Method for analyzing photovoltaic module performance degradation trend and evaluating residual life

Abstract

This invention relates to the field of photovoltaic module testing technology and discloses a method for analyzing the performance degradation trend and assessing the remaining lifetime of photovoltaic modules. The method includes: acquiring the terminal voltage and terminal current sequences during power conversion perturbations; determining the terminal voltage difference and injected charge amount during transient charging intervals, and calculating the dynamic equivalent junction capacitance; reconstructing the parasitic displacement current based on the terminal voltage change rate, and removing the displacement current component from the terminal current sequence to establish the net current sequence; determining the decay time constant of the net current sequence and correcting it with reference to the backsheet temperature to generate an intrinsic degradation index that maps physical damage; calculating the trend acceleration of the intrinsic degradation index and determining the evolution stage, matching the prediction model. This invention, by removing the capacitive response component, eliminates environmental fluctuation interference, establishes the intrinsic degradation state reflecting the failure mechanism, captures the nonlinear inflection point of performance degradation, and enhances the predictability of lifetime assessment.

Description

Technical Field

[0001] This invention belongs to the field of photovoltaic module testing technology, and in particular relates to a method for analyzing the performance degradation trend and assessing the remaining life of photovoltaic modules. Background Technology

[0002] With the continuous growth of photovoltaic power plant installed capacity, the prediction of performance degradation and reliability management of photovoltaic modules during operation have become key links to ensure the profitability of power plants. The mainstream solution usually collects electrical data such as output power, terminal voltage and terminal current of grid-connected inverters, and uses time series models to fit the power degradation trend to estimate the remaining life of the modules. However, such solutions have limitations when dealing with complex weather conditions. Because the transient power drift caused by irradiance fluctuations and temperature changes is deeply coupled with the intrinsic degradation characteristics of materials in the electrical response, it is difficult to extract signals reflecting the intrinsic physical degradation of semiconductors by relying solely on external power characteristics monitoring methods.

[0003] To address the limitations of steady-state data, voltage perturbations during the inverter's maximum power point tracking (MPPT) process are used to probe the transient carrier relaxation characteristics of the photovoltaic (PV) module. Internal lattice defects are characterized by analyzing the current decay constant after the voltage perturbation. However, in real power plant operating environments, due to the voltage regulation effect of the DC bus capacitor at the grid-connected inverter's front end and the influence of the control loop bandwidth, the voltage perturbation applied across the PV module exhibits a non-zero voltage change rate perturbation edge. This voltage perturbation edge excites a charge-discharge displacement current in the space charge region within the PV module, which physically superimposes with the composite conduction current reflecting the intrinsic lifetime. Conventional methods, when extracting electrical characteristics, cannot access the physical coupling mechanism between electric field distortion and charge transport, making it difficult to accurately separate the intrinsic current component using existing grid-connected circuits.

[0004] Therefore, the technical problem to be solved by this invention is how to accurately strip away the parasitic electrical components inside the photovoltaic module and extract the intrinsic parameters reflecting the degradation of the semiconductor surface by utilizing the disturbance excitation under grid-connected operation. Summary of the Invention

[0005] This technical solution provides a method for analyzing the performance degradation trend and assessing the remaining lifetime of photovoltaic modules, comprising the following steps: Step S1: Obtain the timing sequence of the photovoltaic module terminal voltage and the timing sequence of the terminal current, which are controlled by the perturbation action of the power conversion regulation unit. Step S2: Extract the voltage difference of the terminal voltage time series within the transient charging interval, and calculate the time domain integral of the terminal current time series within the transient charging interval to determine the amount of transient injected charge. Step S3: Calculate the ratio of transient injected charge to voltage difference to determine the dynamic equivalent junction capacitance of the photovoltaic module under the current surface aging state; Step S4: Calculate the rate of change of the terminal voltage time sequence relative to time, and determine the parasitic displacement current sequence by multiplying the rate of change of voltage by the dynamic equivalent junction capacitance. Step S5: Calculate the difference between the original acquired terminal current time series and the reconstructed parasitic displacement current series, and establish the net current series characterizing the minority carrier lifetime state after removing the capacitive response component. Step S6: Extract the transient decay time constant from the net current sequence, and perform temperature correction calibration on the transient decay time constant based on the synchronously acquired backsheet temperature to generate an intrinsic degradation index that maps the degree of physical damage in the depletion region of the PN junction inside the photovoltaic module. Step S7: Calculate the trend acceleration of the intrinsic degradation index within the sliding time window, and switch the prediction mode of the lifetime prediction model based on the determination results of the trend acceleration and the preset nonlinear evolution inflection point threshold.

