Tracking angle adjusting method based on photovoltaic panel dust retention light transmittance fluctuation

Abstract

The invention relates to the field of photovoltaic power generation, and discloses a photovoltaic panel dust retention light transmittance fluctuation-based tracking angle adjustment method, which comprises the following steps of S1, acquiring a reference tracking angle of a photovoltaic panel; s2, acquiring a periodic angle disturbance signal with specific injection frequency and amplitude; s3, acquiring a compensation angle of the photovoltaic panel; s4, synthesizing the reference tracking angle, the compensation angle and the periodic angle disturbance signal into an actual tracking angle instruction; and S5, a photovoltaic tracker is driven by a servo motor, so that the photovoltaic tracker adjusts the angle along with the actual tracking angle instruction. The power generation efficiency can be improved by dynamically responding to the dust accumulation influence in real time, and the power change caused by dust accumulation light transmittance fluctuation can be captured in real time and an accurate angle mismatch error signal can be generated by actively injecting a periodic angle disturbance signal and combining synchronous demodulation processing; and the compensation angle is dynamically updated through the integral controller, so that the photovoltaic panel always approaches the actual optimal angle.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photovoltaic power generation, in particular to a tracking angle adjustment method based on dust accumulation and light transmittance fluctuation of a photovoltaic panel. BACKGROUND

[0002] In a photovoltaic power generation system, the tracking angle accuracy of a photovoltaic panel directly determines the solar radiation capture efficiency. The existing technology usually relies on a solar position algorithm (SPA) combined with GPS or NTP data to calculate a reference tracking angle, and then adjusts the angle through a servo motor to drive a tracker to achieve real-time tracking of the sun. However, in actual application, dust is easily accumulated on the surface of the photovoltaic panel, and the dust accumulation will cause light transmittance fluctuation, which in turn will cause the output power of the photovoltaic system to decrease. At the same time, uneven dust distribution may also cause local maximum values in the power curve, resulting in the traditional tracking adjustment method falling into a local optimal trap and failing to accurately match the actual optimal angle, which seriously affects the power generation efficiency.

[0003] The current mainstream angle adjustment method has two key problems: first, it lacks dynamic sensing ability for dust accumulation and light transmittance fluctuation, and usually uses fixed period or manual inspection to adjust the angle, which cannot respond to the power change caused by dust accumulation in real time, and has strong adjustment lag; second, the signal processing and angle optimization mechanism is imperfect, and the power signal is easily disturbed by background noise during collection, and an effective error feedback and nonlinear diagnosis mechanism has not been established, which makes it difficult to accurately identify the angle mismatch problem, resulting in low adjustment accuracy and failing to fully offset the power generation loss caused by dust accumulation. SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides a tracking angle adjustment method based on dust accumulation and light transmittance fluctuation of a photovoltaic panel, which solves the problem of "the prior art does not establish an effective error feedback and nonlinear diagnosis mechanism, and it is difficult to accurately identify the angle mismatch problem".

[0005] To achieve the above purpose, the present application realizes the following technical scheme: a tracking angle adjustment method based on dust accumulation and light transmittance fluctuation of a photovoltaic panel, comprising the following steps: S1: obtaining a reference tracking angle of a photovoltaic panel; S2: obtaining a periodic angle disturbance signal with a specific injection frequency and amplitude; S3: obtaining a compensation angle of the photovoltaic panel; S4: combining the reference tracking angle, the compensation angle and the periodic angle disturbance signal into an actual tracking angle instruction; S5: driving the photovoltaic tracker through a servo motor to adjust the angle following the actual tracking angle instruction; S6: collecting the total output power signal of the photovoltaic system in real time through a power sensor; S7: synchronously demodulating the total output power signal to extract a power harmonic component related to the injection frequency, and generating an angle mismatch error signal; S8: updating the compensation angle based on the angle mismatch error signal.

[0006] Preferably, the step of obtaining the reference tracking angle in S1 comprises: obtaining high-precision timestamp and geographic coordinate data through a GPS module or a network time protocol (NTP) service; and calculating the reference tracking angle through a sun position algorithm (SPA).

