An energy-saving photovoltaic inverter control system and control method

Abstract

The application discloses an energy-saving photovoltaic inverter control system and method, and particularly relates to the technical field of power electronic conversion control, and is used for solving the problem that the existing periodic MPPT disturbance is easily interfered and coupled with external fluctuations under fluctuating light conditions, resulting in dynamic tracking energy loss; by monitoring photovoltaic array output power and DC bus voltage ripple, analyzing the cooperative change characteristics to identify the interference phenomenon, extracting the external fluctuation main frequency and tracking the phase relationship between the control disturbance, and then evaluating the interference coupling risk level, and dynamically adjusting the disturbance period according to the interference coupling risk level, the control rhythm is actively avoided from the external fluctuation main frequency, the harmful interference between the control and the external environmental fluctuation is accurately identified and effectively inhibited, so that the energy capture efficiency of the photovoltaic system under fluctuating light conditions is improved.

Description

Technical Field

[0001] This invention relates to the field of power electronic conversion control technology, and in particular to an energy-saving photovoltaic inverter control system and control method. Background Technology

[0002] As the core power conversion device in a photovoltaic (PV) power generation system, the photovoltaic (PV) inverter's core function is to adjust its operating point in real time through maximum power point tracking (MPPT) control to extract as much electrical energy as possible from the PV array and convert it into AC power that meets grid connection requirements. In actual operating environments, due to factors such as cloud drift, building obstruction, or vegetation shading, the solar intensity received by the PV array is not constant but exhibits dynamic fluctuations. To ensure effective tracking of the maximum power point under fluctuating conditions, the inverter typically employs a periodic MPPT algorithm, such as the perturbation-observation method or the incremental conductance method. This involves periodically applying small perturbations to the output voltage or current of the PV array with a fixed sampling and control cycle, and determining the direction of subsequent perturbations based on the power change trend, thereby gradually approaching the maximum power point.

[0003] However, existing periodic MPPT control methods have drawbacks: when the fluctuation frequency of external illumination changes is close to or has a specific relationship with the inherent disturbance period of the MPPT algorithm, the two will be coupled and interfere with each other. This results in the power change signal detected by the MPPT algorithm being the superposition of the natural fluctuation of illumination and the active control disturbance. This causes a systematic deviation in the algorithm's judgment of the power change trend. As a result, the MPPT algorithm will continuously track a target whose phase lags behind the true maximum power point and cannot stably lock onto the vicinity of the optimal operating point. Instead, it will perform unnecessary and inefficient back-and-forth searches around it, resulting in continuous dynamic tracking energy loss and reducing the overall energy capture efficiency of the photovoltaic system under fluctuating illumination conditions. Summary of the Invention

[0004] This invention addresses the technical problems existing in the prior art by providing an energy-saving photovoltaic inverter control system and control method.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0006] An energy-saving photovoltaic inverter control method includes:

[0007] S1. Monitor and acquire the output power of the photovoltaic array and the DC bus voltage of the photovoltaic inverter to form an output power fluctuation sequence and a DC bus voltage ripple sequence;

[0008] S2. Perform correlation analysis on the output power fluctuation sequence and the DC bus voltage ripple sequence to determine whether there are cooperative change characteristics that characterize the periodic disturbance input.

[0009] S3. When there are cooperative change characteristics, analyze the frequency distribution of the output power fluctuation sequence and extract the main fluctuation frequency;

[0010] S4. Analyze the real-time phase difference between the fluctuation period corresponding to the main fluctuation frequency and the disturbance period of the maximum power point tracking control, and track the change trajectory of the real-time phase difference in the continuous disturbance period.

[0011] S5. Based on the energy accumulation characteristics of the output power fluctuation sequence within the frequency sideband range of the disturbance period and combined with the change trajectory, assess the risk level of interference coupling between the main fluctuation frequency and the disturbance period.

[0012] S6. Adjust the disturbance period of the maximum power point tracking control according to the risk level so that the adjusted disturbance period is offset from the main fluctuation frequency.

[0013] Furthermore, S1 includes:

[0014] The output current and output voltage of the photovoltaic array are collected synchronously with a preset synchronous sampling period, and the DC bus voltage is also collected synchronously.

[0015] The instantaneous power value is calculated based on the synchronously acquired output current and output voltage.

[0016] The instantaneous power value is filtered to form an output power fluctuation sequence;

[0017] The synchronously acquired DC bus voltage is processed by extracting the AC component to form a DC bus voltage ripple sequence.

[0018] Furthermore, S2 includes:

[0019] Time-align the output power fluctuation sequence with the DC bus voltage ripple sequence;

[0020] Within multiple consecutive sliding time windows, the cross-correlation coefficients between the time-aligned output power fluctuation sequence and the DC bus voltage ripple sequence are calculated respectively.

[0021] Determine whether the calculated cross-correlation coefficient continuously exceeds a preset correlation threshold within a continuous sliding time window;

[0022] When the cross-correlation coefficient continuously exceeds the preset correlation threshold within a continuous sliding time window, it is determined that there is a cooperative change feature that characterizes the periodic perturbation input.

[0023] Furthermore, S3 includes:

[0024] Determine the time period corresponding to the cooperative change characteristics in the output power fluctuation sequence;

[0025] The power spectral density is obtained by performing spectral estimation on the output power fluctuation sequence over a time period.

[0026] Identify local peaks in the power spectral density and determine the frequency and amplitude corresponding to each local peak;

[0027] The frequency corresponding to the highest peak whose amplitude exceeds the preset amplitude threshold is determined as the main fluctuation frequency.

[0028] Furthermore, S4 includes:

[0029] Calculate the corresponding fluctuation period based on the main fluctuation frequency;

[0030] Within each disturbance cycle of maximum power point tracking control, the real-time phase offset of the fluctuation cycle relative to the start time of the current disturbance cycle is calculated based on the zero-crossing point or a specific phase point of the output power fluctuation sequence.

[0031] Record the real-time phase offset calculated over multiple consecutive disturbance cycles;

[0032] The real-time phase offsets recorded in multiple consecutive disturbance cycles are connected in chronological order to form the trajectory of the real-time phase difference change in consecutive disturbance cycles.

[0033] Furthermore, S5 includes:

[0034] The frequency sideband range is determined by using the disturbance frequency of maximum power point tracking control as the center frequency and the main fluctuation frequency as the sideband offset.

[0035] The power spectrum integral value of the output power fluctuation sequence in the frequency sideband range is calculated to the total power spectrum integral value in the entire effective analysis frequency band, and the sideband energy amplification factor characterizing the energy accumulation feature is obtained.

[0036] Analyze the change trajectory to determine whether the real-time phase difference exhibits quasi-steady-state characteristics with finite fluctuations around a certain fixed value;

[0037] If the real-time phase difference exhibits quasi-steady-state characteristics, the system is determined to have entered a stable phase-locked state. Based on the magnitude of the sideband energy amplification factor, the risk level of interference coupling characterized by energy accumulation in the stable phase-locked state is assessed.

[0038] If the real-time phase difference does not exhibit quasi-steady-state characteristics, the risk level of dynamic interference coupling is assessed based on the magnitude of the sideband energy amplification factor and the changing trend of the real-time phase difference.

[0039] Furthermore, the sideband energy amplification factor, which characterizes the energy accumulation feature, is obtained in the following way:

[0040] Using the disturbance frequency of maximum power point tracking control as the center frequency and the main fluctuation frequency as the sideband offset, the upper and lower boundary frequencies of the frequency sideband are determined.