[0006] Preferably, step S2 specifically includes: step S21, identifying the voltage step start point and transient charging termination point triggered by the perturbation action of the power conversion adjustment unit; step S22, within the transient charging interval, performing time integration on the terminal current after deducting the steady-state leakage current component to establish the transient injected charge amount.

[0007] Preferably, the temperature correction calibration in step S6 includes: obtaining the carrier mobility temperature compensation coefficient of the semiconductor material corresponding to the photovoltaic module; calculating the product of the carrier mobility temperature compensation coefficient and the transient decay time constant to eliminate the transient influence of the thermal excitation effect caused by irradiance fluctuation on the carrier concentration.

[0008] Preferably, the calculation of trend acceleration in step S7 includes: step S71, selecting a time period covering no less than 5 power conversion adjustment unit perturbation cycles as a sliding time window; step S72, extracting the rate of change of intrinsic degradation index within the sliding time window, and differentiating the rate of change again to generate trend acceleration.

[0009] Preferably, step S1 specifically includes: step S11, using the high-frequency sampling module of the power conversion adjustment unit to capture perturbation transient edge data at a sampling frequency of not less than 100kHz; step S12, performing median filtering on the captured data to eliminate common-mode noise generated by power switch switching.

[0010] Preferably, the prediction mode of the lifetime prediction model in step S7 includes: maintaining the linear degradation assessment model when the trend acceleration is lower than or equal to the nonlinear evolution inflection point threshold; and determining that the photovoltaic module has entered the accelerated degradation stage and switching to the nonlinear lifetime prediction model based on the exponential evolution law when the trend acceleration continues to exceed the nonlinear evolution inflection point threshold for 3 sliding time window periods.

[0011] Preferably, the nonlinear lifetime prediction model includes logic for capturing the breakpoints of physical damage caused by yellowing of the packaging material and water barrier layer penetration failure.

[0012] Preferably, the method further includes: comparing intrinsic degradation indices with the initial health baseline of the storage; and determining the degree of deviation of the photovoltaic module from the initial energy conversion efficiency.

[0013] Preferably, the method further includes: outputting an assessment result indicating the remaining safe operating years of the photovoltaic module; and generating maintenance warning instructions for the photovoltaic module when switching to a nonlinear life prediction model.

[0014] Compared with existing technologies, the photovoltaic module performance degradation trend analysis and remaining life assessment method of the present invention has the following advantages: 1. In the analysis of photovoltaic module performance degradation trend, the voltage perturbation mutation point is determined based on the activation threshold of the second derivative of the terminal voltage sequence. This is then used as the starting anchor point to extract the transient current relaxation sequence. This enables the processing unit to automatically identify the effective perturbation caused by the inverter's maximum power point tracking action, effectively filtering out interference from low-frequency fluctuations on the grid side and random environmental noise. This ensures that the subsequently extracted charge transport features all originate from a defined excitation source, improving the determinism and consistency of feature extraction under complex operating conditions.

[0015] 2. Utilizing the charge pumping effect generated by the voltage differential ramp edge, the dynamic equivalent junction capacitance is calibrated in situ by performing time-domain integration on the current within the first time-series sub-window. The parasitic displacement current is then reconstructed based on the product of the junction capacitance and the time-series derivative of the terminal voltage. This process removes the capacitive response components caused by the inverter DC bus capacitance and perturbation edge passivation from the original sampled signal, thereby outputting a net conduction relaxation sequence that is not affected by external hardware parasitic parameters. This allows the transient decay time constant to directly reflect the intrinsic carrier recombination rate of the semiconductor inside the photovoltaic module.

[0016] 3. By combining the transient decay time constant, which is free from the influence of external impedance, with the synchronously acquired backplane temperature, and by calling the semiconductor carrier mobility temperature compensation coefficient to perform normalization calibration, the transient influence of thermal excitation effect caused by irradiance fluctuation on carrier concentration is eliminated. This allows the irreversible physical degradation caused by the evolution of lattice defect density to be extracted, thereby achieving accurate anchoring of the minority carrier lifetime state in the depletion region of the PN junction of the photovoltaic module. This solves the technical problem in the existing technology that steady-state electrical parameters are easily obscured by environmental fluctuations and cannot reflect intrinsic decay. Attached Figure Description

[0017] Figure 1 This is a flowchart of the component performance degradation analysis and lifetime prediction algorithm of the present invention; Figure 2 This is the logical architecture diagram of the multidimensional feature perception and evolution analysis of this invention. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0019] It should be noted that all directional and positional terms used in this invention, such as: up, down, left, right, front, back, vertical, horizontal, inner, outer, top, bottom, transverse, longitudinal, center, etc., are only used to explain the relative positional relationship and connection between components in a specific state (as shown in the accompanying drawings). They are only for the convenience of describing this invention and do not require that this invention be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention. In addition, the descriptions of "first," "second," etc., in this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated.