[0007] Preferably, the step of collecting the total output power signal in real time through the power sensor in S6 comprises: using a high-frequency Hall current sensor and a Hall voltage sensor installed on the direct-current side of the inverter to perform high-speed sampling at a preset sampling frequency, and obtaining an instantaneous total output power signal.

[0008] Preferably, the step of generating the angle mismatch error signal in S7 comprises: generating a fundamental reference signal consistent with the injection frequency inside the controller; multiplying the total output power signal by the fundamental reference signal to perform mixing processing, and obtaining a mixed signal; and performing low-pass filtering processing on the mixed signal to obtain the angle mismatch error signal.

[0009] Preferably, the step of updating the compensation angle based on the angle mismatch error signal in S8 comprises: taking the angle mismatch error signal as an error input, and calculating through an integral controller; and updating the compensation angle according to the calculation result of the integral controller, with the control target being to make the angle mismatch error signal approach zero.

[0010] Preferably, before performing S2 to obtain the periodic angle disturbance signal, the method further comprises the step of: performing fast Fourier transform (FFT) processing on historical data of the total output power signal of the photovoltaic system to obtain a background noise power spectrum.

[0011] Preferably, the method further comprises the steps of: searching for a minimum point in the background noise power spectrum within a preset injection frequency band range; and determining the frequency corresponding to the minimum point as the injection frequency of the periodic angle disturbance signal.

[0012] Preferably, the method further comprises the steps of: dynamically setting the amplitude of the periodic angle disturbance signal according to the noise floor corresponding to the injection frequency; and setting the gain coefficient and the minimum disturbance amplitude in consideration of the preset.

[0013] Preferably, the synchronous demodulation processing in S7 further comprises: generating a double-frequency reference signal with a frequency being twice the injection frequency inside the controller; multiplying the total output power signal by the double-frequency reference signal, and performing low-pass filtering processing on the multiplication result to obtain a nonlinear diagnosis signal.

[0014] Preferably, the method further comprises the steps of: comparing the amplitude of the nonlinear diagnostic signal with a preset nonlinear threshold; when the amplitude of the nonlinear diagnostic signal is greater than the nonlinear threshold, triggering the control mode switching, suspending the compensation angle updating based on the angle mismatch error signal in S8, and instead performing a global scan optimization within a preset angle range to avoid local maximum.

[0015] The application provides a tracking angle adjustment method based on dust accumulation light transmittance fluctuation of a photovoltaic panel. 1、The application can improve power generation efficiency by real-time dynamic response to dust accumulation, can generate accurate angle mismatch error signals by actively injecting periodic angle disturbance signals, combining synchronous demodulation processing, and capturing power changes caused by dust accumulation light transmittance fluctuation in real time, and can let the photovoltaic panel always approach the actual optimal angle by dynamic updating of the compensation angle by the integral controller, effectively offsetting the power loss caused by dust accumulation.

[0016] 2、The application can greatly reduce the interference of background noise on power signal acquisition by analyzing the background noise power spectrum of the power history data through FFT processing, selecting the frequency with the smallest noise, and dynamically setting the signal amplitude; the introduction of a double-frequency reference signal in the synchronous demodulation process generates a nonlinear diagnostic signal, which can accurately identify local maximum problems and improve adjustment accuracy.