[0041] Calculate the power spectrum integral value of the output power fluctuation sequence in the frequency range between the lower boundary frequency and the upper boundary frequency, and use it as the sideband energy;

[0042] Calculate the total power spectrum integral value of the output power fluctuation sequence over the entire effective analysis frequency band;

[0043] The ratio of the sideband energy to the integral value of the total power spectrum is used as the sideband energy amplification factor.

[0044] Furthermore, the trajectory of change is analyzed to determine whether the real-time phase difference exhibits quasi-steady-state characteristics of finite fluctuations around a fixed value, including:

[0045] Calculate the standard deviation of the trajectory formed by the real-time phase difference over multiple consecutive disturbance cycles;

[0046] The average value of the absolute values ​​of the differences between adjacent real-time phase differences in the changing trajectory is used as the average rate of change.

[0047] When the standard deviation is less than a preset first threshold and the average rate of change is less than a preset second threshold, the real-time phase difference is determined to exhibit quasi-steady-state characteristics.

[0048] Furthermore, S6 includes:

[0049] Based on the risk level, select a disturbance cycle candidate value that is different from the current disturbance cycle from a number of preset disturbance cycle candidate values ​​as the target disturbance cycle;

[0050] Calculate the absolute value of the difference between the frequency corresponding to the target disturbance period and the main fluctuation frequency;

[0051] When the absolute value of the difference is less than the preset safe frequency interval, the step size is adjusted according to the preset disturbance period to increase the absolute value of the difference to the target direction, and the target disturbance period is recalculated.

[0052] The disturbance period of the maximum power point tracking control is updated to the final determined target disturbance period.

[0053] On the other hand, the present invention provides an energy-saving photovoltaic inverter control system, comprising:

[0054] The sequence forming module is used to monitor and acquire the output power of the photovoltaic array and the DC bus voltage of the photovoltaic inverter, and to form the output power fluctuation sequence and the DC bus voltage ripple sequence.

[0055] The feature judgment module is used to perform correlation analysis on the output power fluctuation sequence and the DC bus voltage ripple sequence to determine whether there are cooperative change features that characterize periodic disturbance inputs.

[0056] The frequency extraction module is used to analyze the frequency distribution of the output power fluctuation sequence and extract the main fluctuation frequency when there are cooperative variation characteristics.

[0057] The trajectory tracking module is used to analyze the real-time phase difference between the fluctuation period corresponding to the main fluctuation frequency and the disturbance period of the maximum power point tracking control, and to track the trajectory of the real-time phase difference within the continuous disturbance period.

[0058] The risk level assessment module is used to assess the risk level of interference coupling between the main fluctuation frequency and the disturbance period based on the energy accumulation characteristics of the output power fluctuation sequence in the frequency sideband range of the disturbance period and the change trajectory.

[0059] The level adjustment module is used to adjust the disturbance period of the maximum power point tracking control according to the risk level, so that the adjusted disturbance period is staggered from the main fluctuation frequency.

[0060] The beneficial effects of this invention are:

[0061] 1. By monitoring the coordinated changes in photovoltaic array output power fluctuations and DC bus voltage ripple in real time, it is possible to accurately identify whether there is harmful interference coupling between external light fluctuations and MPPT periodic disturbances. This breaks through the limitations of traditional methods that rely solely on a single power signal for decision-making. Through cross-validation of the two related physical quantities, it effectively distinguishes whether the power change originates from natural fluctuations in external light or from the control disturbance itself, thus providing a reliable criterion for subsequent intervention. This enables the system to identify the root cause of interference in complex fluctuation environments, avoiding misjudgment at the source.

[0062] 2. After identifying interference risks, the system dynamically assesses the risk level and adaptively adjusts the MPPT disturbance cycle accordingly. This enables the controller's internal operating cycle to actively avoid the main frequency of external fluctuations, achieving real-time matching between control parameters and the operating environment. This breaks the cycle of continuous frequency beating and energy loss that may be caused by the original fixed-cycle disturbance mode, allowing the photovoltaic inverter to maintain stable and efficient maximum power point tracking under fluctuating illumination conditions, and significantly improving the overall energy capture efficiency of the system. Attached Figure Description

[0063] Figure 1 This is a flowchart of an energy-saving photovoltaic inverter control method according to the present invention;

[0064] Figure 2 This is a schematic diagram of the structure of an energy-saving photovoltaic inverter control system according to the present invention. Detailed Implementation

[0065] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0066] Example 1: Figure 1 This invention provides an energy-saving photovoltaic inverter control method, comprising:

[0067] S1. Monitor and acquire the output power of the photovoltaic array and the DC bus voltage of the photovoltaic inverter to form an output power fluctuation sequence and a DC bus voltage ripple sequence;

[0068] S2. Perform correlation analysis on the output power fluctuation sequence and the DC bus voltage ripple sequence to determine whether there are cooperative change characteristics that characterize the periodic disturbance input.

[0069] S3. When there are cooperative change characteristics, analyze the frequency distribution of the output power fluctuation sequence and extract the main fluctuation frequency;

[0070] S4. Analyze the real-time phase difference between the fluctuation period corresponding to the main fluctuation frequency and the disturbance period of the maximum power point tracking control, and track the trajectory of the real-time phase difference in the continuous disturbance period.

[0071] S5. Based on the energy accumulation characteristics of the output power fluctuation sequence within the frequency sideband range of the disturbance period and combined with the change trajectory, assess the risk level of interference coupling between the main fluctuation frequency and the disturbance period.

[0072] S6. Adjust the disturbance period of the maximum power point tracking control according to the risk level so that the adjusted disturbance period is offset from the main fluctuation frequency.

[0073] S1. Monitor and acquire the output power of the photovoltaic array and the DC bus voltage of the photovoltaic inverter to form an output power fluctuation sequence and a DC bus voltage ripple sequence. The specific implementation is as follows:

[0074] The output power of the photovoltaic array and the DC bus voltage of the photovoltaic inverter are monitored and acquired to form an output power fluctuation sequence and a DC bus voltage ripple sequence. This process is performed with a preset synchronous sampling period. The synchronous sampling period is set based on the disturbance period of the maximum power point tracking control, and its purpose is to ensure effective sampling of the dynamic process caused by the control disturbance. Specifically, the synchronous sampling period should be much smaller than the disturbance period, usually set to one-tenth to one-twentieth of the disturbance period. For example, if the disturbance period of the maximum power point tracking control is 0.1 seconds, the synchronous sampling period can be set to 0.005 seconds. This setting is based on the sampling theorem to ensure that the signal components corresponding to the disturbance period can be recovered without distortion. In actual implementation, a periodic interrupt signal is generated by a timer inside the controller. In each interrupt service routine, the analog-to-digital conversion operation of the output current and output voltage of the photovoltaic array and the DC bus voltage of the photovoltaic inverter is triggered synchronously, thereby ensuring that the data of these three physical quantities are collected simultaneously at each sampling moment.

