that allows the system to maintain the linear response assumption. When the measured signal-to-noise ratio fails to meet the standard, it indicates that the current injected signal amplitude is too small, causing the response to be interfered with by background noise. The system executes amplitude closed-loop adjustment logic to proportionally increase the amplitude of the next injection, provided that it does not exceed the small-signal constraint upper limit.

[0107] Updated perturbation amplitude A new =min(A old *(SNR target / SNR meas ),A max );

[0108] In the formula, min is the function that takes the minimum value, A old The perturbation amplitude used in the current injection, SNR target For target signal-to-noise ratio, SNR meas For the actual measured signal-to-noise ratio, A max The upper limit of the small signal constraint.

[0109] For example, when the measured signal-to-noise ratio calculated after the system performs injection at a certain intermediate frequency point does not reach the target signal-to-noise ratio, the system increases the disturbance amplitude proportionally according to the above formula. Since the increased amplitude is still less than the upper limit of the small signal constraint, the system re-performs injection and measurement at that frequency point with the new amplitude, so that the measurement accuracy at that frequency point meets the engineering requirements.

[0110] Furthermore, in some optional implementations, when the measured signal-to-noise ratio is higher than the target signal-to-noise ratio, the system can choose to reduce the subsequent injection amplitude to minimize the impact of the measurement action on the grid-connected system.

[0111] Updated perturbation amplitude A new =max(A old *(SNR target / SNR meas ),Amin );

[0112] In the formula, max is the function that takes the maximum value, and A old The perturbation amplitude used in the current injection, SNR target For target signal-to-noise ratio, SNR meas For the actual measured signal-to-noise ratio, A min This is a preset minimum disturbance amplitude limit to prevent calculation errors. Through the aforementioned bidirectional adaptive amplitude modulation mechanism, the system achieves the optimal balance between measurement accuracy and the degree of power grid disturbance in single-frequency point-by-point scanning mode.

[0113] Example 5 provides another implementation method for the scanning mode: for the multi-frequency parallel scanning mode, the link-aware multi-frequency signal grouping algorithm is described in detail. The remaining process is the same as in Example 2, and can be referred to the foregoing description, so it will not be repeated here. In one possible implementation, the following steps are included:

[0114] Optionally, when the frequency sweep mode is configured as a multi-frequency parallel sweep mode, the step of determining the sweep frequency in the step of generating a wideband disturbance command based on the grid-connected system operating status information includes: constraining the sweep frequency to a preset frequency grid; constructing a candidate frequency group based on the sweep frequency on the frequency grid, and determining the set of second-order and third-order intermodulation products of the frequency within the group; calculating the frequency difference between the product frequencies in the set of second-order and third-order intermodulation products and any sweep frequency; and determining that the candidate frequency group satisfies the intermodulation safety constraint when the frequency difference is greater than the preset intermodulation protection bandwidth.

[0115] Furthermore, in multi-frequency parallel scanning mode, to eliminate spectral leakage during multi-frequency synchronous sampling, the system predefines a base frequency resolution. The system constrains all scanning frequencies to nodes that are integer multiples of this base frequency resolution, forming a frequency grid. The following constraint equations are used:

[0116] f k =n k *df0;

[0117] In the formula, f k For the k-th scan frequency, n k df0 is a positive integer representing the corresponding frequency grid number, and df0 is the preset base frequency resolution. This constraint ensures that all scan frequencies within the subsequent observation window contain an integer number of complete signal cycles.

[0118] As an optional implementation, when multiple frequencies are combined and injected into a grid-connected system, nonlinear characteristics can generate intermodulation interference. To avoid interference, for any candidate frequency group, the system extracts each independent frequency within the group and calculates the linear combination results between the independent frequencies to determine the sets of second- and third-order intermodulation products. Specifically, the sets of second- and third-order intermodulation products include the sum frequency, difference frequency, and third-order intermodulation frequency of any two different frequencies within the group. Further, the system iterates through each product frequency in the sets of second- and third-order intermodulation products and calculates the absolute frequency difference between the product frequency and all globally defined scanning frequencies. The system sets an intermodulation protection bandwidth to provide a safety margin. The intermodulation protection bandwidth is set to a value that is at least twice the fundamental frequency resolution. When all calculated frequency differences are greater than the intermodulation protection bandwidth, the system determines that the current candidate frequency group will not cause spectral leakage pollution to the target scanning frequency point and that the candidate frequency group meets the intermodulation safety constraints.