[0020] In the description of this invention, unless otherwise explicitly specified and limited, the terms installation, connection, and linking should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections; they can refer to direct connections or indirect connections through an intermediate medium; they can refer to the internal connection of two components. For those skilled in the art, the specific meaning of the above terms in this invention can be understood according to the specific circumstances.

[0021] In the description of this specification, references to the terms "an embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example, and the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0022] A method for analyzing the performance degradation trend and assessing the remaining lifetime of photovoltaic modules includes the following steps: Step S1: Obtain the timing sequence of the photovoltaic module terminal voltage and the timing sequence of the terminal current, which are controlled by the perturbation action of the power conversion regulation unit. Step S2: Extract the voltage difference of the terminal voltage time series within the transient charging interval, and calculate the time domain integral of the terminal current time series within the transient charging interval to determine the amount of transient injected charge. Step S3: Calculate the ratio of transient injected charge to voltage difference to determine the dynamic equivalent junction capacitance of the photovoltaic module under the current surface aging state; Step S4: Calculate the rate of change of the terminal voltage time sequence relative to time, and determine the parasitic displacement current sequence by multiplying the rate of change of voltage by the dynamic equivalent junction capacitance. Step S5: Calculate the difference between the original acquired terminal current time series and the reconstructed parasitic displacement current series, and establish the net current series characterizing the minority carrier lifetime state after removing the capacitive response component. Step S6: Extract the transient decay time constant from the net current sequence, and perform temperature correction calibration on the transient decay time constant based on the synchronously acquired backsheet temperature to generate an intrinsic degradation index that maps the degree of physical damage in the depletion region of the PN junction inside the photovoltaic module. Step S7: Calculate the trend acceleration of the intrinsic degradation index within the sliding time window, and switch the prediction mode of the lifetime prediction model based on the determination results of the trend acceleration and the preset nonlinear evolution inflection point threshold.

[0023] Preferably, step S2 specifically includes: step S21, identifying the voltage step start point and transient charging termination point triggered by the perturbation action of the power conversion adjustment unit; step S22, within the transient charging interval, performing time integration on the terminal current after deducting the steady-state leakage current component to establish the transient injected charge amount.

[0024] Preferably, the quantization logic for determining the parasitic displacement current sequence in step S4 follows the following formula: ,in, The instantaneous displacement current in the parasitic displacement current sequence. Let u(t) be the dynamic equivalent junction capacitance determined in step S3, u(t) be the instantaneous terminal voltage in the terminal voltage time sequence, and t be the sampling time.

[0025] Preferably, the temperature correction calibration in step S6 includes: obtaining the carrier mobility temperature compensation coefficient of the semiconductor material corresponding to the photovoltaic module; calculating the product of the carrier mobility temperature compensation coefficient and the transient decay time constant to eliminate the transient influence of the thermal excitation effect caused by irradiance fluctuation on the carrier concentration.

[0026] Preferably, the calculation of trend acceleration in step S7 includes: step S71, selecting a time period covering no less than 5 power conversion adjustment unit perturbation cycles as a sliding time window; step S72, extracting the rate of change of intrinsic degradation index within the sliding time window, and differentiating the rate of change again to generate trend acceleration.

[0027] Preferably, step S1 specifically includes: step S11, using the high-frequency sampling module of the power conversion adjustment unit to capture perturbation transient edge data at a sampling frequency of not less than 100kHz; step S12, performing median filtering on the captured data to eliminate common-mode noise generated by power switch switching.

[0028] Preferably, the prediction mode of the lifetime prediction model in step S7 includes: maintaining the linear degradation assessment model when the trend acceleration is lower than or equal to the nonlinear evolution inflection point threshold; and determining that the photovoltaic module has entered the accelerated degradation stage and switching to the nonlinear lifetime prediction model based on the exponential evolution law when the trend acceleration continues to exceed the nonlinear evolution inflection point threshold for 3 sliding time window periods.

[0029] Preferably, the nonlinear lifetime prediction model includes logic for capturing the breakpoints of physical damage caused by yellowing of the packaging material and water barrier layer penetration failure.

[0030] Preferably, the method further includes: comparing intrinsic degradation indices with the initial health baseline of the storage; and determining the degree of deviation of the photovoltaic module from the initial energy conversion efficiency.

[0031] Preferably, the method further includes: outputting an assessment result indicating the remaining safe operating years of the photovoltaic module; and generating maintenance warning instructions for the photovoltaic module when switching to a nonlinear life prediction model.