[0075] The instantaneous power value is calculated based on the synchronously acquired output current and output voltage. The method for calculating the instantaneous power value is to multiply the output voltage and output current sample values ​​belonging to the same sampling time. For example, at a specific sampling time, if the acquired output voltage is 300 volts and the acquired output current is 10 amps, then the calculated instantaneous power value at that time is 300 × 10 = 3000 watts. This multiplication operation is performed on each pair of sampling points arranged in chronological order, thereby generating a sequence of instantaneous power values ​​equal to the number of sampling points. The physical meaning of each data point in this sequence is the output power of the photovoltaic array at that sampling time, and its unit is watts. This instantaneous power value sequence contains all the information about the output power of the photovoltaic array, and its fluctuations reflect the combined effects of various factors such as changes in external illumination, active disturbances of the maximum power point tracking controller, and circuit switching noise.

[0076] The instantaneous power values ​​are filtered to form the output power fluctuation sequence. The purpose of filtering is to separate the low-frequency fluctuation components, mainly caused by the maximum power point tracking (MPPT) disturbance and the gradual change in external illumination, from the instantaneous power value sequence, while suppressing high-frequency switching noise and other high-frequency interference. Filtering is implemented using a digital low-pass filter. The key design parameter of the digital low-pass filter is its cutoff frequency. This cutoff frequency needs to be set higher than the main frequency components of interest to the MPPT disturbance. For example, the cutoff frequency can be set to five times the MPPT disturbance frequency. If the disturbance frequency is 10 Hz, the cutoff frequency of the low-pass filter can be set to 50 Hz. The filter type and order are selected based on the trade-off between transition band steepness and computational complexity; for example, a fourth-order Butterworth low-pass filter can be chosen. The instantaneous power value sequence calculated above is used as input and fed into the designed digital low-pass filter for filtering. The output sequence of the filter is the output power fluctuation sequence. This sequence retains the low-frequency fluctuation characteristics of the original power signal, and its data points correspond one-to-one with the original sampling times.

[0077] The synchronously acquired DC bus voltage undergoes AC component extraction processing to form a DC bus voltage ripple sequence. The purpose of AC component extraction is to remove the DC component from the DC bus voltage, thus obtaining the AC ripple portion reflecting voltage fluctuations. One implementation method is to use a digital high-pass filter. The cutoff frequency of the digital high-pass filter needs to be set below the maximum power point tracking control (MPPT) disturbance frequency to ensure that the ripple component corresponding to the disturbance frequency can pass through. For example, the cutoff frequency of the high-pass filter can be set to half of the MPPT disturbance frequency. If the disturbance frequency is 10 Hz, the cutoff frequency of the high-pass filter can be set to 5 Hz. The original DC bus voltage sequence obtained synchronously is input into this high-pass filter, and its output sequence is the DC bus voltage ripple sequence. Another equivalent implementation method is to use moving average subtraction. First, a moving average is calculated on the original DC bus voltage sequence to estimate its DC component. The time length of the moving average window should be greater than the MPPT disturbance period; for example, the window length can be set to twice the disturbance period. In the specific calculation, for each point in the sequence, the arithmetic mean of itself and several adjacent points is calculated as the estimated value of the DC component at that point. Then, the moving average value corresponding to each point in the original DC bus voltage sequence is subtracted, and the resulting difference sequence is the DC bus voltage ripple sequence. Regardless of the specific method used, the final DC bus voltage ripple sequence eliminates stable DC bias, highlights voltage fluctuations caused by control disturbances and system dynamics, and each data point is strictly synchronized in time with the data points of the output power fluctuation sequence.

[0078] S2. Perform correlation analysis on the output power fluctuation sequence and the DC bus voltage ripple sequence to determine whether there are cooperative variation characteristics representing periodic disturbance inputs. Specifically, this is implemented as follows:

[0079] A correlation analysis is performed on the output power fluctuation sequence and the DC bus voltage ripple sequence to determine whether there are cooperative variation characteristics representing periodic disturbance inputs. First, the output power fluctuation sequence and the DC bus voltage ripple sequence are time-aligned. Time alignment is necessary because although the two sequences are obtained with the same synchronous sampling period in step S1, in the actual data processing flow, slight delays in data buffer management, processing task scheduling, or acquisition channels may cause misalignment in the time index between the two sequences. Time alignment is accomplished by comparing the timestamps or sampling indices of the two sequences. One specific implementation assumes that the two sequences have the same and known sampling period, and alignment is achieved by finding the index offset that maximizes the cross-correlation value of a reference segment at the beginning of the two sequences. A more direct method is to stamp each sampling point with the same high-precision timer timestamp during the hardware acquisition stage, and then directly align the data points based on the matching timestamps. In this embodiment, a hypothetical alignment method based on a common sampling index is adopted, that is, it is directly assumed that the output power fluctuation sequence data points with the same sequence index number generated in step S1 and the DC bus voltage ripple sequence data points correspond to the exact same physical time. This operation ensures that the subsequent analysis uses two sets of data that are strictly synchronized in time.

[0080] Within multiple consecutive sliding time windows, the cross-correlation coefficients between the time-aligned output power fluctuation sequence and the DC bus voltage ripple sequence are calculated. The sliding time window is a fixed-length data interval that slides backward with a fixed step size from the starting point of the aligned long sequence. The window length is a crucial parameter, set sufficiently to capture any potential periodic disturbance patterns. The window length can be set based on the disturbance period of the maximum power point tracking control; for example, it can be set to the number of data points corresponding to twice the disturbance period. If the disturbance period is 0.1 seconds and the synchronous sampling period is 0.005 seconds, then one disturbance period corresponds to 20 sampling points, and the window length can be set to 40 sampling points. The sliding window's movement step size can be set to 1 sampling point to achieve high-resolution continuous analysis. For each defined window position, the corresponding output power fluctuation sequence segment and DC bus voltage ripple sequence segment within that window are extracted. The cross-correlation coefficient between these two segments is calculated. The calculation of the cross-correlation coefficient involves the following core steps: First, calculate the arithmetic mean of all data points in each of the two sub-segments to obtain their respective means; second, subtract the mean of each sub-segment from each data point in that sub-segment to obtain two sub-segment sequences with zero means; then, calculate the sum of the products of corresponding data points in these two zero-mean sequences; next, calculate the standard deviation of each of the two zero-mean sequences; finally, divide the sum of the previously calculated products by the product of the following three terms: the standard deviation of the first zero-mean sequence, the standard deviation of the second zero-mean sequence, and the square root of the number of data points within the window. The cross-correlation coefficient obtained through this series of calculations is a dimensionless value between -1 and +1. The absolute value of the value represents the strength of the linear correlation between the two sequences within that window time, and the positive or negative sign represents the consistency of the direction of change. Repeating the above calculations for each sliding window yields a sequence of cross-correlation coefficients arranged in window order.

[0081] The determination process assesses whether the calculated cross-correlation coefficient consistently exceeds a preset correlation threshold within a continuous sliding time window. This preset correlation threshold is used to statistically determine the significance of the correlation, and its setting requires a balance between detection sensitivity and noise immunity. This correlation threshold can be determined by analyzing historical or simulation data. For example, in a large amount of known stable operating condition data without strong periodic disturbance coupling, the cross-correlation coefficient between the output power fluctuation sequence and the DC bus voltage ripple sequence can be calculated, and their distribution can be statistically analyzed. The high percentile value of this distribution can be set as the correlation threshold; for example, the 97.5th percentile could be set as the correlation threshold, which might be around 0.65. Another setting basis is based on experience, such as setting the correlation threshold between 0.6 and 0.75. The determination of a sustained exceedance requires not only that the cross-correlation value is greater than the correlation threshold, but also that this situation occurs consecutively within multiple adjacent sliding windows. The number of consecutive windows is another determination parameter; for example, it can be set to require three or more consecutive windows to meet the condition. This consecutiveness parameter is set to avoid single-point misjudgments caused by accidental noise spikes or transient interference, thereby improving the stability and reliability of the determination conclusion. Therefore, the judgment logic is that in the cross-correlation number sequence, there must exist a continuous subsequence in which each cross-correlation value is greater than the preset correlation threshold, and the length of the continuous subsequence is not less than the preset number of continuous windows.