[0119] Optionally, the step of determining the frequency group of the multi-frequency signal further includes: extracting the logarithmic link gain of each scanning frequency based on the prior parameters of the injection link; under the premise of satisfying the intermodulation safety constraints, calculating the range of logarithmic link gain within the candidate frequency group for the unassigned frequencies on the frequency grid after they are added to the candidate frequency group; assigning the unassigned frequencies to the candidate frequency group with the smallest corresponding logarithmic link gain range, determining the frequency group of the multi-frequency signal, and reducing the amplitude dynamic range of frequencies within the same group on the silicon carbide perturbation injection unit side.

[0120] Accordingly, since the amplitude response of the injection link changes with frequency, the required driving amplitude at different frequencies differs on the injection unit side. Without frequency intervention, the dynamic range of amplitude at the injection unit side for frequencies in the same group would be too large, thus increasing the signal's peak factor. To address this issue, the system introduces a link gain sensing mechanism during the grouping process. The system extracts the pre-calibrated injection link transfer function and calculates the logarithmic link gain corresponding to each scanning frequency. The specific formula for calculating the logarithmic gain is as follows:

[0121] g k =ln(|H T (w k )|);

[0122] In the formula, g k Let H be the logarithmic link gain at the k-th scan frequency, ln be the natural logarithm function, and |H| be the logarithmic link gain. T (w kThe value represents the calibrated amplitude gain of the injected link at the scanning frequency. Logarithmic operations are used because the peak factor of multi-frequency signals is sensitive to amplitude ratios, and the proportional relationship of amplitude gains can be transformed into a difference relationship through logarithmic operations. During frequency allocation, the system sorts all frequencies to be scanned in ascending order of value and checks the created candidate frequency groups one by one. For any candidate frequency group that satisfies intermodulation safety constraints, the system calculates the new logarithmic link gain range formed after adding the frequency to be allocated to that group. The specific range update formula is as follows:

[0123] D g_plus =max(g max ,g k )-min(g _min ,g _k );

[0124] In the formula, D _g_plus The range of intra-group logarithmic link gain after the frequency to be allocated is given; max is the function to find the maximum value; min is the function to find the minimum value; g _max To determine the maximum logarithmic link gain within the current candidate frequency group before inclusion, g _min To determine the minimum logarithmic link gain within the current candidate frequency group before inclusion, g _k This is the logarithmic link gain for the frequency to be allocated. This range value reflects the dispersion between the maximum and minimum amplitudes on the injected unit side within the group. Furthermore, the system defines the negative of the range value as the comprehensive score. The comprehensive score formula is as follows:

[0125] Score = -1 * D g_plus ;

[0126] In the formula, Score is the overall score, -1 is the constant multiplier, and D g_plus The system calculates the logarithmic link gain range. It compares the overall score of the frequency to be assigned among all feasible candidate frequency groups and assigns the frequency to the candidate frequency group with the highest overall score that satisfies the intermodulation safety constraints. When none of the candidate frequency groups satisfy the intermodulation safety constraints, the system creates a completely new independent frequency group for the frequency to be assigned. Through iterative allocation based on the range minimization criterion, the system outputs the final determined multi-frequency signal group.

[0127] In some alternative implementations, each candidate frequency group dynamically maintains its current maximum and minimum logarithmic gain values ​​in memory. Whenever a new frequency is assigned to the group, the system performs a real-time numerical comparison and replacement update operation. This dynamic maintenance mechanism reduces the range calculation process to constant time complexity, improving the efficiency of grouping calculations when scanning a large number of frequency points.

[0128] Example 6 further details the phase optimization and derating control mechanism in multi-frequency parallel scanning mode. In one possible implementation, it includes the following steps:

[0129] Optionally, when the frequency sweep mode is configured as a multi-frequency parallel scanning mode, in the step of generating a wideband disturbance command based on the grid-connected system operating status information, the step of determining the phase of the multi-frequency superimposed signal includes: constructing a grid-connected point domain discrete signal based on the unnormalized target amplitude at the grid connection point determined according to the background noise power spectrum in the prior parameters of the injected link and a preset initial phase, and performing time-domain clipping and phase extraction operations to update the phase at the grid connection point.

