Laminated solar cell performance prediction and band gap parameter optimization method and system

Abstract

The invention discloses a laminated solar cell performance prediction and band gap parameter optimization method and system. The method comprises the following steps: acquiring solar spectrum time sequence data of at least one geographic position and including a plurality of moments; through an innovative'pre-calculation-re-query 'mode, pre-calculation is performed on each solar spectrum in the time sequence data to generate a basic parameter atlas of a preset unijunction band gap energy range, a plurality of to-be-evaluated laminated cell band gaps are combined, and performance parameters of each sub-cell are obtained by querying the basic parameter atlas. Calculating the instantaneous power at each time point; and finally, carrying out time integration on the instantaneous power of each laminated solar cell band gap combination to obtain the annual total generating capacity, and determining the optimal band gap combination according to the annual total generating capacity. According to the method, the band gap parameter of the laminated solar cell can be efficiently and accurately optimized based on the real dynamic spectrum data, so that the result is closer to the actual application.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of solar cell design, and more particularly relates to a method for predicting the performance of a stacked solar cell and optimizing bandgap parameters. BACKGROUND

[0002] Stacked solar cells can effectively utilize photons in different energy ranges of the solar spectrum by stacking sub-cells with different bandgaps (Eg), thereby breaking through the theoretical efficiency limit of single-junction solar cells and becoming a key technology for realizing next-generation high-efficiency photovoltaic power generation. In the design of stacked solar cells, the bandgap combination of the top and bottom sub-cells is a core parameter that determines the photoelectric conversion efficiency: the bandgap matching degree directly affects key performance indicators such as photon absorption efficiency, carrier separation efficiency, and open-circuit voltage. Therefore, accurately finding and evaluating the optimal bandgap combination is a core link for pushing stacked solar cells from theoretical design to practical high-performance application.

[0003] However, the existing technology has the following significant defects in the process of finding and evaluating the optimal bandgap combination:

[0004] On the one hand, the simulation scenario is idealized, and the simulation results are seriously out of line with actual working conditions: Currently, the industry generally uses a single standard solar spectrum (such as AM1.5G) to simulate and determine the optimal bandgap. However, in actual applications, the spectrum received by solar cells will constantly change due to factors such as geographical location, season, weather, and day and night time. Using a single static standard spectrum for design, the results cannot accurately reflect the true power generation performance of the cell under dynamic changes throughout the year, resulting in a large deviation between theoretical design and actual performance.

[0005] On the other hand, the calculation efficiency is low and difficult to support large-scale precise optimization: In order to more accurately evaluate the performance of stacked solar cells in different regions around the world, real spectrum data at an hourly level or even higher resolution throughout the year needs to be introduced, and different bandgap combinations within a wide range need to be scanned, which requires a huge amount of calculation. Traditional simulation methods usually use a serialized nested loop calculation method, and in the face of massive data and complex models, the calculation process is extremely time-consuming, often taking hours or even days to complete the simulation of a region, severely restricting the research and development efficiency and the speed of optimization iteration, and unable to meet the needs of rapid technological research and development and industrialization.

[0006] In summary, the existing technology cannot simultaneously meet the core needs of "based on real dynamic working conditions" and "highly efficient and precise calculation", and there is an urgent need to develop a method and system for predicting the performance of stacked solar cells and optimizing bandgap parameters that can integrate real dynamic spectrum data, have high calculation efficiency, and support automated analysis. SUMMARY

[0007] In view of the above defects or improvement needs of the prior art, the present application provides a kind of stacked solar cell performance prediction and band gap parameter optimization method and system, by innovative "pre-computing-requery" mode, the originally high coupling calculation is decoupled into spectrum independent preprocessing and fast table operation, combined with the hybrid architecture of multi-process parallel and JIT technology, the number of most time-consuming integral operation is reduced from N times to N times, greatly improve the calculation efficiency, can break through the bottleneck of traditional serial calculation time-consuming;On the other hand, the present application discards the idealization mode of traditional single static standard spectrum, adopts the real dynamic spectrum data of annual hour level, takes the annual total power generation as the optimization goal, so that the optimization result is more suitable for dynamic working conditions such as geographical location, season, weather in actual application, and reduces the deviation between theoretical and actual power generation performance;The method of the present application is based on ideal SQ theory, but has high expansibility, can adapt to actual device scenarios including non-radiative recombination and other complex losses, and can generate multi-dimensional analysis report, provide customized guidance for stacked cell design in different regions, and finally provide efficient and accurate technical support for high-performance stacked solar cell research and industrialization.