[0017] 3、The application can directly use the Hall sensor on the DC side of the inverter to collect power signals, rely on the existing servo motor and controller to realize function upgrade, and has strong compatibility; at the same time, the saturation limitation of the integral controller and the timing recovery mechanism of the global scan ensure that the system can still operate stably in complex dust accumulation environment, which can reduce manual inspection and adjustment frequency and reduce maintenance cost. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 The flowchart of the application. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the specification of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0020] Embodiment: Please refer to the drawings in the specification of the application Figure 1 The embodiment of the application provides a tracking angle adjustment method based on dust accumulation light transmittance fluctuation of a photovoltaic panel, which comprises the following steps: S1: Obtain the reference tracking angle of the photovoltaic panel, which is the initial theoretical basis for the photovoltaic system to track the sun, providing accurate reference for subsequent angle adjustment, avoiding insufficient solar radiation capture due to initial direction deviation; S2: Obtain a periodic angle perturbation signal with a specific injection frequency and amplitude. By actively injecting the signal, the impact of dust accumulation on power generation can be detected in real time, providing a key carrier for subsequent identification of angle mismatch, allowing the system to dynamically perceive light transmittance fluctuations; S3: Obtain the compensation angle of the photovoltaic panel, which is used to correct the deviation between the reference tracking angle and the actual optimal angle. Reasonably setting the initial compensation angle can shorten the adjustment convergence time and help the system quickly enter the high-efficiency power generation state; S4: Combine the reference tracking angle, compensation angle, and periodic angle perturbation signal into the actual tracking angle command. This command integrates theoretical reference, deviation correction, and active detection requirements, allowing the photovoltaic panel to track the sun while detecting the impact of dust accumulation through perturbation feedback, laying the foundation for precise adjustment; S5: Drive the photovoltaic tracker through a servo motor to adjust the angle according to the actual tracking angle command. The high-precision drive of the servo motor ensures that the angle follows without lag, avoiding solar radiation waste due to slow mechanical response and ensuring adjustment accuracy; S6: Collect the total output power signal of the photovoltaic system in real time through a power sensor. The power signal is the core data reflecting the impact of dust transmittance fluctuations, and real-time collection can capture power changes in time, providing original basis for subsequent demodulation processing and error signal generation; S7: Perform synchronous demodulation processing on the total output power signal to extract power harmonic components related to the injection frequency and generate an angle mismatch error signal. Synchronous demodulation can accurately separate the power fluctuations caused by dust, and the error signal can quantify the angle deviation, indicating the direction of compensation adjustment; S8: Update the compensation angle based on the angle mismatch error signal. By continuously updating the compensation angle, the tracking deviation can be continuously corrected, allowing the photovoltaic panel to always approach the optimal angle and maximize the power generation efficiency loss caused by dust transmittance fluctuations.

[0021] Further, the reference tracking angle in S1 needs to be acquired by first acquiring high-precision timestamp and geographic coordinate data through a GPS module or a network time protocol (NTP) service. The GPS module selects an industrial-grade module with a positioning accuracy of 1 meter and a time synchronization accuracy of ±10 ns. The longitude collected needs to be accurate to degrees and retain 6 decimal places, and the latitude also needs to be accurate to degrees and retain 6 decimal places. The timestamp format is year-month-day hour: minute: second and is accurate to seconds. Such high-precision data can provide a reliable time and space basis for solar position calculation and avoid errors in reference angle calculation due to data deviation. Then, the reference tracking angle is calculated through a solar position algorithm (SPA). The SPA algorithm needs to derive key parameters such as the solar declination angle and the hour angle first. The solar declination angle calculation formula is

[0022] where n is the accumulated day of the year, January 1 is 1, and December 31 is 365 or 366. The hour angle calculation formula is

[0023] and the unit is degrees. The true solar time needs to be corrected according to the local time difference. The solar altitude angle calculation formula is

[0024] where φ is the installation inclination angle of the photovoltaic panel and is a fixed value (such as 35°). The solar azimuth angle calculation formula is

[0025] The reference tracking angle calculation formula is

[0026] Through these parameter derivation and formula calculation, the reference tracking angle can be matched with the actual solar position, providing an accurate initial direction for subsequent adjustment.