[0130] Correspondingly, in multi-frequency parallel scanning, the superposition of multiple frequency components in the time domain easily generates high peak values, leading to signal distortion or converter overmodulation. To optimize the peak factor, the system establishes a discrete signal model in the grid connection point domain. The system assigns unnormalized target amplitudes to frequencies within each group based on the background noise level at the grid connection point. Furthermore, the system obtains a preset initial phase. To accelerate the iterative convergence process, the system pre-calculates the theoretically optimal phase. Before constructing the discrete signal in the grid connection point domain based on the unnormalized target amplitude and initial phase at the grid connection point, specifically, the system calculates the theoretically optimal phase using the Schrödinger phase formula based on the ascending order of frequencies within each multi-frequency signal frequency group, and uses this as the initial phase to compress the initial time-domain peak value of the multi-frequency superimposed signal before initiating the iteration. The formula for calculating the theoretically optimal phase is as follows:

[0131] psi g_i =-3.14*i*(i-1) / M g ;

[0132] In the formula, psi g_i M represents the theoretically optimal phase of the i-th frequency within the g-th group, where i is the ascending index of that frequency within the group, and M... g Let g be the total number of frequencies contained in the g-th group, and -3.14 be a constant, i.e., a negative approximation of pi. The initial phase determined by the Schroeder phase formula can effectively compress the initial time-domain peak value of the multi-frequency superimposed signal before starting the iteration.

[0133] Furthermore, the system performs high-density time discretization sampling within a set minimum common period, constructing a grid-connected point domain discrete signal based on the unnormalized target amplitude and initial phase. The system then performs a time-domain clipping operation on this discrete signal. Specifically, the system sets a peak-shaving threshold, restricting all time-domain sampled data with absolute values ​​exceeding the threshold to the threshold value itself, while retaining their original positive and negative signs. The peak-shaving threshold can be set as a preset proportional coefficient multiple of the sum of the unnormalized target amplitudes on the grid-connected point side, with the preset proportional coefficient determined by those skilled in the art based on the target peak factor requirements. Sampled data that do not exceed the peak-shaving threshold retain their original values. After clipping, the system performs a discrete Fourier transform on the clipped discrete signal, extracting the complex phase angles corresponding to each scan frequency component, and updating these complex phase angles to the current grid-connected point side phase.

[0134] Optionally, by combining the amplitude gain and phase shift information in the prior parameters of the injection link, the updated phase at the grid connection point is mapped to the injection unit side to construct a discrete signal in the injection unit domain.

[0135] Furthermore, after obtaining the updated grid-connection point-side phase, the system needs to consider the physical impact of the injection link. The system extracts the amplitude gain and phase shift information corresponding to each scanning frequency point from the prior parameters of the injection link. The system subtracts the corresponding phase shift information from the updated grid-connection point-side phase through subtraction, mapping the phase to the injection unit side. Simultaneously, the system divides the unnormalized target amplitude at the grid-connection point by the amplitude gain to inversely deduce the reference amplitude at the injection unit side. Based on the mapped phase and the inversely derived reference amplitude, the system re-performs discretization reconstruction to construct a discrete signal in the injection unit domain that reflects the actual output waveform of the silicon carbide perturbation injection unit.

[0136] Optionally, when the discrete signal in the injected unit domain exceeds the preset upper limit of the DC bus voltage constraint, the time domain pruning and phase extraction operation of the injected domain is performed and mapped back to the grid connection point side; when the discrete signal in the injected unit domain does not exceed the upper limit of the DC bus voltage constraint, the phase of the current injected unit side remains unchanged.

[0137] Specifically, the system iterates through all time sampling points of the constructed discrete signal in the injection unit domain and detects its maximum peak absolute value. The system obtains a preset upper limit for the DC bus voltage constraint, which is obtained by multiplying the actual configured DC bus voltage value by the maximum allowable modulation ratio. When the detected maximum peak absolute value exceeds the upper limit for the DC bus voltage constraint, overmodulation risk is determined. At this time, the system performs a second time-domain clipping operation in the injection unit domain, with the peak clipping threshold set to the upper limit for the DC bus voltage constraint. After clipping, the system performs a discrete Fourier transform again to extract the new phase on the injection unit side and remaps the new phase back to the grid connection point side by adding phase shift information, completing the iterative calculation of the dual-domain coupling cycle. If the maximum peak absolute value does not exceed the upper limit, the clipping operation in this embodiment is skipped.

[0138] Optionally, by alternately performing the above-mentioned pruning and phase mapping operations between the grid connection point domain and the injection unit domain until the preset dual-domain joint convergence condition is met, the optimized grid connection point side phase and the corresponding injection unit side phase are obtained, and a broadband disturbance command is generated accordingly.