[0008] In order to achieve the above-mentioned purpose, one aspect of the present application provides a kind of stacked solar cell performance prediction and band gap parameter optimization method, including the following steps:

[0009] S1: automatically scan and load standard format containing at least one geographical location, annual hour level solar spectrum time series data file;At the same time, analyze the metadata in the solar spectrum time series data file, convert the solar energy spectrum of each geographical location and each hour level into photon energy domain flux spectrum;

[0010] S2: based on SQ theory, pre-compute and generate key basic parameter atlas covering wide range of single-junction band gap energy for each hour level of photon energy domain flux spectrum processed in step S1;

[0011] S3: for any stacked band gap combination, obtain the key performance parameters of each sub-cell from the key basic parameter atlas by interpolation and fast query, and simulate and calculate the instantaneous performance under different band gap combinations by combining the hybrid parallel computing architecture of multi-process parallel processing and just-in-time compilation technology;

[0012] S4: group and aggregate the annual instantaneous performance data of each stacked solar cell band gap combination, obtain the annual total power generation of each band gap combination, and determine the optimal band gap combination according to the annual total power generation;

[0013] S5: automatically generate performance thermodynamic diagram containing annual power generation and top, bottom cell band gap relationship of different regions and different years, and multi-dimensional analysis report of stacked mismatch current density statistical data at each time under different band gap combinations.

[0014] Further, the metadata in step S1 includes the solar energy spectrum in the wavelength domain at minute or hour level under actual geographical conditions, the geographical position of the stacked solar cell application, and the time zone.

[0015] Further, in step S1, the input solar energy spectrum S λ (λ) is converted into a photon energy domain flux spectrum Φ E (E), and the conversion relationship is:

[0016]

[0017] In the formula, Φ λ (λ) is the photon wavelength domain flux spectrum, with the unit of s -1 ·m -2 ·nm -1 , λ is the wavelength of light, E is the photon energy, h is the Planck constant, and c is the speed of light in vacuum.

[0018] Further, in step S2, the preset single-junction cell band gap energy range is 0.5 eV to 4.5 eV.

[0019] Further, step S2 includes:

[0020] The short-circuit current density is obtained by integrating and calculating the photon energy domain flux spectrum in the interval where the energy is greater than the band gap E g , and a short-circuit current density spectrum is generated;

[0021] The reverse saturation current density is calculated according to the Planck blackbody radiation spectrum, and a reverse saturation current density spectrum is generated;

[0022] Based on the short-circuit current density and the reverse saturation current density, the open-circuit voltage and the theoretical maximum efficiency are further calculated, and an open-circuit voltage spectrum and a theoretical maximum efficiency spectrum are generated;

[0023] The calculation formula of the short-circuit current density J sc (E g ) is:

[0024]

[0025] In the formula, q is the elementary charge, and E is the photon energy.

[0026] The calculation formula of the reverse saturation current density J0(E g ) is:

[0027]

[0028] In the formula, B (T) is the Planck blackbody radiation spectrum at the cell operating temperature T; EQE ELis the external quantum efficiency of electroluminescence; is the Planck distribution of blackbody radiation; k is the Boltzmann constant;

[0029] open-circuit voltage V oc The calculation formula is:

[0030]

[0031] wherein, is the thermal voltage.