[0032] Example 1: In the specific operating condition of a megawatt-level grid-connected photovoltaic power station, where the photovoltaic modules are deployed in a desert environment with extreme high temperatures and severe irradiance fluctuations, the system faces the challenge of voltage perturbation edge passivation caused by the voltage stabilization effect of the inverter's DC bus capacitor. This non-ideal step excitation induces a space charge region charge-discharge displacement current within the photovoltaic module, which physically superimposes with the composite conduction current characterizing the intrinsic lifetime of the semiconductor. Traditional steady-state power fitting methods struggle to isolate this electrical component without adding external detection hardware. The processing unit uses a high-frequency sampling module of the power conversion adjustment unit at a frequency of not less than 100kΩ. The sampling frequency of Hz acquires the timing sequence of the photovoltaic module terminal voltage and terminal current, which are controlled by the perturbation action of the power conversion regulation unit. When the system detects the voltage fluctuation generated by the inverter performing the maximum power point tracking task, it calculates the timing second derivative of the terminal voltage timing sequence and compares it with the preset second derivative activation threshold. When the second derivative exceeds the activation threshold, the voltage perturbation mutation point is determined as the starting anchor point, and the transient current relaxation sequence with a duration of Δt is extracted. This step automatically identifies the effective perturbation trigger time through the nonlinear mutation characteristics of the second derivative, eliminating the interference of low-frequency fluctuations on the sampling starting point of the grid side.

[0033] In the extracted transient current relaxation sequence, the processing unit identifies the voltage step start point triggered by the perturbation action of the power conversion regulation unit and the transient charging termination point. It extracts the voltage difference in the terminal voltage time sequence within the transient charging interval. The system calculates the time-domain integral of the terminal current within this interval (after deducting the steady-state leakage current component) to establish the transient injected charge. It then calculates the ratio of the transient injected charge to the voltage difference, calibrating the dynamic equivalent junction capacitance of the photovoltaic module under its current surface aging state in situ. This process transforms the perturbation edge passivation phenomenon, originally considered interference, into a probe, deconstructing the capacitive response characteristics inside the module through the charge pumping effect. This is done before performing the aforementioned terminal current deduction operation. At this time, the processing unit retrieves the average terminal current sampling value within the 10 power frequency steady-state cycles prior to the voltage perturbation trigger through the backtracking memory, using it as the reference for the steady-state leakage current component. In the mathematical operation of this time-domain integration, this reference value is subtracted from the transient terminal current data at each sampling point within the transient charging interval. This eliminates the systemic offset interference caused by the background DC leakage current of the grid-connected circuit on the cumulative calculation of transient injected charge. After obtaining the dynamic equivalent junction capacitance, the processing unit calculates the voltage change rate of the terminal voltage time series relative to time, and determines the parasitic displacement current series by multiplying the voltage change rate by the dynamic equivalent junction capacitance. Its quantization logic follows the following formula: ,in, The instantaneous displacement current in the parasitic displacement current sequence. The system calculates the difference between the original acquired terminal current time sequence and the reconstructed parasitic displacement current sequence. By performing differential subtraction, the capacitive response component is removed, and the net current sequence characterizing the minority carrier lifetime state is established. This mechanism eliminates the physical masking of the current relaxation waveform by the inverter front-end DC bus capacitance and external impedance, so that the subsequently extracted transient decay time constant directly reflects the physical damage degree of the PN junction depletion region inside the component.

[0034] The transient decay time constant is extracted from the net current sequence, and the product of the transient decay time constant is calculated using the built-in semiconductor carrier mobility temperature compensation coefficient based on the synchronously acquired backplane temperature. According to the semiconductor lattice phonon scattering theory, carrier mobility exhibits a nonlinear decay with increasing lattice temperature. The temperature compensation coefficient logic follows the formula: ,in, This is the temperature compensation coefficient for semiconductor carrier mobility. To simultaneously collect the absolute temperature of the photovoltaic module backsheet, To store the absolute temperature of the standard test reference, parameter m is a dimensionless empirical constant determined by the lattice phonon scattering mechanism, extracted to a value between 1.5 and 2.7 based on a pre-calibrated set of physical property parameters of monocrystalline silicon material. The processing unit multiplies the extracted transient decay time constant with the calculated temperature compensation coefficient, and uses mathematical relationships to directly remove the transient interference component of the thermally excited layer on the carrier concentration, generating an intrinsic degradation index that maps the internal physical damage of the photovoltaic module. The system calculates the trend acceleration of this intrinsic degradation index within a sliding time window, which covers no less than 5 power conversion adjustment unit perturbation cycles. When the trend acceleration continuously exceeds the preset nonlinear evolution inflection point threshold for 3 sliding time window cycles, the module is determined to have entered the accelerated decay stage, and the system automatically evaluates the linear decay. The model is switched to a nonlinear lifetime prediction model based on exponential evolution and outputs the assessment results of the remaining safe operating years. Yellowing of the encapsulation material of photovoltaic modules leads to strong absorption of short-wavelength photons in the transmission spectrum, causing a nonlinear drop in the injection rate of photogenerated carriers in the shallow region of the PN junction. Meanwhile, the failure of the water barrier layer allows trace amounts of external water vapor to invade the surface of the crystalline silicon, triggering chemical hydrolysis of the passivation layer and inducing a large number of interface dangling bonds, which act as high-density nonradiative recombination centers. Both of these overall encapsulation failure mechanisms modulate the recombination rate of minority carriers on the surface, which is reflected in the surface electrical level as a cliff drop in the transient decay time constant. This allows the system to accurately capture and verify the accelerated decay inflection point of the physical encapsulation structure based solely on the abnormal abrupt change curve of the intrinsic electrical degradation index.