[0139] As an optional implementation, following the above logic, a closed-loop iterative cycle is established between the two physical domains. Each iteration gradually reduces the peak factor of the overall signal, allowing the signal shape to adapt to the physical constraints on both sides simultaneously. To determine when the iteration stops, the system sets a dual-domain joint convergence condition. The preset dual-domain joint convergence condition includes two sub-conditions: First, the absolute value of the change in the peak factor on the grid connection point side calculated in two adjacent iterations is less than a preset peak factor convergence threshold. In this embodiment, the peak factor convergence threshold is set to 0.01. Second, the peak value of the currently calculated discrete signal in the injection unit domain is not greater than the upper limit of the DC bus voltage constraint. Only when both of the above sub-conditions are met simultaneously does the system determine that the iteration is successful. The final output phase group is used to synthesize a multi-frequency modulated wave and is encapsulated in a wideband perturbation command for execution.

[0140] Optionally, the preset dual-domain joint convergence conditions include: the change in peak factor at the grid connection point between two adjacent iterations is less than the preset peak factor convergence threshold, and the peak value of the injected unit domain discrete signal does not exceed the upper limit of the DC bus voltage constraint; when the alternating execution of pruning and phase mapping operations reaches the maximum number of iterations and still fails to simultaneously meet the dual-domain joint convergence conditions, a degradation derating strategy is executed, the phase combination of the last iteration is extracted, and based on the peak value exceeding the limit ratio of the injected unit domain discrete signal, a global derating scaling is applied to the amplitude of the entire group of multi-frequency superimposed signals to meet the upper limit of the DC bus voltage constraint.

[0141] Furthermore, if the system fails to converge after reaching the preset maximum number of iterations due to excessively poor frequency combination characteristics or insufficient DC bus voltage, in this embodiment, the maximum number of iterations is set to 200, and the system activates a safety fallback mechanism. At this point, the system stops iterating, extracts the phase data generated in the 200th iteration, and calculates the proportional constant at the time of the current limit exceedance. The formula for calculating the derating scaling factor is as follows:

[0142] β=V SiC_lim / max val ;

[0143] In the formula, β is the global derating scaling factor, and V SiC_lim The maximum value is the preset upper limit of the DC bus voltage constraint. val This represents the maximum absolute value of the discrete signal in the injected unit domain at each discrete sampling point.

[0144] The system multiplies the target amplitude at the grid connection point for all frequencies within the entire group by this derating scaling factor. By proportionally reducing the injected power at all frequencies, the output signal at the injection unit side is limited to the physical capability range, preventing device damage and ensuring continuous execution of impedance measurement under extreme conditions.

[0145] Example 7 further details the broadband impedance refinement process based on baseline subtraction. In one possible implementation, the grid connection point voltage and current signals before and after the small-signal disturbance superposition are acquired, baseline subtraction is performed, and the pure disturbance response component is obtained, such as... Figure 4 As shown, it includes the following steps:

[0146] Optionally, within the baseline acquisition window where disturbance injection is not enabled, background frequency domain components are extracted from the grid connection point voltage and current signals based on preset Fourier extraction coefficients.

[0147] Accordingly, to eliminate the interference of background harmonics, which are prevalent in the power grid, on the response to weak disturbances, the system introduces a baseline subtraction mechanism and defines a baseline acquisition window. The starting point of this window is set before the silicon carbide disturbance injection unit begins its injection action, or during the period of silence after the injection at the previous frequency point. Within this baseline acquisition window, the system synchronously acquires and records the voltage and current signals at the grid connection point. The system uses a discrete Fourier transform algorithm to extract the background frequency domain components corresponding to the current scanning frequency. To ensure that the extracted results correspond to the complex amplitude values ​​of the one-sided Fourier coefficients, the system uses a uniform Fourier extraction coefficient, which is the ratio of 2 to the time window length. The specific formula for extracting the voltage background frequency domain components is as follows:

[0148] V base (j*w k )=(2 / T w_kIntegral(v) PCC (t)*exp(-j*w k *t),t b ,t b +T w_k );

[0149] In the formula, V base (j*w k ) represents the grid connection point voltage at frequency w k The background frequency domain component at T w_k The baseline acquisition window length is given by Integral, which is the integration operation. PCC (t) represents the instantaneous voltage at the grid connection point, exp is the natural exponential function, -j is the negative imaginary unit, and w k Let t be the angular frequency of the k-th scan point, and t be the time variable. b t represents the start time of the baseline acquisition window. b +T w_k This is the end time of the baseline acquisition window.