[0032] Further, the simulation of the transient performance under different bandgap combinations in step S3 combines the mixed parallel computing architecture of multi-process parallel processing and just-in-time compilation technology, and includes:

[0033] A total bandgap combination task queue containing all cities and all hourly spectral data is constructed, the total number of bandgap combinations to be calculated is read, the range of the top cell bandgap E g1 and the bottom cell bandgap E g2 is set, and the total bandgap combination tasks are dynamically allocated to all available CPU cores of the computer through a multi-process executor, realizing coarse-grained parallel processing;

[0034] In each parallel process, all bandgap combinations (E g1 , E g2 ) to be evaluated are traversed, and for each bandgap combination, the required basic parameters are directly obtained by performing fast one-dimensional linear interpolation on the key basic parameter maps generated in step S2;

[0035] The core numerical calculation function for solving the maximum output power is compiled into highly optimized local machine code using just-in-time compilation technology, realizing fine-grained calculation acceleration; for a double-ended stacked device, the relationship of total voltage = sum of sub-cell voltages is clear, the optimal working point is determined by solving the maximum power density, and then the transient output power and stack mismatch current density parameters at the time of the optimal working point are calculated;

[0036] Each CPU core stores the current bandgap combination, time period, current density corresponding to the optimal working point, total voltage, transient output power, and stack mismatch current density parameters in the local cache, and after completing all assigned tasks, they are summarized into the transient performance database of the main process.

[0037] Further, the optimal working point is determined by efficiently solving the maximum power density using an iterative search algorithm in step S3, and the transient output power and stack mismatch current density parameters at the time of the optimal working point are calculated; including:

[0038] The top cell and the bottom cell absorb the solar spectrum according to the band gap energy difference, and generate the top cell short-circuit current density and the bottom cell short-circuit current density, respectively.

[0039] The power density is calculated by the product of the total voltage and the current density; the total voltage of the double-end stacked cell is calculated by the sum of the sub-cell voltages, and the instantaneous output power of the double-end stacked cell under the spectrum at this moment is obtained by solving the current density and the total voltage corresponding to the maximum power;

[0040] The stacked mismatch current density is calculated by the absolute value of the difference between the top cell short-circuit current density and the bottom cell short-circuit current density.

[0041] Further, the step S4 comprises:

[0042] The annual total power generation of each band gap combination (E g1 ,E g2 ) is obtained by accumulating the instantaneous output power density at all discrete time points;

[0043] The annual total power generation of different band gap combinations is compared by traversing the entire band gap parameter space, and the globally optimal band gap combination that can realize maximum energy output is screened out.

[0044] Further, the annual total power generation table Y(E g1 ,E g2 ) is shown as:

[0045]

[0046] Wherein, E g1 is the band gap of the top cell, E g2 is the band gap of the bottom cell, Q is the total number of time points, Δt i is the time step; t i is the time; and P is the power generation.

[0047] The second aspect of the present application provides a stacked solar cell performance prediction and band gap parameter optimization system for realizing the stacked solar cell performance prediction and band gap parameter optimization method, comprising:

[0048] A data processing unit is used for automatically scanning and loading a standard format annual hourly solar spectrum time series data file containing at least one geographic location; in this process, the metadata in the data file is parsed, and the solar energy spectrum of each geographic location and each hourly level is accurately converted into a photon energy domain flux spectrum;

[0049] A parameter calculation unit is used for pre-calculation based on the SQ theory for each hourly level photon energy domain flux spectrum output by the data processing unit; a key basic parameter atlas covering a wide range of single-junction band gap energies is generated.

[0050] A performance simulation unit is configured to quickly query and obtain the key performance parameters of each sub-cell from the key basic parameter atlas generated by the parameter calculation unit for any given stack band gap combination, and simultaneously simulate and calculate the instantaneous performance under different band gap combinations, including the instantaneous power and stack mismatch current density parameters, by combining the hybrid parallel computing architecture of multi-process parallel processing and just-in-time (JIT) compilation technology.

[0051] A result aggregation and optimization unit is configured to collect the annual instantaneous performance data of each stack solar cell band gap combination output by the performance simulation unit, and perform grouping aggregation operation to obtain the annual total power generation of each band gap combination, and determine the optimal band gap combination from the numerous band gap combinations by an optimization algorithm according to the annual total power generation, so as to realize the global optimization of the performance of the stack solar cell.

[0052] A report generation unit is configured to automatically generate a multi-dimensional analysis report according to the optimal band gap combination determined by the result aggregation and optimization unit and the process data provided by the performance simulation unit, and the report content covers the annual power generation in different regions and different years and the performance thermodynamic diagram of the relationship between the band gap of the top cell and the bottom cell, which intuitively shows the influence law of the region, time and band gap combination on the power generation, and meanwhile, generates the statistical data of the stack mismatch current density at each time under different band gap combinations to assist in analyzing the reasons for the performance difference of each band gap combination, and provides comprehensive and intuitive reference basis for researchers and industry decision makers.