[0035] Example 2: In this example, a cumulative irradiance of 2500 kWh / m² is selected. 2Furthermore, for monocrystalline silicon photovoltaic modules with a service life of up to 12 years, the test platform utilizes a programmable DC electronic load system to simulate dynamic impedance characteristics, and combines this with a synchronous data acquisition module with a sampling accuracy of 16 bits and a sampling rate of 250kHz to acquire electrical fingerprint data. The data originates from the physical experimental platform, whose voltage measurement resolution is better than 0.1mV, meeting the requirement of capturing microsecond-level voltage step edges. Regarding the sampling period... The setting is controlled by the carrier recombination relaxation time constant and the switching ripple frequency of the power conversion adjustment unit, due to the sampling period. It is necessary to balance signal reconstruction fidelity with the computational load of the processing unit. When the carrier relaxation time is shortened, the sampling period needs to be adjusted to reduce aliasing error when extracting the decay time constant. The sampling period is determined for this specific working condition, and its value tends towards the lower limit of its range. With a time window of 4μs, it is achieved within the timing window. The transient charging range is covered by 50 sampling points to achieve dynamic equivalent junction capacitance. To simulate electromagnetic interference in an industrial environment, the closed loop of the computational logic injects Gaussian noise with a signal-to-noise ratio of 25dB into the terminal current timing sequence.

[0036] The experimental group established three comparison systems: the control group used a method based on steady-state output power fitting; the experimental group of this invention used a net current analysis method after removing parasitic displacement current; and the partially missing control group skipped the dynamic equivalent junction capacitance. The real-time calculation steps, using only the nominal capacitance value of the components at the factory, show that when the power conversion regulation unit generates a non-ideal voltage excitation with an edge passivation slope of 50V / ms, the control group produces an 18.6% residual in remaining lifetime assessment. The present invention's sample group identifies the voltage step start point and the transient charge termination point, calculates the transient injected charge, divides it by the voltage difference, and obtains the dynamic equivalent junction capacitance through in-situ calibration. It is 4.62 μF.

[0037] Goodness of fit between the net current sequence after removing the capacitive response component and the waveform of the physical evolution law The value reached 0.992. In contrast, the partially missing control group, due to its failure to respond to the capacitance evolution caused by the narrowing of the junction width during aging, showed a deviation of 35.2% between the nominal and measured values, resulting in a reconstructed parasitic displacement current sequence. Compensation has been generated, among which The instantaneous displacement current caused the final evaluation deviation rate to rise to 12.4%. Boundary verification of the sampling frequency showed that when the sampling frequency was reduced from 100kHz to below the performance inflection point of 10kHz, it was unable to capture the rate of voltage change. High-frequency transient characteristics, dynamic equivalent junction capacitance The calculation error increased from 2.1% nonlinearity to 45.8%, of which The error, which represents the rate of change of voltage over time, causes phase distortion in the net current sequence after stripping, suppressing the extraction of the transient decay time constant reflecting the physical damage of the PN junction. However, within the window of 100kHz to 250kHz, the extraction accuracy of the intrinsic degradation index tends to stabilize, and the sensitivity to environmental radiation fluctuations decreases by 72.5%. The experimental results confirm that in-situ calibration of the dynamic equivalent junction capacitance can offset non-ideal perturbation losses.