[0082] When the cross-correlation coefficient consistently exceeds a preset correlation threshold within a continuous sliding time window, a cooperative variation characteristic representing periodic disturbance input is determined to exist. This determination is a classification decision based on statistical rules. It indicates that, over a considerable continuous period, the low-frequency fluctuation pattern of the photovoltaic array output power and the ripple pattern of the DC bus voltage exhibit a highly similar and stable synchronous variation pattern. This stable and high-intensity correlation cannot be explained by inherent system characteristics or random noise; its physical root points to a common periodic excitation source acting simultaneously on the photovoltaic array output port and the DC bus circuit. In the context of a maximum power point tracking control system, this excitation source is precisely the voltage or current disturbance periodically injected by the controller to search for the maximum power point. Therefore, this determination technically characterizes that the current operating state of the system is significantly affected by periodic beat frequencies or oscillation modes caused by internal control disturbances. This conclusion is a necessary condition and key basis for triggering subsequent frequency analysis, phase tracking, and risk assessment steps.

[0083] S3. When there are cooperative variation characteristics, analyze the frequency distribution of the output power fluctuation sequence and extract the main fluctuation frequency. The specific implementation is as follows:

[0084] When step S2 determines the existence of cooperative change characteristics, the frequency distribution of the output power fluctuation sequence is analyzed and the main fluctuation frequency is extracted. First, it is necessary to determine the time period corresponding to the cooperative change characteristics in the output power fluctuation sequence. This time period determination directly depends on the analysis results of step S2. Specifically, in step S2, the process of calculating the cross-correlation coefficient through a sliding window and determining if it continuously exceeds a preset correlation threshold has identified a continuous window sequence. The time period corresponding to the cooperative change characteristics is the sum of the original output power fluctuation sequence data covered by these continuous determination windows. The implementation method is to record the starting index position of the first sliding window that meets the conditions in step S2 within the output power fluctuation sequence, and the ending index position of the last sliding window that meets the conditions. For example, if the first determination window covers data points from index 1001 to 1040, the window sliding step size is 1, and 5 windows are consecutively determined, then the last determination window covers data points from index 1005 to 1044. Therefore, the data corresponding to this time period is all the data in the output power fluctuation sequence from index point 1001 to index point 1044. This continuous subsequence is extracted and used as input data for subsequent frequency analysis. This operation ensures that the data segment being analyzed is precisely the period in which the cooperative variation characteristics are most significant.

[0085] The power spectral density is obtained by spectral estimation of the extracted output power fluctuation sequence segments within the time period. The spectral estimation employs a periodogram method based on Fast Fourier Transform (FFT). First, the input segment data is preprocessed to eliminate DC offset and end-effects. Preprocessing includes detrending: linear fitting is performed on the data segments to obtain a straight line representing their trend; then, the value corresponding to this line is subtracted from the original data to obtain zero-mean fluctuation data. Next, windowing is applied to reduce spectral leakage. The Hanning window function can be selected, and the window function sequence is multiplied point-by-point with the detrended data sequence. The number of points for the FFT is determined by padding with zeros to the original data length, typically up to a power of 2. For example, if the original data length is 44 points, it can be padded to 64 points to improve frequency display resolution. Then, the windowed and zero-padded sequence is subjected to the FFT to obtain the complex spectrum. The method for calculating power spectral density is as follows: take the modulus of the complex number corresponding to each frequency point in the Fast Fourier Transform (FFT) result, square the modulus, divide it by the product of the total energy of the window function used and the number of FFT points, and normalize by the sampling frequency as needed. The final result is a discrete power spectral density sequence, where the horizontal axis represents frequency in Hertz (Hz), and the frequency interval (i.e., frequency resolution) is obtained by dividing the sampling frequency by the number of FFT points. The vertical axis represents the estimated power spectral density value, which can be in watts per Hertz (W / Hz), characterizing the distribution intensity of signal power at different frequency points.

[0086] Local peaks in the power spectral density are identified, and the frequency and amplitude corresponding to each local peak are determined. A local peak is a frequency point in the power spectral density sequence where the value is greater than the values ​​of its preceding and following frequencies. The identification process is accomplished by traversing all internal points in the power spectral density sequence except for the first and last points. To eliminate spurious peaks caused by small fluctuations or noise, a minimum peak prominence condition can be set. The minimum peak prominence means that a peak must be higher than the lowest value of five frequencies on either side of it by a certain minimum difference. This minimum difference can be set based on a proportion of the overall dynamic range of the power spectral density, for example, it can be set to 2% of the maximum value of the entire power spectral density sequence. For each point that satisfies both the local maximum condition and the minimum peak prominence condition, its corresponding frequency value is recorded. This frequency value is calculated based on the point's index position in the sequence, frequency resolution, and fundamental frequency. For example, if the fundamental frequency is 0 Hz, the frequency resolution is 0.5 Hz, and the peak point index is 20, then its corresponding frequency is 10 Hz. Simultaneously, the power spectral density value corresponding to that point is recorded as the amplitude of the peak. Through this step, a series of candidate peaks can be obtained, each peak being characterized by both frequency and amplitude values.

[0087] The frequency corresponding to the highest peak whose amplitude exceeds a preset amplitude threshold is determined as the main fluctuation frequency. The preset amplitude threshold is used to filter out frequency components with significant energy. The preset amplitude threshold can be set based on several criteria. One criterion is an absolute energy threshold, such as 0.1 watts per hertz (W / Hertz), with peaks below this value considered background noise. Another more adaptive criterion is a relative proportion, such as setting the preset amplitude threshold to 1.5 times the median amplitude of all identified local peaks, or to 10% of the maximum amplitude of the entire power spectral density sequence. For example, if the maximum amplitude of the power spectral density is 50 W / Hertz, the preset amplitude threshold can be set to 5 W / Hertz. In implementation, firstly, the amplitudes of all identified local peaks are compared with the preset amplitude threshold, filtering out peaks with amplitudes greater than the preset amplitude threshold to form a set of significant peaks. Then, within this set of significant peaks, the peak with the largest amplitude value is searched. If multiple maximum values ​​with equal amplitudes exist, the one with the lower frequency can be selected first, or a comprehensive judgment can be made by referring to the temporal periodicity of the data within the coordinated change characteristic period in step S2. The frequency value corresponding to this finally selected significant peak is determined as the main fluctuation frequency. The main fluctuation frequency is a specific value, such as 10.5 Hz. It represents the frequency of the most important and energy-concentrated periodic component of the output power fluctuation within the characteristic period being analyzed. It is the core input parameter for period calculation and phase analysis in subsequent steps.