[0027] Furthermore, in step S6, the real-time acquisition of the total output power signal via power sensors requires the use of high-frequency Hall current and Hall voltage sensors installed on the DC side of the inverter. The DC side of the inverter is the core component of the photovoltaic system's power output; installing sensors here allows for direct acquisition of the system's true power data, avoiding measurement deviations caused by line losses. The current sensor should be selected based on a range covering the system's maximum output current, such as 0-100A, while also meeting the requirements of a sensitivity of 40mV / A and a linearity error ≤0.5%. The voltage sensor should be selected based on a range matching the inverter's DC side voltage, such as 0-1000V. The transformer ratio is 1:2500 with a linearity error ≤0.2%. Both are connected using shielded twisted-pair cables, with the shield grounded at one end. This connection method effectively reduces the impact of electromagnetic interference on signal acquisition, ensuring accurate current and voltage data. High-speed sampling is performed according to a preset sampling frequency, which must satisfy the Nyquist sampling theorem and is usually set to 20Hz. If the upper limit of the injection frequency is subsequently adjusted, the sampling frequency must be increased synchronously. For example, if the upper limit of the injection frequency is 10Hz, the sampling frequency must be ≥40Hz. This avoids signal aliasing and completely preserves the fluctuation information in the power signal. The instantaneous total output power signal is obtained through the formula...

[0028] The calculation yields U(t) as the sampled instantaneous voltage in V and I(t) as the sampled instantaneous current in A. The collected U(t) and I(t) data are subjected to a third-order Butterworth low-pass filter with a cutoff frequency of 10Hz to remove high-frequency noise before being substituted into the formula. The filtered current and voltage data are more stable, ensuring that the calculated instantaneous total output power signal accurately reflects the system's power generation status, providing high-quality data support for subsequent demodulation processing.

[0029] Furthermore, in step S7, generating the angle mismatch error signal requires first generating a fundamental frequency reference signal consistent with the injected frequency within the controller. The controller uses a microcontroller with direct digital synthesis (DDS) capabilities, such as the STM32H743. The formula for generating the fundamental frequency reference signal is as follows:

[0030] in The injection frequency of the periodic angular disturbance signal is expressed in Hz. The signal frequency accuracy must be ≤ ±0.1Hz, and the phase stability ≤ ±1° / min. This high-precision reference signal ensures accurate frequency and phase matching with the disturbance signal, laying the foundation for subsequent mixing processing. The total output power signal is then multiplied by the base frequency reference signal for mixing. This mixing process is implemented through the controller's internal multiplier module, using the following formula:

[0031] The process can convert the power harmonic component related to the injection frequency in the total output power signal into a low-frequency component, facilitating subsequent filtering separation; finally, the mixed frequency signal is subjected to low-pass filtering processing, and the low-pass filtering adopts a 4th order Butterworth filter with a cutoff frequency set to When the cutoff frequency is 0.46 Hz, the filter needs to ensure that the high-frequency component is attenuated by ≥40 dB, and the final angle mismatch error signal formula is:

[0032] Wherein represents low-pass filtering processing. After low-pass filtering, high-frequency noise interference can be effectively removed, and low-frequency components related to angle mismatch are retained, and the generated angle mismatch error signal can clearly reflect the deviation of the current tracking angle from the optimal angle, providing a clear basis for updating the compensation angle.

[0033] Further, updating the compensation angle based on the angle mismatch error signal in S8 requires the angle mismatch error signal as the error input. The error input can accurately inform the controller of the size and direction of the current angle deviation, avoiding blind adjustment; calculation is performed through an integral controller, and the integral controller adopts a proportional-integral (PI) control algorithm, and the core formula is:

[0034] Wherein is the compensation angle of the current control period and the unit is °, is the compensation angle of the last control period and the unit is °, is the proportional coefficient, usually 0.02° / W, used for fast response to error, is the integral coefficient, usually 0.05° / (W・s), used to eliminate steady-state error, is the integral value of the angle mismatch error signal from the initial time to the current time and the unit is The control algorithm combines the rapidity of proportional control and the static error elimination characteristic of integral control, and can ensure accurate adjustment of the compensation angle; the control period Δt is set to 1s, and after each calculation, it is necessary to determine whether the compensation angle exceeds the reasonable range, usually the reasonable range is -5° to 5°, if , or , the integral term accumulation needs to be suspended to avoid integral saturation, which will cause the compensation angle to adjust overshoot and affect system stability. Through such judgment and limitation, the compensation angle adjustment can be ensured to be stable, and the control target can be finally achieved, i.e. the angle mismatch error signal tends to zero, so that the photovoltaic panel is always at the optimal power generation angle.