[0038] Example 3: In the operation and maintenance scenario of photovoltaic power plants deployed in high-altitude environments with strong ultraviolet radiation, monocrystalline silicon photovoltaic modules face the risk of nonlinear failure caused by photodegradation of encapsulation materials. Due to individual differences in initial minority carrier lifetime and junction capacitance baselines among modules from different manufacturing batches, the system faces the challenge of establishing degradation criteria based on the individual physical properties of the modules. The processing unit uses a calibration procedure based on statistical distribution to determine the threshold of the nonlinear evolution inflection point. Its initial state is defined as the intrinsic degradation index sequence within the first month of grid-connected operation of the photovoltaic power plant. This sequence reflects the intrinsic physical baseline of the module when no structural damage has occurred. The system obtains the degradation rate distribution by calculating the first difference of this sequence and uses a Gaussian kernel function to estimate the probability density of the degradation rate, selecting the rate value corresponding to the 99.7% cumulative probability distribution as the background. The upper limit of scene noise is determined, and a multiplicative weighting is performed based on the band distortion coefficient of the monocrystalline silicon semiconductor to establish the nonlinear evolution inflection point threshold of the specific component. This procedure anchors the judgment on the initial physical property response of the individual component, eliminating the interference of manufacturing tolerances on the lifetime assessment logic. In actual operation, the band distortion coefficient is obtained by the back-end server calling a pre-established material property database. The values ​​in this database are derived by measuring the dimensionless ratio of the actual bandgap width shrinkage to the ideal pure silicon lattice bandgap in the same batch of monocrystalline silicon cells under standard accelerated ultraviolet aging tests in a laboratory environment. Its conventional value range is marked between 0.85 and 0.92. By performing multiplicative compensation of this coefficient, the system can accurately correct the nonlinear evolution error of complex doped components in the field that deviates from the ideal physical reference.

[0039] For the analytical process of the net current sequence, the processing unit uses the Levenberg-Marquardt nonlinear least squares algorithm to extract the transient decay time constant τ. This algorithm minimizes the sum of squared residuals between the measured net current sequence and the physical evolution model through iterative optimization. Its mathematical mapping logic follows the following formula: Where j(t) is the current density of the net current sequence at sampling time t, A is the proportionality coefficient related to the concentration of photogenerated carriers, B is the steady-state bias term reflecting the characteristics of dark current, and τ is the transient decay time constant characterizing minority carrier lifetime. The system sets the iteration termination condition as the relative change rate of τ in two adjacent calculations being less than 0.1%, thus outputting an intrinsic degradation index with physical determinism. This step realizes the white-box extraction of the minority carrier recombination dynamics process in the depletion region of the PN junction. When the trend acceleration of the intrinsic degradation index is detected to exceed the nonlinear evolution inflection point threshold, the system switches to a lifetime prediction model based on exponential evolution. The processing unit fits the defect evolution rate function according to the second-order time derivative of the current intrinsic degradation index and applies the physical damage accumulation exponential dynamic evolution model. The mathematical expression logic of the model follows the formula: Where P(t) is the predicted instantaneous output power of the photovoltaic module. To determine the baseline output power at the starting point of the accelerated degradation stage of a photovoltaic module, parameter λ is the initial damage coefficient characterizing the aging start of the encapsulation material, parameter γ is the nonlinear degradation rate exponent, and parameter t is the cumulative operating time calculated since entering the accelerated degradation stage. The absolute value of the trend acceleration during trigger mode switching is extracted as the numerical input of parameter γ. The instantaneous output power of the target state is set to 80% of the initial nominal power and substituted into the left side of the formula. The final state quantization value of the time parameter t is obtained by solving the evolution model equation. The remaining time for the module power output to drop to 80% of the initial value is calculated and determined as the remaining lifetime assessment value. This prediction method utilizes the nonlinear characteristics of defect evolution to capture the yellowing conductivity of the encapsulation material. The accelerated decay inflection point enables the power plant operation and maintenance system to output predictive conclusions with material logic support for different aging chemical conditions. To resolve the conflict of physical dimensions in the above parameter extraction process, before extracting the absolute value of the trend acceleration as the actual input parameter γ, the processing unit automatically calls the preset time normalization constant to perform product calibration. The scalar value of this constant is set to 1, and its physical dimension is time unit. By multiplying it with the trend acceleration whose dimension includes the negative square of time, it is seamlessly transformed into the decay rate exponent that meets the dimensionless requirement of the independent variable of the exponential model. This ensures the closed loop of the algorithm logic and mathematical self-consistency when substituting the overall trend statistics into the surface exponential dynamic evolution.