[0088] S4. Analyze the real-time phase difference between the fluctuation period corresponding to the main fluctuation frequency and the disturbance period of the maximum power point tracking control, and track the trajectory of the real-time phase difference within the continuous disturbance period. The specific implementation is as follows:

[0089] The real-time phase difference between the fluctuation period corresponding to the dominant fluctuation frequency and the disturbance period of the maximum power point tracking control is analyzed, and the trajectory of the real-time phase difference within continuous disturbance periods is tracked. First, the fluctuation period is calculated based on the dominant fluctuation frequency. The dominant fluctuation frequency is a specific value output in step S3, for example, 5.2 Hz. The fluctuation period is the reciprocal of this frequency, calculated by dividing the number 1 by the dominant fluctuation frequency. Dimensionally, frequency is measured in Hertz (Hz), i.e., cycles per second; therefore, the period is measured in seconds. For example, when the dominant fluctuation frequency is 5.2 Hz, the fluctuation period is approximately 1 / 5.2 ≈ 0.1923 seconds. This fluctuation period numerically represents the theoretical time required for the power fluctuation caused by external illumination disturbance to complete one full cycle. The disturbance period of the maximum power point tracking control is a known control parameter, for example, 0.2 seconds, representing the time interval between the controller's periodic actions of actively adjusting the power point. These two period values ​​serve as the basis for subsequent phase difference calculations.

[0090] Within each disturbance cycle of the maximum power point tracking (MPPT) control, the real-time phase offset of the fluctuation cycle relative to the start of the current disturbance cycle is calculated, using the zero-crossing point or a specific phase point of the output power fluctuation sequence as a reference. This calculation first requires determining two key time reference points. The first reference point is the start of the current disturbance cycle. This moment can be explicitly marked by the MPPT controller in its internal logic, for example, by generating a timestamp signal at the start of the control cycle, or by estimating it based on a known, fixed disturbance cycle length and a global timer. The second reference point is the phase characteristic point in the output power fluctuation sequence that represents the start of a fluctuation cycle. A zero-crossing point is a commonly used characteristic point, defined as the moment when the sequence value changes from positive to negative or vice versa. The method for detecting zero-crossing points involves iterating through the data points of the output power fluctuation sequence, searching for index positions that satisfy the following conditions: the data value at this position multiplied by the data value at the previous position is less than or equal to zero; the data value at this position is less than or equal to a small threshold close to zero to handle fluctuations near zero; and the absolute value of the data value at this position is less than the absolute value of the data value at the previous position to ensure that a true zero-crossing is captured, rather than a small fluctuation near a peak. Specific phase points can also be local maxima (peaks) or local minima (troughs) of the sequence. After determining the start time Tdiststart of the current disturbance period, the first zero-crossing point or specific phase point that meets the conditions is found in the output power fluctuation sequence data that is later than Tdiststart, and its time is denoted as Twaveref. The formula for calculating the real-time phase offset is as follows: First, calculate the time difference DeltaT = Twaveref - Tdiststart. Then, divide this time difference by the fluctuation period determined in step S3, denoted as Twave. That is, the real-time phase offset = DeltaT / Twave, which represents the proportion by which the phase reference point of the fluctuation signal leads the start point of the disturbance period. This proportion is usually multiplied by 360 degrees to convert it into an angle value. If the calculated proportion is greater than 1, it indicates that the fluctuation period is less than the disturbance period, and the phase difference has exceeded one complete 360-degree cycle. In this case, the decimal part of the proportion is usually multiplied by 360 degrees as the effective phase difference.

[0091] Record the real-time phase shift calculated over multiple consecutive perturbation cycles. The number of consecutive perturbation cycles needs to be large enough to form a statistically significant change trajectory; for example, record the real-time phase shifts corresponding to the most recent 50 or 100 perturbation cycles. In practice, for each completed perturbation cycle, calculate a real-time phase shift using the method described above; this value can be stored as a scale value or a converted angle value. Store these values ​​sequentially in a first-in-first-out (FIFO) data buffer or a list structure according to the order in which their corresponding perturbation cycles occur. During recording, the perturbation cycle number or absolute start time can be associated with the data. For example, record a phase difference of 43.2 degrees for the first cycle, 45.1 degrees for the second cycle, and so on. This step systematically accumulates a sequence of raw observational data on the evolution of phase relationships over time.

[0092] The real-time phase offsets recorded in multiple consecutive disturbance cycles are connected in chronological order to form the trajectory of the real-time phase difference over the consecutive disturbance cycles. Logically, the connection operation means using the disturbance cycle number or the start time of these cycles as the horizontal axis and the corresponding real-time phase offset as the vertical axis to form a discrete sequence of data points. This sequence is the trajectory. For example, the horizontal axis represents cycle numbers 1 to 50, and the vertical axis represents the corresponding phase difference angle value sequence [43.2, 45.1, 40.8, ..., 55.0]. This trajectory visually reveals the dynamic behavior of the relative phase relationship between the main fluctuation and the controller disturbance. If two cycles are integer multiples of each other or very close, the phase difference may fluctuate within a small range around a fixed value. If the two cycles are mismatched, the phase difference usually shows a monotonically increasing or decreasing trend, and its rate of change reflects the small difference between the two cycles. This trajectory is one of the core input data for state judgment in subsequent step S5, such as determining whether quasi-steady-state characteristics are present and risk assessment; its morphological characteristics directly reflect the dynamic characteristics of system interference coupling.

[0093] S5. Based on the energy accumulation characteristics of the output power fluctuation sequence within the frequency sideband range of the disturbance period and combined with the change trajectory, assess the risk level of interference coupling between the main fluctuation frequency and the disturbance period. Specifically, this is implemented as follows:

[0094] Based on the energy accumulation characteristics of the output power fluctuation sequence within the frequency sideband range of the disturbance period and combined with the change trajectory, the risk level of interference coupling between the dominant fluctuation frequency and the disturbance period is assessed. First, the frequency sideband range is determined using the disturbance frequency of the maximum power point tracking control as the center frequency and the dominant fluctuation frequency as the sideband offset. The disturbance frequency of the maximum power point tracking control is the frequency of the periodic disturbance action set by the controller; it is a known or measurable parameter, for example, a value of 10 Hz corresponding to a period of 0.1 seconds. The dominant fluctuation frequency is the frequency value extracted in step S3 that characterizes the dominant periodic fluctuation in the output power, for example, a value of 5 Hz. The frequency sideband range is typically defined as the frequency band formed by offsetting one dominant fluctuation frequency above and below the disturbance frequency. Specifically, the lower boundary frequency equals the disturbance frequency minus the dominant fluctuation frequency; the upper boundary frequency equals the disturbance frequency plus the dominant fluctuation frequency. For example, subtracting the dominant wave frequency of 5 Hz from the perturbation frequency of 10 Hz yields the lower boundary frequency of 5 Hz; adding the dominant wave frequency of 5 Hz to the perturbation frequency of 10 Hz yields the upper boundary frequency of 15 Hz. The resulting frequency sideband range is from 5 Hz to 15 Hz. In some implementations, to focus on the core energy, the sideband range can be narrowed, for example, offset by 0.8 times the dominant wave frequency both above and below the perturbation frequency. That is, the sideband range extends from the perturbation frequency minus 0.8 times the dominant wave frequency to the perturbation frequency plus 0.8 times the dominant wave frequency. This narrowing factor of 0.8 can be adjusted based on the system's observational experience with beat frequency energy distribution, for example, selected between 0.7 and 0.95.