[0053] Overall, compared with the prior art, the above technical solutions conceived by the present application can achieve the following beneficial effects:

[0054] (1) The present application innovatively proposes a "pre-computation-re-query" mode to realize a key breakthrough from the calculation logic level, aiming at the problem of huge calculation amount and long time consumption caused by massive dynamic spectrum data and wide range of band gap combination scanning in the performance simulation of traditional stack solar cells, and the number of integral operations is greatly reduced from N*M times to only N times by pre-computing the short-circuit current density atlas and the reverse saturation current density atlas based on the SQ theory for each independent hour-level spectrum, and when evaluating the performance of any band gap combination, the integral operation does not need to be repeated, and only the parameters need to be quickly queried from the pre-constructed atlas by interpolation and then the results can be obtained by simple algebraic operation, thereby reducing the amount of repeated calculation from the root. Meanwhile, by combining the hybrid parallel computing architecture of multi-process parallel processing and just-in-time (JIT) technology, the calculation-intensive tasks are allocated to multiple CPU cores, and the core algorithm is optimized at the machine code level, which further improves the simulation speed, effectively solves the efficiency bottleneck of traditional serial calculation "several hours to several days to complete the simulation of a single region", significantly shortens the band gap parameter optimization iteration period of the stack solar cell, and provides support for rapid technical research and development.

[0055] (2) The conventional method generally uses a single static standard solar spectrum (such as AM1.5G) for simulation, which cannot reflect the dynamic changes of the solar spectrum in actual application due to factors such as geographical location, season, weather, day and night time, etc., causing a large deviation between theoretical design and actual power generation performance. The present application discards this idealized mode and uses real solar spectrum time series data at an hourly level or even higher resolution throughout the year as input to accurately restore the dynamic lighting environment under different application scenarios. At the same time, the "annual total power generation" is taken as the core optimization target, replacing the efficiency index under the traditional single spectrum, which is more in line with the needs of the photovoltaic industry "oriented by actual power generation income". The optimal bandgap combination obtained by the scheme of the present application can match the annual power generation demand under real dynamic spectrum, effectively reducing the deviation between theoretical design and actual installed power generation efficiency, improving the commercial application value of the stacked solar cell, and avoiding the waste of design resources caused by scenario disconnection.

[0056] (3) The core computing framework of the present application is based on the ideal SQ theory, such as assuming that the electroluminescence external quantum efficiency is 1, and not considering additional losses such as non-radiative recombination, but is not limited to ideal scenarios. The "pre-computation-re-query" mode and hybrid parallel architecture are universal and can be easily extended to simulations involving non-radiative recombination, stacked / parallel resistance, and other complex actual device loss mechanisms. Only the calculation logic of related loss parameters needs to be supplemented in the basic parameter map construction stage to adapt to the performance evaluation needs closer to the real device. On the other hand, the method of the present application supports loading spectrum data of "at least one geographical location", which can simultaneously complete bandgap optimization under multiple regional and climate scenarios, and generate multi-dimensional analysis reports (including bandgap-power generation heat maps and mismatched current statistics in different regions) to provide customized bandgap design schemes for photovoltaic projects in different regions, adapting to the industrialization layout needs of stacked solar cells worldwide.

[0057] (4) The complete process of real dynamic spectrum input, efficient parallel computing, annual power generation optimization, and multi-dimensional report output of the present application not only can efficiently screen out the optimal bandgap combination that adapts to a specific scenario, but also can intuitively present the correlation between bandgap combination and power generation, and the time distribution characteristics of stacked mismatched current through analysis reports, providing a clear optimization direction for the research and development end. For bandgap combinations with high mismatched current proportion, the bandgaps of sub-cells can be adjusted to reduce current mismatch. For spectral differences in different regions, the top / bottom sub-cell bandgaps can be customized. This technical output from data to conclusion and from conclusion to guidance can effectively link the theoretical design and engineering application of stacked solar cells, help researchers avoid blind tests, reduce research and development costs, and accelerate the conversion process of high-performance stacked solar cells from laboratory technology to industrial products. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 A schematic diagram of a structure of a stacked solar cell;