[0040] Example 4: In the on-site deployment of a newly built grid-connected photovoltaic power station, the modulation waveform differences of the power conversion regulation unit and the parasitic parameters of the DC bus exhibit random characteristics. The processing unit initiates the pre-calibration procedure on the first complete sunny day of grid-connected operation. The system acquires the no-load disturbance sequence of the power conversion regulation unit during the irradiance stabilization period, calculates the probability distribution characteristics of the first and second time series derivatives of the terminal voltage time series, and uses the maximum likelihood estimation method to extract the amplitude envelope of the background electromagnetic noise. 1.5 times the amplitude envelope is established as the second-order derivative specific to this site. The technical basis for setting the activation threshold at 1.5 times is that, according to statistical verification, the parasitic switching ripple generated by the cross-coupling between the high-frequency hardware on site and the grid side follows a Gaussian distribution in the no-load time sequence. Based on the three sigma statistical principle of normal distribution, 1.5 times the amplitude envelope is selected as the upper limit boundary for judgment. This can cover more than 99.7% of the extreme values ​​of background random fluctuations in mathematical probability, thereby ensuring that the activation threshold maintains a high sensitivity to non-ideal micro-perturbation steps caused by the tracking action of the true maximum power point while effectively isolating common-mode noise false triggering.

[0041] When the processing unit initiates the nonlinear least squares algorithm to extract minority carrier lifetime indicators, the system establishes a parameter initialization model based on the first set of perturbation response data collected on-site. This model extracts the mean terminal current of the transient charging interval as the initial input value of the proportional coefficient A, where A is the proportional coefficient. It also maps the steady-state current component at the end of the transient current relaxation sequence to the initial reference value of the steady-state bias term B, where B is the steady-state bias term. At the same time, it uses the logarithmic slope of the net current sequence to estimate the order of magnitude of the transient decay time constant τ. The system constructs an initial vector based on real-time physical measurements to guide the iterative optimization process and calculates periodic corrections to the sliding prior value according to changes in ambient temperature to extract component aging characteristics.

[0042] Example 5: In the field commissioning scenario of a distributed photovoltaic rooftop system, the processing unit faces the risk of sampling aliasing caused by the difference in switching frequencies of different models of power conversion and regulation units. To establish a sampling frequency suitable for the specific hardware topology... Including the sliding time window length, the system initiates the pre-frequency scanning procedure. After the photovoltaic module reaches a quasi-steady state, the processing unit triggers the step action of the power conversion regulation unit. The high-frequency sampling module captures the spectral energy distribution during the voltage step process and identifies the dominant switching frequency of the power conversion regulation unit. And the energy attenuation inflection point of higher harmonic components, the sampling frequency Established to be no less than the dominant switching frequency It is 10 times that of the time required to calculate the physical settling time required for the transient voltage to evolve from the initial point to 99% of the steady-state value. ,in, Sampling frequency, To dominate the switching frequency, To adjust the physical adjustment time, the sliding time window length is set to 1.2 times the physical adjustment time. A data capture boundary for specific energy conversion environments is established at the hardware level. When the system detects a variable operating condition where the fluctuation rate of environmental irradiance exceeds 10% within a preset sampling interval, the processing unit calls the dynamic gain self-calibration program. By calculating the peak difference of transient charging current under different irradiance levels, a dynamic equivalent junction capacitance is generated. The optical drift compensation coefficient β, where, β is the dynamic equivalent junction capacitance and β is the nonlinear modulation coefficient. This procedure utilizes the physical compression effect of photogenerated carriers on the depletion region width to establish a mapping between the measured current response and the calibration curve in the register. In the logic operation of stripping parasitic displacement current, the system calculates amplitude weighted correction for the reconstructed displacement current sequence based on real-time irradiance data, eliminating the minority carrier lifetime assessment drift caused by cloud cover, and keeping the deviation of the generated intrinsic degradation index within the technical error range of 0.5% under different meteorological conditions.

[0043] The processing unit performs discrete second-order difference operations on the generated intrinsic degradation index sequence to establish the trend acceleration. The system calculates the difference between the intrinsic degradation index at the current sampling time and the index at the previous sampling time to obtain the first-order degradation rate of the index. The first-order degradation rate is then subjected to difference calculation again and divided by the square of the perturbation interval to establish the trend acceleration characterizing the drastic change of the index. This procedure transforms the continuous physical process into discretized numerical calculation logic, and its quantization logic follows the following formula: Where a(n) is the trend acceleration at time n, X(n) is the intrinsic degradation index at time n, and ΔT is the triggering period of the perturbation action of the power conversion adjustment unit. The processing unit injects the calculated trend acceleration into the judgment loop and compares it with the nonlinear evolution inflection point threshold pre-stored in the register. When the trend acceleration exceeds the threshold for three consecutive sliding time window periods, the system determines that the photovoltaic module has entered the accelerated failure stage. The processing unit triggers the model switching instruction. In the accelerated degradation mode, the lifetime prediction model calculates the curvature of the power degradation trajectory based on the acceleration of the current intrinsic degradation index and predicts the remaining service time when the output power of the photovoltaic module drops to 80% of the nominal value.