[0150] Optionally, within the injection measurement window after disturbance injection is enabled and transient decay has occurred, the total frequency domain component is extracted from the grid connection point voltage and current signal based on the Fourier extraction coefficients, wherein the time length of the injection measurement window is equal to that of the baseline acquisition window.

[0151] Furthermore, after baseline acquisition is completed, the silicon carbide perturbation injection unit initiates perturbation injection. The system waits for a preset transient decay time to ensure the circuit response enters a steady state. Optionally, the transient decay time can be set to a preset multiple of the time constant of the equivalent RL circuit composed of the injection transformer and filter. A typical value of the preset multiple is such that the transient component decays to below a preset percentage of the steady-state amplitude. Those skilled in the art can determine the transient decay time based on the actual injection link parameters. Further, the system activates the injection measurement window and simultaneously acquires the grid connection point voltage and current signals. Using preset Fourier extraction coefficients and extraction algorithms, the system calculates the total frequency domain component. To ensure the alignment and effectiveness of the baseline subtraction operation in the frequency domain, the time length of the injection measurement window needs to be constrained to be equal to the time length of the baseline acquisition window.

[0152] Furthermore, to eliminate spectral leakage, the system imposes physical constraints on the time window length. The time lengths of the baseline acquisition window and the injection measurement window are set to integer multiples of the reciprocal of the preset fundamental frequency resolution, ensuring that all scanning frequencies contain a complete signal cycle within the time length, thus eliminating spectral leakage effects during the frequency domain extraction process.

[0153] Wherein, the time length T is equal to that of the injection measurement window and the baseline acquisition window. w_k =m k *(2*3.14 / wk );

[0154] In the formula, T w_k To ensure that the injection measurement window and the baseline acquisition window have the same time length, m k Let m be the number of complete disturbance signal cycles contained within the observation window. k Greater than or equal to the preset minimum number of periods, where 2 * 3.14 is a constant, i.e., twice the value of pi, w k Let be the angular frequency of the k-th scanning point. In practical engineering configurations, the minimum number of cycles is usually set to no less than 5 to ensure sufficient averaging effect on random noise.

[0155] Optionally, a complex subtraction operation is performed between the total frequency domain component and the background frequency domain component to obtain the pure disturbance response component that eliminates background harmonic interference.

[0156] The system extracts the total frequency domain component within the measurement window and subtracts the background frequency domain component within the baseline window to obtain the pure disturbance response component. This is calculated using the following formula:

[0157] V p (j*w k )=V total (j*w k )-V base (j*w k );

[0158] In the formula, V p (j*w k V represents the frequency domain component of the pure disturbance voltage after baseline subtraction. total (j*w k ) represents the frequency domain component of the total voltage, V base (j*w k The background voltage frequency domain component is represented by ( ). Similarly, the pure disturbance current frequency domain component is obtained using the same subtraction operation. Through the above differential operation, the contamination of the target frequency measurement by the inherent 50Hz fundamental frequency and its integer multiples of harmonics of the power grid can be removed.

[0159] Furthermore, after calculating the equivalent port impedance, a relative stability verification needs to be performed. The original absolute threshold criterion often exhibits inconsistent constraint strength when dealing with significant differences in impedance amplitude across a wide frequency band. Therefore, this embodiment adopts a relative threshold criterion and adds a zero-impedance protection mechanism. The specific relative stability criterion formula is as follows:

[0160] abs(Z m -Z m_minus_1 ) / max(abs(Z m_minus_1 ),Z floor )≤e r ;

[0161] In the formula, abs is the absolute value function, and Z... m Z represents the result of the m-th impedance measurement. m_minus_1 This represents the impedance measurement result of the (m-1)th time, where max is the function for maximizing the value, and Z... floor e is the lower limit protection value for impedance. r This is a dimensionless relative stability threshold. When the impedance amplitude is greater than the protection value, the criterion degenerates into the condition that the relative deviation does not exceed the set proportion; when the impedance amplitude is extremely small and close to zero, the denominator is covered by the protection value to prevent calculation abnormalities or criterion failure caused by dividing by zero.

[0162] Example 8 further details the process of pre-scan calibration and cross-group shared pilot consistency verification. In one possible implementation, before the steps of acquiring the grid-connected system operating status information and the pre-calibrated injection link prior parameters, the process further includes pre-calibrating the injection link prior parameters, such as... Figure 5 As shown, it includes the following steps:

[0163] Optionally, under the steady-state operation of the grid-connected system and without disturbance injection, the grid connection point voltage is collected to estimate the broadband background noise power spectrum at each scanning frequency point.