[0095] The ratio of the power spectral density integral value of the output power fluctuation sequence within the frequency sideband range to the total power spectral density integral value across the entire effective analysis frequency band is calculated to obtain the sideband energy amplification factor characterizing the energy accumulation feature. This calculation is based on the power spectral density estimation result of the output power fluctuation sequence obtained in step S3. The power spectral density integral is achieved through numerical integration. For a discrete power spectral density sequence, its frequency resolution is Δf, in Hertz. The sideband energy is calculated as follows: in the power spectral density sequence, find all data points with frequency values ​​greater than or equal to the lower boundary frequency and less than or equal to the upper boundary frequency, sum the power spectral density values ​​corresponding to these data points sequentially, and then multiply the sum by the frequency resolution Δf to obtain the sideband energy Eside, in Watts. The total power spectral density integral value is calculated as follows: within the entire effective analysis frequency band, for example from 0 Hertz to the Nyquist frequency (half of the sampling frequency), sum the power spectral density values ​​corresponding to all frequency points, and then multiply the sum by the frequency resolution Δf to obtain the total energy Etotal, also in Watts. The sideband energy amplification factor Kband is equal to the sideband energy Eside divided by the total energy Etotal, i.e., Kband = Eside / Etotal. This factor is a dimensionless number between 0 and 1. For example, if Eside is calculated to be 30 watts and Etotal to be 120 watts, then Kband is 0.25. The larger this factor is, the higher the proportion of system fluctuation energy concentrated in the sidebands offset by one main fluctuation frequency on both sides of the disturbance frequency. This directly reflects the intensity of energy modulation and concentration caused by the interference between the main fluctuation and the periodic disturbance.

[0096] Analyze the change trajectory to determine whether the real-time phase difference exhibits quasi-steady-state characteristics, exhibiting finite fluctuations around a fixed value. The change trajectory is a sequence output from step S4, composed of real-time phase difference values ​​corresponding to N consecutive disturbance cycles, arranged in chronological order. These real-time phase difference values ​​are typically converted to angle values, denoted as Φ1, Φ2, ..., ΦN. Quasi-steady-state characteristics refer to the sequence's values ​​fluctuating around a stable value with limited amplitude and no obvious monotonic trend. The judgment process first calculates the standard deviation of the sequence. The standard deviation σ is calculated as follows: first, calculate the arithmetic mean μ of the sequence, where μ equals the sum of all phase difference values ​​Φi divided by the number N, where i represents the index of the real-time phase difference value; then calculate the deviation (Φi-μ) of each phase difference value from the mean; calculate the sum of squares of these deviations; divide the sum of squares by (N-1); finally, take the square root of the result to obtain the standard deviation σ, with the same unit as the phase difference. Next, calculate the average rate of change of the sequence. The average rate of change ν is calculated as follows: The absolute values ​​of the differences between two adjacent values ​​in the sequence are calculated sequentially, i.e., |Φ2-Φ1|, |Φ3-Φ2|, ..., |ΦN-Φ{N-1}|, resulting in N-1 variables. Then, the arithmetic mean of these N-1 variables is calculated, which is the average rate of change ν, expressed in degrees per disturbance cycle. A first threshold σth and a second threshold νth need to be preset. The first threshold σth is used to limit the standard deviation, and its setting can be based on the typical fluctuation range of the phase difference under normal dynamic conditions in a non-phase-locked system. For example, the 90th or 95th percentile of the standard deviation of the phase difference during non-phase-locked periods can be taken as the first threshold σth by analyzing a large amount of historical data. This value might be 15 degrees. The second threshold νth is used to limit the average rate of change, and its setting can be based on the phase drift rate per cycle caused by the minimum theoretical frequency difference between the disturbance cycle and the main fluctuation cycle. For example, if the minimum resolvable frequency difference is 0.01 Hz and the disturbance period is 0.1 seconds, then the phase drift per period is approximately 0.36 degrees. The second threshold νth can be set to several times this value, such as 2 degrees per disturbance period. When both the calculated standard deviation σ is less than the preset first threshold σth and the calculated average rate of change ν is less than the preset second threshold νth, the real-time phase difference is determined to exhibit quasi-steady-state characteristics. Otherwise, it is determined not to exhibit quasi-steady-state characteristics.

[0097] If the real-time phase difference exhibits quasi-steady state characteristics, it is determined that the system has entered the stable phase-locked state, and based on the magnitude of the sideband energy amplification factor, the risk level of interference coupling characterized by energy aggregation in the stable phase-locked state is evaluated. The stable phase-locked state means that the relative phase relationship between the main fluctuation and the controller perturbation is locked within a narrow interval, and at this time, the energy loss pattern generated by the beat frequency effect is stable and continuous. The risk assessment in this state is mainly based on the sideband energy amplification factor Kband because the degree of energy aggregation is directly related to the severity of efficiency loss. The assessment is completed by comparing Kband with a preset grading threshold. For example, the first grading threshold Klow is set to 0.2, and the second grading threshold Khigh is set to 0.4. These thresholds can be determined by simulating or experimenting to establish a corresponding relationship curve of the average efficiency loss of the system under different Kband values. For example, when Kband is lower than 0.2, the observed average efficiency loss is less than 1%; when Kband is between 0.2 and 0.4, the efficiency loss is between 1% and 3%; when Kband is higher than 0.4, the efficiency loss is greater than 3%. The risk level is set according to this corresponding relationship: if Kband < Klow, it is evaluated as a low risk level; if Klow ≤ Kband < Khigh, it is evaluated as a medium risk level; if Kband ≥ Khigh, it is evaluated as a high risk level.

[0098] If the real-time phase difference does not exhibit quasi-steady-state characteristics, the risk level of dynamic interference coupling is assessed based on the magnitude of the sideband energy amplification factor and the trend of the real-time phase difference. At this point, the system is not phase-locked, the phase difference continuously changes, and the risk is determined by both the instantaneous energy accumulation intensity and the direction of phase evolution. The trend of the real-time phase difference, S, can be obtained by calculating the slope of the linear regression of the change trajectory, in degrees per disturbance cycle. The slope S is calculated by performing a least-squares linear fit on the disturbance cycle number (1, 2, ..., N) and the corresponding phase difference value sequence (Φ1, Φ2, ..., ΦN). The absolute value of S indicates the speed of phase drift, and the sign indicates the direction. Dynamic risk assessment requires a combination of Kband and |S|. A two-dimensional risk assessment table is predefined. For example, Kband is divided into three levels: low (<0.15), medium (0.15~0.3), and high (>0.3); |S| is divided into three levels: slow (<5 degrees / cycle), medium (5~15 degrees / cycle), and fast (>15 degrees / cycle). Then, low, medium, and high risk levels are assigned to the nine combinations. Another quantification method is to calculate the dynamic risk index Rdynamic = w1 × Kband + w2 × (|S| / Smax), where w1 and w2 are preset weighting coefficients and w1 + w2 = 1, and Smax is a preset maximum normalized reference slope, for example, 30 degrees per cycle. The setting of the weighting coefficients w1 and w2 reflects the relative importance of the two factors in the dynamic process, which can be determined by analyzing the degree of instantaneous power loss of the system under different (Kband, |S|) combinations in historical data using regression analysis. For example, w1 = 0.6 and w2 = 0.4 might be determined. After calculating Rdynamic, the final risk level is determined based on the preset interval into which its value falls, for example, [0, 0.3) for low risk, [0.3, 0.6) for medium risk, and [0.6, 1] for high risk. Through the above logic, a clear risk level assessment result is output.