[0044] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. A method for analyzing the performance degradation trend and assessing the remaining life of photovoltaic modules, characterized in that, Includes the following steps: Step S1: Obtain the timing sequence of the photovoltaic module terminal voltage and the timing sequence of the terminal current, which are controlled by the perturbation action of the power conversion regulation unit. Step S2: Extract the voltage difference of the terminal voltage time series within the transient charging interval, and calculate the time domain integral of the terminal current time series within the transient charging interval to determine the amount of transient injected charge. Step S3: Calculate the ratio of transient injected charge to voltage difference to determine the dynamic equivalent junction capacitance of the photovoltaic module under the current surface aging state; Step S4: Calculate the rate of change of the terminal voltage time sequence relative to time, and determine the parasitic displacement current sequence by multiplying the rate of change of voltage by the dynamic equivalent junction capacitance. Step S5: Calculate the difference between the original acquired terminal current time series and the reconstructed parasitic displacement current series, and establish the net current series characterizing the minority carrier lifetime state after removing the capacitive response component. Step S6: Extract the transient decay time constant from the net current sequence, and perform temperature correction calibration on the transient decay time constant based on the synchronously acquired backsheet temperature to generate an intrinsic degradation index that maps the degree of physical damage in the depletion region of the PN junction inside the photovoltaic module. Step S7: Calculate the trend acceleration of the intrinsic degradation index within the sliding time window, and switch the prediction mode of the lifetime prediction model based on the determination results of the trend acceleration and the preset nonlinear evolution inflection point threshold.

2. The method for analyzing the performance degradation trend and assessing the remaining life of a photovoltaic module according to claim 1, characterized in that, Step S2 specifically includes: Step S21, identifying the voltage step start point and transient charging termination point triggered by the perturbation action of the power conversion regulation unit; Step S22, within the transient charging interval, performing time integration on the terminal current after deducting the steady-state leakage current component to establish the transient injected charge amount.

3. The method for analyzing the performance degradation trend and assessing the remaining life of a photovoltaic module according to claim 1, characterized in that, The temperature correction calibration in step S6 includes: obtaining the carrier mobility temperature compensation coefficient of the semiconductor material corresponding to the photovoltaic module; calculating the product of the carrier mobility temperature compensation coefficient and the transient decay time constant to eliminate the transient influence of the thermal excitation effect caused by irradiance fluctuation on the carrier concentration.

4. The method for analyzing the performance degradation trend and assessing the remaining life of a photovoltaic module according to claim 1, characterized in that, The calculation of trend acceleration in step S7 includes: step S71, selecting a time period covering no less than 5 power conversion adjustment unit perturbation cycles as a sliding time window; step S72, extracting the rate of change of intrinsic degradation index within the sliding time window, and differentiating the rate of change again to generate trend acceleration.

5. The method for analyzing the performance degradation trend and assessing the remaining life of a photovoltaic module according to claim 1, characterized in that, Step S1 specifically includes: Step S11, using the high-frequency sampling module of the power conversion adjustment unit to capture the transient edge data of the perturbation at a sampling frequency of not less than 100kHz; Step S12, performing median filtering on the captured data to eliminate common-mode noise generated by power switch switching.

6. The method for analyzing the performance degradation trend and assessing the remaining life of a photovoltaic module according to claim 1, characterized in that, The prediction mode of the lifetime prediction model in step S7 includes: maintaining the linear degradation assessment model when the trend acceleration is lower than or equal to the nonlinear evolution inflection point threshold; and determining that the photovoltaic module has entered the accelerated degradation stage and switching to the nonlinear lifetime prediction model based on the exponential evolution law when the trend acceleration continues to exceed the nonlinear evolution inflection point threshold for 3 sliding time window periods.

7. The method for analyzing the performance degradation trend and assessing the remaining life of a photovoltaic module according to claim 6, characterized in that, The nonlinear lifetime prediction model includes logic for capturing the breakpoints of physical damage caused by yellowing of the packaging material and water barrier layer penetration failure.

8. The method for analyzing the performance degradation trend and assessing the remaining life of a photovoltaic module according to claim 1, characterized in that, The method also includes: comparing intrinsic degradation indices with the initial health baseline of the storage; and determining the degree of deviation of the photovoltaic module from its initial energy conversion efficiency.

9. The method for analyzing the performance degradation trend and assessing the remaining life of a photovoltaic module according to claim 6, characterized in that, The method also includes: outputting assessment results indicating the remaining safe operating years of photovoltaic modules; and generating maintenance warning instructions for photovoltaic modules when switching to a nonlinear life prediction model.

Images