[0035] Further, before performing S2 to obtain the periodic angle disturbance signal, the history data of the total output power signal of the photovoltaic system also needs to be subjected to fast Fourier transform (FFT) processing to obtain the background noise power spectrum. First, the total output power history data of the last 48 hours is called, the data sampling interval is consistent with the real-time sampling interval, such as 0.05 s, and the data of the night light period (at this time the power is ≤100 W) and the extreme weather period (at this time the wind speed is >10 m / s and the precipitation is >5 mm / h) needs to be excluded. The power data of these periods has no reference value, and after exclusion, the accuracy of subsequent analysis can be ensured. Then, the remaining data is subjected to detrending processing, which is to subtract the sliding mean value, and the sliding window size is set to 1000 points. The detrending processing can eliminate the long-term drift in the power data and make the data better reflect the noise characteristics. The FFT processing adopts 1024-point Hanning window weighting. The Hanning window can reduce the frequency spectrum leakage and improve the FFT processing accuracy. The frequency resolution is The background noise power spectrum is calculated as The unit is W² / Hz, and the calculation formula is:

[0036] is the pre-processed power history data, represents the square of the modulus of the FFT result, and finally the power spectrum data in the frequency range of 0-10 Hz is retained for subsequent analysis. This frequency range covers the possible injection frequency band of the disturbance signal. The obtained background noise power spectrum can clearly reflect the noise intensity at each frequency, providing a basis for selecting a low-noise injection frequency for subsequent analysis, and avoiding the disturbance signal being masked by noise.

[0037] Further, the minimum point in the background noise power spectrum in the preset injection frequency band range also needs to be searched, and the frequency corresponding to the minimum point is determined as the injection frequency of the periodic angle disturbance signal. The preset injection frequency band is usually set to [0.5 Hz, 5 Hz]. This range can avoid low-frequency mechanical resonance (usually <0.5 Hz) and high-frequency signal attenuation (usually >5 Hz), and at the same time, known interference frequencies such as power grid harmonics (such as 50 Hz, 100 Hz) and motor operation need to be excluded. These interference frequencies will affect the detection of the disturbance signal, and after exclusion, the rationality of the selection of the injection frequency can be ensured. The golden section method is used to search for the minimum point, the initial search interval is [0.5 Hz, 5 Hz], the iteration accuracy is set to 0.1 Hz, the noise power values of the two golden section points in the calculation interval are calculated each time, and the search range is gradually reduced. The golden section method can efficiently find the minimum point and reduce the calculation amount. The formula for finally determining the injection frequency is:

[0038] wherein representing the variable that minimizes the function value, determining After that, 10 minutes of real-time power data need to be continuously collected to verify whether the deviation of the real-time noise power at the frequency from the historical data is ≤20%. If the deviation exceeds the limit, a new search is needed. This verification step can ensure that the selected injection frequency is still in a low noise state in the real-time environment, providing a guarantee for the effective detection of subsequent disturbance signals.

[0039] Further, the amplitude of the periodic angular disturbance signal also needs to be dynamically set according to the noise floor corresponding to the injection frequency, and the preset gain coefficient and minimum disturbance amplitude need to be considered when setting. First, the injection frequency is extracted from the background noise power spectrum corresponding to the noise floor , with units of W² / Hz. The noise floor reflects the interference intensity at the injection frequency and is a key basis for setting the amplitude. Then, the disturbance amplitude is calculated by the formula:

[0040] where A is the amplitude of the periodic angular disturbance signal and has units of °, k_A is the gain coefficient, and the empirical value is usually taken which can be optimized through experiments, representing the square root operation, and the minimum disturbance amplitude is usually taken as 0.5°, ensuring that the signal can be detected. This formula can match the amplitude setting with the noise intensity, avoiding a signal that is too weak or too strong. After calculation, the amplitude needs to be limited. If A>2°, it is forcibly set to 2° to prevent the amplitude from being too large, causing the photovoltaic panel to adjust frequently and significantly, increasing mechanical wear and tear. If A<0.3°, it is forcibly set to 0.3° to ensure that the signal-to-noise ratio is ≥10dB, allowing the disturbance signal to be clearly identified. The final formula for the periodic angular disturbance signal is:

[0041] where is the disturbance angle at time t and has units of °. Through such amplitude setting, a disturbance signal that meets the detection requirements and does not affect system stability can be generated, providing a reliable tool for subsequent angle mismatch detection.

[0042] Further, the synchronous demodulation processing in the S7 also needs to generate a double-frequency reference signal with a frequency that is twice the injection frequency inside the controller. Then, the total output power signal is multiplied by the double-frequency reference signal, and the multiplication result is subjected to low-pass filtering processing to obtain a nonlinear diagnostic signal. The double-frequency reference signal is generated by the DDS module of the controller, and the formula is:

[0043] The frequency is exactly 2 times the injection frequency, and the double-frequency reference signal is Phase synchronization ensures the accuracy of demodulation processing; the mixing formula is:

[0044] This process can separate the nonlinear components in the power signal; low-pass filtering uses the same method as the generation... Using a 4th-order Butterworth filter with the same parameters to ensure consistent filtering performance, the final formula for the nonlinear diagnostic signal is:

[0045] That is, when the dust accumulation on the photovoltaic panel is uniformly distributed, the power curve exhibits a single-peak characteristic. When the amplitude is ≤0.3W, and uneven dust distribution causes a local maximum value in the power curve, The amplitude will increase significantly, through The amplitude change can be used as a basis for judging the nonlinear state of the system and provide a reference for subsequent control mode switching.

[0046] Furthermore, the amplitude of the nonlinear diagnostic signal needs to be compared with a preset nonlinear threshold. When the amplitude of the nonlinear diagnostic signal exceeds the nonlinear threshold, a control mode switch is triggered, pausing the compensation angle update based on the angle mismatch error signal in S8. Instead, a global scan optimization is performed within a preset angle range to avoid local maxima. The preset nonlinear threshold is used. It is usually set to 0.9W, which is the value under normal linear conditions. A threshold set to three times the maximum value effectively distinguishes between normal and non-linear states, preventing false triggering; the comparison logic is... This indicates that the system is in a linear state, and the original compensation angle update logic is maintained. This indicates that the system may be trapped in a local maximum, requiring a global scan for optimization. The scan optimization steps are to first pause the integral controller and record the current compensation angle. To facilitate subsequent recovery; then set the scan range. The current baseline tracking angle covers the optimal angle deviation that dust accumulation may cause. The photovoltaic panel angle is adjusted point by point in 0.1° increments, with a 2-second pause at each increment to ensure stable power acquisition and to collect the corresponding power values. Then find the point of maximum power, using the formula:

[0047] in The variable that maximizes the function value is represented; the maximum power point found corresponds to the true optimal angle; update the compensation angle:

[0048] The compensation angle can be corrected to the optimal value; finally, the integral controller is restored, and the proportional coefficient and integral coefficient are reduced, such as Run for 10 minutes to ensure that the system is stable at the globally optimal angle. If the system is still running at the threshold value, re-perform the scan optimization after 30 minutes. The globally optimized scan can jump out of the local maximum value trap, ensuring that the system always runs at the most efficient angle.

[0049] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.​

Claims

1. A method for adjusting the tracking angle based on the fluctuation of the light transmittance of a photovoltaic panel due to dust accumulation, characterized in that, The method comprises the following steps: S1: obtaining a reference tracking angle of the photovoltaic panel; S2: obtaining a periodic angle perturbation signal with a specific injection frequency and amplitude; S3: obtaining a compensation angle of the photovoltaic panel; S4: synthesizing the reference tracking angle, the compensation angle and the periodic angle perturbation signal into an actual tracking angle instruction; S5: driving the photovoltaic tracker through a servo motor to adjust the angle according to the actual tracking angle instruction; S6: collecting a total output power signal of the photovoltaic system in real time through a power sensor; S7: performing synchronous demodulation processing on the total output power signal to extract a power harmonic component related to the injection frequency, and generating an angle mismatch error signal; S8: updating the compensation angle based on the angle mismatch error signal.