[0099] S6. Adjust the disturbance period of the maximum power point tracking control according to the risk level so that the adjusted disturbance period is staggered from the main fluctuation frequency. Specifically, the implementation is as follows:

[0100] The disturbance period of the maximum power point tracking control is adjusted according to the risk level to ensure that the adjusted disturbance period is offset from the dominant fluctuation frequency. First, based on the risk level, a candidate disturbance period different from the current disturbance period is selected from a set of preset candidate disturbance periods as the target disturbance period. The risk level is a qualitative or quantitative result output after the evaluation in step S5, such as low risk, medium risk, or high risk, or a risk level represented by numbers 1, 2, and 3. The preset set of candidate disturbance periods is a set of discrete period values ​​predetermined and stored in the controller during the control system design phase. This set is set based on the resolution of the controller's hardware timer, the allowable disturbance frequency range for stable system operation, and the requirements for dynamic response speed. For example, if the minimum allowable disturbance period for the controller is 0.05 seconds and the maximum is 0.3 seconds, then several values ​​can be selected within this range at fixed or non-uniform intervals to form a candidate set, such as 0.05 seconds, 0.08 seconds, 0.1 seconds, 0.15 seconds, 0.2 seconds, 0.25 seconds, and 0.3 seconds. The logic for selecting the target perturbation period is as follows: the higher the risk level, the greater the potential threat the current interference coupling state poses to system efficiency. Therefore, it is necessary to select a candidate value that differs more significantly from the current perturbation period in numerical value, in order to more significantly change the system's operating point and thus more effectively escape the harmful coupling state. In practice, the absolute difference between the currently used perturbation period and each candidate value in the candidate set is calculated. Then, selection rules are formulated based on the risk level. For example, it can be predefined that: low risk level corresponds to selecting the candidate value corresponding to the smallest absolute difference, but excluding the current value itself corresponding to a difference of 0; medium risk level corresponds to selecting the candidate value corresponding to the difference in the middle position after all differences are sorted from smallest to largest; high risk level corresponds to selecting the candidate value corresponding to the largest absolute difference. Based on this rule, a preliminary target perturbation period is determined from the candidate set.

[0101] Calculate the absolute value of the difference between the frequency corresponding to the target disturbance period and the dominant fluctuation frequency. The frequency corresponding to the target disturbance period is equal to 1 divided by the value of the target disturbance period. The dominant fluctuation frequency is the value determined in step S3. The method for calculating the absolute value of the difference is to subtract the dominant fluctuation frequency from the target disturbance frequency and then take the absolute value of the result. This value quantifies the separation distance between the adjusted controller disturbance rhythm and the external dominant fluctuation rhythm in the frequency dimension.

[0102] When the absolute value of the frequency difference is less than the preset safe frequency interval, the step size is adjusted according to the preset disturbance period to increase the absolute value of the frequency difference towards the target direction, and the target disturbance period is recalculated. The preset safe frequency interval is a threshold value used to ensure effective frequency separation and avoid significant interference coupling caused by adjacent frequencies again. The setting of this interval needs to be based on the system's sensitivity to frequency proximity. One method is to observe the sideband energy amplification factor caused by the beat frequency effect in the system output power fluctuation at different frequency intervals through simulation or experiment, and set the minimum frequency interval corresponding to the reduction of the sideband energy amplification factor to an acceptable level as the safe frequency interval, for example, this value might be 2 Hz. The judgment condition is: whether the calculated absolute value of the frequency difference is less than the preset safe frequency interval. If it is less, it means that the initially selected target disturbance period has failed to meet the requirement of effective frequency separation, and a secondary adjustment is required. The secondary adjustment is based on a preset disturbance period adjustment step size. This step size is a positive time increment, for example, 0.02 seconds. The setting requires a trade-off between adjustment speed and system stability. Too small a step size may lead to a slow adjustment process, while too large a step size may cause transient shocks in the control loop. The goal of the recalculation is to increase the absolute value of the difference between the frequency corresponding to the new target disturbance period and the main fluctuation frequency. During implementation, it is determined whether the current target disturbance frequency is higher or lower than the main fluctuation frequency. If the target frequency is higher than the main fluctuation frequency, the target frequency should be further increased to increase the absolute value of the difference; conversely, the target frequency should be further decreased. Within the allowable period range, the current target period value is added to or subtracted from the preset adjustment step size, with the addition or subtraction determined according to the direction of increasing the difference, resulting in a new candidate period value. Then, the frequency corresponding to this new period is calculated, and the absolute value of its difference from the main fluctuation frequency is recalculated. If the new absolute value of the difference is still less than the safe frequency interval, this adjustment process is repeated, changing the step size in the same direction each time, until the absolute value of the difference is greater than or equal to the safe frequency interval, or reaches the boundary of the allowable period range. Finally, the period value that meets the conditions is determined as the final target disturbance period. If the adjustment in the initial direction has reached the boundary and still does not meet the conditions, you can try adjusting in the opposite direction.

[0103] The disturbance period of the maximum power point tracking control is updated to the final determined target disturbance period. The final determined target disturbance period is a period value obtained after the above selection and possible recalculation iterations, ensuring that the absolute value of the difference between its corresponding frequency and the main fluctuation frequency is not less than a preset safe frequency interval. The update operation is achieved by modifying the timer or counter parameters in the controller responsible for generating periodic disturbance signals. Specifically, the controller software writes the final target disturbance period value into the corresponding hardware timer period register, or updates the timing variable used to control the disturbance interval in the software loop. For example, if the final target disturbance period is 0.12 seconds, the timer is set to trigger again 0.12 seconds after each interruption, thereby controlling the sending period of the disturbance command. From the next disturbance period, the controller will operate according to this new period. This update operation actively changes the controller's internal rhythm, maintaining a sufficient safe distance in the frequency domain between it and the main frequency of external environmental fluctuations, thereby disrupting the harmful interference coupling conditions identified in steps S2 to S5, enabling the system to enter a more stable and efficient operating state, and realizing energy-saving adaptive control based on real-time state perception and risk assessment.

[0104] Example 2: Figure 2 A schematic diagram of an energy-saving photovoltaic inverter control system according to the present invention is provided. The energy-saving photovoltaic inverter control system includes:

[0105] The sequence forming module is used to monitor and acquire the output power of the photovoltaic array and the DC bus voltage of the photovoltaic inverter, and to form the output power fluctuation sequence and the DC bus voltage ripple sequence.

[0106] The feature judgment module is used to perform correlation analysis on the output power fluctuation sequence and the DC bus voltage ripple sequence to determine whether there are cooperative change features that characterize periodic disturbance inputs.

[0107] The frequency extraction module is used to analyze the frequency distribution of the output power fluctuation sequence and extract the main fluctuation frequency when there are cooperative variation characteristics.

[0108] The trajectory tracking module is used to analyze the real-time phase difference between the fluctuation period corresponding to the main fluctuation frequency and the disturbance period of the maximum power point tracking control, and to track the trajectory of the real-time phase difference within the continuous disturbance period.

[0109] The risk level assessment module is used to assess the risk level of interference coupling between the main fluctuation frequency and the disturbance period based on the energy accumulation characteristics of the output power fluctuation sequence in the frequency sideband range of the disturbance period and the change trajectory.

[0110] The level adjustment module is used to adjust the disturbance period of the maximum power point tracking control according to the risk level, so that the adjusted disturbance period is staggered from the main fluctuation frequency.