[0164] Accordingly, to enable signal-to-noise ratio calculation and adaptive amplitude adjustment in subsequent scans, the system performs offline noise estimation before the formal frequency sweep begins. The system monitors the operating status of the energy storage converter unit, confirming that it remains at the predetermined steady-state operating point, and that the silicon carbide perturbation injection unit remains silent. In this state, the system defines a global noise estimation window and continuously acquires the grid-connected voltage sequence. By performing frequency domain transformation on the acquired discrete-time sequence, the system calculates the power spectral density of each scan frequency and its adjacent frequency bands. Based on this power spectral density, the system can deduce the effective value of the background noise at the corresponding frequency point, calculated using the following formula:

[0165] σ noise =sqrt(Integral(S vv ,d w ));

[0166] In the formula, σ noise Let S be the effective value of the background noise at the scanned frequency point, sqrt be the square root function, and Integral be the definite integral function performed in the adjacent frequency band centered at the scanned frequency point. vv To estimate the obtained broadband background noise power spectral density, d w ω is the integral variable for angular frequency.

[0167] Optionally, under no-load or light-load conditions of the grid-connected system, the silicon carbide disturbance injection unit is controlled to output calibration signals sequentially and synchronously acquire the response voltage at the grid connection point in order to extract the injection link transfer function at each scanning frequency point.

[0168] Accordingly, due to the magnetizing inductance and leakage inductance of the injection transformer and the frequency-varying characteristics of the filter, the disturbance signal will experience amplitude attenuation and phase shift during transmission. To obtain compensation parameters, the system executes a calibration procedure under no-load or light-load conditions that do not affect the main power exchange. The system controls the silicon carbide disturbance injection unit to sequentially output known sinusoidal test signals at each scanning frequency. This sinusoidal test signal has a given input amplitude and initial phase. The synchronous acquisition system extracts the complex frequency domain value of the actual response voltage formed at the grid connection point at the corresponding frequency. The specific transfer function calculation formula is as follows:

[0169] H T =V cal / A cal ;

[0170] In the formula, H T V is the complex gain of the injected link transfer function at a given scan frequency. cal A is the complex frequency domain extracted value of the grid connection point response voltage. cal Input the amplitude of the given calibration signal.

[0171] Optionally, the broadband background noise power spectrum and the injection link transfer function are used together to construct pre-calibrated injection link prior parameters.

[0172] Specifically, the calculated background noise data and complex gain data at each frequency point are structured and stored to form a lookup table or discrete data sequence. This dataset constitutes the injected link prior parameters. In subsequent online broadband frequency sweeping, the system no longer performs global frequency band scanning estimation, but instead retrieves this pre-calibrated parameter set to guide the noise equalization amplitude allocation of multi-frequency signals, link gain-aware grouping, and dual-domain iterative phase optimization.

[0173] In one aspect of this application, the step of performing impedance measurement validity and relative stability verification in conjunction with a pre-configured decision threshold specifically includes: optionally, when dividing the scanning frequency points into multiple frequency groups for batch injection measurement, embedding the same pre-configured shared pilot frequency in each frequency group.

[0174] Furthermore, since the multi-frequency parallel scanning mechanism requires dividing all wideband frequency points into multiple frequency groups for sequential execution, the entire frequency scanning process has a time span. During this period, the operating conditions of the new energy power station may drift, disrupting the physical consistency of the spliced ​​impedance spectrum. To address this issue, the system employs a pilot verification mechanism. The system selects a subset of frequencies from the total scanned frequency point set as shared pilot frequencies. These selected shared pilot frequencies are distributed across the low-frequency, mid-frequency, and high-frequency segments of the target measurement frequency band. When executing the frequency grouping algorithm, the system replicates and assigns these shared pilot frequencies to each independent frequency group. During the embedding process, the system synchronously verifies the intermodulation safety constraints after adding the shared pilot frequencies, ensuring that the newly constructed frequency groups do not generate spectral leakage.

[0175] Optionally, the pilot impedances measured at different frequency groups of the shared pilot frequency can be extracted.

[0176] Accordingly, when performing disturbance injection and response measurements for each frequency group in chronological order, the system treats the shared pilot frequency as a regular scanning frequency point, performs baseline subtraction and complex division operations to obtain the equivalent port impedance at that frequency point. Since the same shared pilot frequency is triggered for measurement in multiple frequency groups, the system records the impedance data set corresponding to that frequency in different group measurement time windows in the memory unit, and uses it as a reference sample for determining the stability of the operating condition.