2. The method of claim 1, wherein the method is based on the fluctuation of the light transmittance of the photovoltaic panel due to dust accumulation. The step of obtaining the reference tracking angle in S1 comprises: obtaining high-precision time stamp and geographic coordinate data through a GPS module or a network time protocol service; and calculating the reference tracking angle through a sun position algorithm.

3. The method of claim 1, wherein the method is based on the fluctuation of the light transmittance of the photovoltaic panel due to dust accumulation. The step of collecting the total output power signal in real time through the power sensor in S6 comprises: using a high-frequency Hall current sensor and a Hall voltage sensor installed on a direct-current side of an inverter to perform high-speed sampling at a preset sampling frequency, and obtaining an instantaneous total output power signal.

4. The method of claim 1, wherein the method is based on the fluctuation of the light transmittance of the photovoltaic panel due to dust accumulation. The step of generating the angle mismatch error signal in S7 comprises: generating a base frequency reference signal consistent with the injection frequency inside a controller; multiplying the total output power signal and the base frequency reference signal to perform mixing processing, and obtaining a mixed signal; and performing low-pass filtering processing on the mixed signal to obtain the angle mismatch error signal.

5. The method of claim 4, wherein the tracking angle is adjusted based on the fluctuation of the light transmittance of the photovoltaic panel. The step of updating the compensation angle based on the angle mismatch error signal in S8 comprises: taking the angle mismatch error signal as an error input, and performing calculation through an integral controller; and updating the compensation angle according to the calculation result of the integral controller, so that the angle mismatch error signal tends to be zero.

6. The method of tracking angle adjustment based on fluctuation of light transmittance of dust accumulation on a photovoltaic panel according to claim 1, characterized in that, Before performing S2 to obtain the periodic angle perturbation signal, the method further comprises the step of: performing fast Fourier transform processing on historical data of the total output power signal of the photovoltaic system to obtain a background noise power spectrum.

7. The method of tracking angle adjustment based on fluctuation of light transmittance of dust accumulation on a photovoltaic panel according to claim 6, characterized in that, The method further comprises the following steps: searching for a minimum point in the background noise power spectrum within a preset injection frequency band range; and determining a frequency corresponding to the minimum point as the injection frequency of the periodic angle perturbation signal.

8. The method of claim 7, wherein the tracking angle is adjusted based on the fluctuation of the light transmittance of the photovoltaic panel. The method further comprises the following steps: dynamically setting the amplitude of the periodic angle perturbation signal according to a noise floor corresponding to the injection frequency; and setting the gain coefficient and a minimum perturbation amplitude in consideration of the preset gain coefficient and the minimum perturbation amplitude.

9. The method of claim 5, wherein the method is based on the fluctuation of the light transmittance of the dust accumulated on the photovoltaic panel. The synchronous demodulation processing in S7 further comprises: generating a double-frequency reference signal with a frequency being twice the injection frequency inside the controller; multiplying the total output power signal and the double-frequency reference signal, and performing low-pass filtering processing on the multiplication result to obtain a nonlinear diagnosis signal.

10. The method of tracking angle adjustment based on fluctuation of light transmittance of a photovoltaic panel due to dust accumulation according to claim 9, wherein, The method further comprises the following steps: comparing the amplitude of the nonlinear diagnosis signal with a preset nonlinear threshold value; when the amplitude of the nonlinear diagnosis signal is greater than the nonlinear threshold value, triggering control mode switching, and suspending the compensation angle updating based on the angle mismatch error signal in S8, and instead performing a global scanning optimization within a preset angle range to avoid a local maximum value.

Images