[0111] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0112] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0113] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission; wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission includes infrared, microwave, etc. Computer-readable storage media can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0114] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0115] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0116] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0117] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0118] If a function is implemented as a software module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0119] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0120] In conclusion, the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A control method for an energy-saving photovoltaic inverter, characterized in that, include: S1. Monitor and acquire the output power of the photovoltaic array and the DC bus voltage of the photovoltaic inverter to form an output power fluctuation sequence and a DC bus voltage ripple sequence; S2. Perform correlation analysis on the output power fluctuation sequence and the DC bus voltage ripple sequence to determine whether there are cooperative change characteristics that characterize the periodic disturbance input. S3. When cooperative variation characteristics exist, analyze the frequency distribution of the output power fluctuation sequence and extract the main fluctuation frequency, including: Determine the time period corresponding to the cooperative change characteristics in the output power fluctuation sequence; The power spectral density is obtained by performing spectral estimation on the output power fluctuation sequence over a time period. Identify local peaks in the power spectral density and determine the frequency and amplitude corresponding to each local peak; The frequency corresponding to the highest peak whose amplitude exceeds the preset amplitude threshold is determined as the main fluctuation frequency. S4. Analyze the real-time phase difference between the fluctuation period corresponding to the main fluctuation frequency and the disturbance period of the maximum power point tracking control, and track the change trajectory of the real-time phase difference in the continuous disturbance period. S5. Based on the energy accumulation characteristics of the output power fluctuation sequence within the frequency sideband range of the disturbance period and combined with the change trajectory, assess the risk level of interference coupling between the main fluctuation frequency and the disturbance period, including: The frequency sideband range is determined by using the disturbance frequency of maximum power point tracking control as the center frequency and the main fluctuation frequency as the sideband offset. The power spectrum integral value of the output power fluctuation sequence in the frequency sideband range is calculated to the total power spectrum integral value in the entire effective analysis frequency band, and the sideband energy amplification factor characterizing the energy accumulation feature is obtained. Analyze the change trajectory to determine whether the real-time phase difference exhibits quasi-steady-state characteristics with finite fluctuations around a certain fixed value; If the real-time phase difference exhibits quasi-steady-state characteristics, the system is determined to have entered a stable phase-locked state. Based on the magnitude of the sideband energy amplification factor, the risk level of interference coupling characterized by energy accumulation in the stable phase-locked state is assessed. If the real-time phase difference does not exhibit quasi-steady-state characteristics, the risk level of dynamic interference coupling is assessed based on the magnitude of the sideband energy amplification factor and the changing trend of the real-time phase difference. S6. Adjust the disturbance period of the maximum power point tracking control according to the risk level so that the adjusted disturbance period is offset from the main fluctuation frequency.

2. The energy-saving photovoltaic inverter control method according to claim 1, characterized in that, S1 includes: The output current and output voltage of the photovoltaic array are collected synchronously with a preset synchronous sampling period, and the DC bus voltage is also collected synchronously. The instantaneous power value is calculated based on the synchronously acquired output current and output voltage. The instantaneous power value is filtered to form an output power fluctuation sequence; The synchronously acquired DC bus voltage is processed by extracting the AC component to form a DC bus voltage ripple sequence.

3. The energy-saving photovoltaic inverter control method according to claim 1, characterized in that, S2 include: Time-align the output power fluctuation sequence with the DC bus voltage ripple sequence; Within multiple consecutive sliding time windows, the cross-correlation coefficients between the time-aligned output power fluctuation sequence and the DC bus voltage ripple sequence are calculated respectively. Determine whether the calculated cross-correlation coefficient continuously exceeds a preset correlation threshold within a continuous sliding time window; When the cross-correlation coefficient continuously exceeds the preset correlation threshold within a continuous sliding time window, it is determined that there is a cooperative change feature that characterizes the periodic perturbation input.

4. The energy-saving photovoltaic inverter control method according to claim 1, characterized in that, S4 includes: Calculate the corresponding fluctuation period based on the main fluctuation frequency; Within each disturbance cycle of maximum power point tracking control, the real-time phase offset of the fluctuation cycle relative to the start time of the current disturbance cycle is calculated based on the zero-crossing point or a specific phase point of the output power fluctuation sequence. Record the real-time phase offset calculated over multiple consecutive disturbance cycles; The real-time phase offsets recorded in multiple consecutive disturbance cycles are connected in chronological order to form the trajectory of the real-time phase difference change in consecutive disturbance cycles.

5. The energy-saving photovoltaic inverter control method according to claim 1, characterized in that, The sideband energy amplification factor, which characterizes the energy accumulation feature, is obtained in the following way: Using the disturbance frequency of maximum power point tracking control as the center frequency and the main fluctuation frequency as the sideband offset, the upper and lower boundary frequencies of the frequency sideband are determined. Calculate the power spectrum integral value of the output power fluctuation sequence in the frequency range between the lower boundary frequency and the upper boundary frequency, and use it as the sideband energy; Calculate the total power spectrum integral value of the output power fluctuation sequence over the entire effective analysis frequency band; The ratio of the sideband energy to the integral value of the total power spectrum is used as the sideband energy amplification factor.

6. The energy-saving photovoltaic inverter control method according to claim 1, characterized in that, Analyze the trajectory of change to determine whether the real-time phase difference exhibits quasi-steady-state characteristics of finite fluctuations around a fixed value, including: Calculate the standard deviation of the trajectory formed by the real-time phase difference over multiple consecutive disturbance cycles; The average value of the absolute values ​​of the differences between adjacent real-time phase differences in the changing trajectory is used as the average rate of change. When the standard deviation is less than a preset first threshold and the average rate of change is less than a preset second threshold, the real-time phase difference is determined to exhibit quasi-steady-state characteristics.

7. The energy-saving photovoltaic inverter control method according to claim 1, characterized in that, S6 include: Based on the risk level, select a disturbance cycle candidate value that is different from the current disturbance cycle from a number of preset disturbance cycle candidate values ​​as the target disturbance cycle; Calculate the absolute value of the difference between the frequency corresponding to the target disturbance period and the main fluctuation frequency; When the absolute value of the difference is less than the preset safe frequency interval, the step size is adjusted according to the preset disturbance period to increase the absolute value of the difference to the target direction, and the target disturbance period is recalculated. The disturbance period of the maximum power point tracking control is updated to the final determined target disturbance period.

8. An energy-saving photovoltaic inverter control system, used to implement the energy-saving photovoltaic inverter control method according to any one of claims 1-7, characterized in that, include: The sequence forming module is used to monitor and acquire the output power of the photovoltaic array and the DC bus voltage of the photovoltaic inverter, and to form the output power fluctuation sequence and the DC bus voltage ripple sequence. The feature judgment module is used to perform correlation analysis on the output power fluctuation sequence and the DC bus voltage ripple sequence to determine whether there are cooperative change features that characterize the periodic disturbance input. The frequency extraction module is used to analyze the frequency distribution of the output power fluctuation sequence and extract the main fluctuation frequency when there are cooperative variation characteristics. The trajectory tracking module is used to analyze the real-time phase difference between the fluctuation period corresponding to the main fluctuation frequency and the disturbance period of the maximum power point tracking control, and to track the trajectory of the real-time phase difference within the continuous disturbance period. The risk level assessment module is used to assess the risk level of interference coupling between the main fluctuation frequency and the disturbance period based on the energy accumulation characteristics of the output power fluctuation sequence in the frequency sideband range of the disturbance period and the change trajectory. The risk level adjustment module is used to adjust the disturbance period of the maximum power point tracking control according to the risk level, so that the adjusted disturbance period is staggered from the main fluctuation frequency.

Images