[0177] Optionally, the cross-group relative deviation of the pilot impedance between different frequency groups can be calculated.

[0178] Specifically, the results of the same shared pilot frequency measured in the first frequency group and the second frequency group are extracted, and differential and normalization calculations are performed to obtain a dimensionless value reflecting the degree of deviation between the two measurement results. The specific deviation calculation formula is as follows:

[0179] Δ p =abs(Z g -Z g_prime ) / max(abs(Z g ),Z floor );

[0180] In the formula, Δ p The cross-group relative deviation is given by abs, which is the absolute value function, and Z is the absolute value function. g Z is the pilot impedance measured in the first frequency group to share the pilot frequency. g_prime The shared pilot frequency is the pilot impedance measured in the second frequency group, where max is the function for maximizing the value, and Z is the pilot impedance. floor This is the preset lower limit protection value for impedance.

[0181] Optionally, when the relative deviation between groups exceeds a preset consistency threshold, the system operating condition is determined to have drifted, and the corresponding frequency group after the drift inflection point is marked as the group to be retested based on the time node when the deviation exceeds the standard.

[0182] Further, a preset consistency threshold is obtained. In this embodiment, the consistency threshold is set to 0.10. The system iterates through the relative deviations between all pilot frequencies in adjacent or similar time pairs and compares them with the consistency threshold. When a deviation exceeding the consistency threshold is detected, the system determines that the grid topology or the operating point of the energy storage converter unit has changed during the time interval between the two measurements. The system traces the pair number that caused the deviation, finds the first frequency group that shifted, and adds it and all subsequent frequency groups to the retest mark. The system schedules the silicon carbide disturbance injection unit to re-execute the baseline acquisition and injection measurement process on the set of frequency points marked as retest groups.

[0183] In one aspect of this application, after calculating the cross-group relative deviation of the pilot impedance between different frequency groups, the method further includes: optionally, when all cross-group relative deviations do not exceed the consistency threshold, calculating the average value of the pilot impedance measured in multiple frequency groups for a given shared pilot frequency, and using the average value as the output impedance of the shared pilot frequency in the broadband impedance spectrum at the final grid connection point.

[0184] Accordingly, if the relative deviations across groups at all checkpoints are within the consistency threshold throughout the entire batch frequency sweep cycle, the system determines that the system operating condition remains stable throughout the cycle, and the splicing of multiple sets of data is legitimate. At this point, in order to further reduce random errors in the measurement process, the system extracts the impedance data from all previous measurements for the same shared pilot frequency that is included in multiple frequency groups and performs an arithmetic average operation.

[0185] Correspondingly, the output impedance Z avg =sum(Z g ) / N p ;

[0186] In the formula, sum is the summation function, and Z... g Let N be the pilot impedance measured in each frequency group. p This represents the total number of frequency groups containing the shared pilot frequency. The system writes the calculated output impedance into the final grid-connected point broadband impedance spectrum, completing the post-processing quality verification and data fusion of the impedance scan task.

[0187] It should be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the present invention will not describe the various possible combinations separately.

Claims

1. A broadband impedance scanning method for new energy sources, characterized in that, include: Obtain the operating status information of the grid-connected system and the pre-calibrated prior parameters of the injection link; Based on the grid-connected system's operating status information, a wideband disturbance command is generated, and the silicon carbide disturbance injection unit is controlled to superimpose a small signal disturbance through the injection transformer at the grid connection point according to the wideband disturbance command. Based on the prior parameters of the injected link and the broadband disturbance command, the feedforward compensation term is determined and introduced into the control loop of the energy storage converter unit that maintains steady-state grid connection, so as to counteract the closed-loop suppression effect of the energy storage converter unit on small-signal disturbances. Collect the grid connection point voltage and current signals before and after the small signal disturbance is superimposed, perform baseline subtraction processing, and obtain the pure disturbance response component; The equivalent port impedance of the grid connection point is calculated based on the pure disturbance response components, and the effectiveness and relative stability of the impedance measurement are verified by combining the pre-configured judgment threshold.

2. The method according to claim 1, characterized in that, The silicon carbide disturbance injection unit is connected to the primary winding of the injection transformer via a second-order low-pass filter; Among them, the cutoff frequency of the second-order low-pass filter is greater than the highest scanning frequency of the small signal disturbance, and less than the preset ratio of the switching frequency of the silicon carbide disturbance injection unit. The ratio of the power frequency voltage drop generated by the equivalent leakage inductance of the secondary side of the injected transformer under the rated current of the main power channel to the rated voltage at the grid connection point is less than the preset voltage drop ratio threshold.

3. The method according to claim 1, characterized in that, The steps of determining the feedforward compensation term based on the injection link prior parameters and broadband disturbance commands, and introducing the feedforward compensation term into the control loop of the energy storage converter unit that maintains steady-state grid connection, include: Based on the injection link transfer function in the prior parameters of the injection link and the feedforward channel transfer function of the energy storage converter, a comprehensive calculation is performed to obtain the frequency adaptive equivalent conversion coefficient that includes amplitude matching and phase alignment characteristics. Based on the upper limit frequency of the effective feedforward bandwidth of the energy storage converter, the scanning frequency points are divided into a feedforward compensable subset and a residual subset. For the frequency points within the feedforward compensable subset, a feedforward compensation term is constructed based on the frequency adaptive equivalent conversion coefficient and the wideband disturbance command, and the feedforward compensation term is subtracted from the original voltage feedforward reference of the energy storage converter unit.

4. The method according to claim 1, characterized in that, In the step of generating broadband disturbance commands based on grid-connected system operating status information, the generated broadband disturbance commands include frequency sweep mode configuration, which includes single-frequency point-by-point scanning mode and multi-frequency parallel scanning mode.

5. The method according to claim 4, characterized in that, When the frequency sweep mode is configured as multi-frequency parallel sweep mode, the step of determining the sweep frequency in the step of generating wideband disturbance commands based on grid-connected system operating status information includes: Constrain the scanning frequency to a preset frequency grid; Candidate frequency groups are constructed based on the scanning frequencies on the frequency grid, and the sets of second- and third-order intermodulation products of the frequencies within the group are determined. Calculate the frequency difference between the product frequencies in the sets of second- and third-order intermodulation products and any scanning frequency; When the frequency difference is greater than the preset intermodulation protection bandwidth, the candidate frequency group is determined to meet the intermodulation safety constraints.

6. The method according to claim 1, characterized in that, The steps for acquiring grid-connected point voltage and current signals before and after small-signal disturbance superposition, performing baseline subtraction processing, and obtaining pure disturbance response components include: Within the baseline acquisition window without disturbance injection enabled, background frequency domain components are extracted from the grid connection point voltage and current signals based on preset Fourier extraction coefficients. Within the injection measurement window after disturbance injection is enabled and transient decay has occurred, the total frequency domain component is extracted from the grid connection point voltage and current signals based on the Fourier extraction coefficients, wherein the time length of the injection measurement window is equal to that of the baseline acquisition window. By performing a complex subtraction operation between the total frequency domain component and the background frequency domain component, the pure disturbance response component that eliminates background harmonic interference is obtained.

7. The method according to claim 1, characterized in that, Before the steps of obtaining the grid-connected system operating status information and the pre-calibrated injection link prior parameters, the process also includes a step of pre-calibrating the injection link prior parameters, specifically including: Under the steady-state operation of the grid-connected system and without disturbance injection, the grid connection point voltage is collected to estimate the broadband background noise power spectrum at each scanning frequency point; Under no-load or light-load conditions of the grid-connected system, the silicon carbide disturbance injection unit is controlled to output calibration signals sequentially and synchronously collect the response voltage at the grid connection point in order to extract the injection link transfer function at each scanning frequency point. The broadband background noise power spectrum and the injection link transfer function are used together to construct the pre-calibrated injection link prior parameters.

8. The method according to claim 1, characterized in that, The steps for verifying the validity and relative stability of impedance measurements by combining pre-configured decision thresholds include: When dividing the scanning frequency points into multiple frequency groups for batch injection measurement, the same shared pilot frequency is embedded in each frequency group; Extract the pilot impedance measured at different frequency groups for the shared pilot frequency; Calculate the cross-group relative deviation of the pilot impedance between different frequency groups; When the relative deviation between groups exceeds the preset consistency threshold, the system operating condition is determined to have drifted, and the corresponding frequency group after the drift inflection point is marked as the group to be retested based on the time node when the deviation exceeds the standard.

9. The method according to claim 8, characterized in that, After calculating the cross-group relative deviation of the pilot impedance between different frequency groups, the following steps are also included: When all cross-group relative deviations do not exceed the consistency threshold, for a given shared pilot frequency, calculate the average value of the pilot impedance measured in multiple frequency groups, and use the average value as the output impedance of the shared pilot frequency in the broadband impedance spectrum at the final grid connection point.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein the program, when executed, performs the method of any one of claims 1 to 9.

Images