[0059] Figure 2 A schematic diagram of a flow of a stacked solar cell performance prediction and band gap parameter optimization method according to an embodiment of the present application;

[0060] Figure 3 A schematic diagram of a structure of a stacked solar cell performance prediction and band gap parameter optimization system according to an embodiment of the present application;

[0061] Figure 4 A schematic diagram of a principle of a "pre-computation-re-query" algorithm in a stacked solar cell performance prediction and band gap parameter optimization method according to an embodiment of the present application;

[0062] Figure 5 A performance heat map of annual power generation in Urumqi when changing the band gaps of the top cell and the bottom cell at the same time;

[0063] Figure 6 A performance heat map of annual power generation in Kuala Lumpur when changing the band gaps of the top cell and the bottom cell at the same time;

[0064] Figure 7 A performance graph of annual power generation of three stacked structures of perovskite / silicon (1.12 eV), perovskite / perovskite (1.22 eV) and perovskite / organic stacked solar cell (1.33 eV) with different band gaps when only changing the band gap of the top cell in the Urumqi region;

[0065] Figure 8 A performance graph of annual power generation of three stacked structures of perovskite / silicon (1.12 eV), perovskite / perovskite (1.22 eV) and perovskite / organic stacked solar cell (1.33 eV) with different band gaps when only changing the band gap of the top cell in the Kuala Lumpur region;

[0066] Figure 9 A full-year current mismatch analysis graph of a band gap combination of perovskite / silicon in the Urumqi region under standard spectrum matching conditions;

[0067] Figure 10 A full-year current mismatch analysis graph of a band gap combination of perovskite / perovskite in the Urumqi region under standard spectrum matching conditions;

[0068] Figure 11 A full-year current mismatch analysis graph of a band gap combination of perovskite / organic stacked solar cell in the Urumqi region under standard spectrum matching conditions. DETAILED DESCRIPTION

[0069] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not used to limit the present application. In addition, the technical features involved in the various embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0070] As shown in Fig. 1, it is a schematic diagram of the structure of a stacked solar cell; the stacked solar cell includes glass / transparent conductive oxide Glass / TCO; Glass is glass, and TCO (Transparent Conductive Oxide) is a transparent conductive oxide, which is a transparent conductive electrode allowing light to pass through and conduct current; it provides physical support for the cell, realizes light incidence and current output at the same time, and is a "window" connecting the cell with the outside world; Figure 1

[0071] The hole transport layer HTL (Hole Transport Layer) is used to transport the holes generated by the WBG layer to the electrode, reduce the hole recombination, and improve the carrier collection efficiency;

[0072] The wide band gap layer WBG (Wide Band Gap) uses the wide band gap characteristic to absorb high-energy photons to generate electron-hole pairs, and is an "upper unit" for realizing spectral band utilization in the stacked cell;

[0073] The interconnect layer ICL (Interconnect Layer) is used to connect the upper and lower sub-cells (WBG and NBG), realize the recombination and re-separation of carriers, make the electrons of the upper sub-cell and the holes of the lower sub-cell recombine here, and at the same time make the other type of carriers continue to transport, so as to guarantee the overall current path of the stacked cell;

[0074] The narrow band gap layer NBG (Narrow Band Gap) is used to supplement the absorption of low-energy photons in the solar spectrum, expand the utilization range of the cell to the spectrum, and cooperate with the WBG layer to realize the layered absorption of different energy photons and improve the overall photoelectric conversion efficiency;

[0075] The electron transport layer ETL (Electron Transport Layer) is used to transport electrons (negative charged carriers), transport the electrons generated by the NBG layer to the electrode, reduce the electron recombination, and help the effective collection of carriers;

[0076] The back electrode is used to collect the electrons transported from the ETL, and forms a current loop with the front electrode (Glass / TCO) to output the electric energy generated by the cell;

[0077] ​These layer structures collectively constitute a stacked solar cell, which realizes high-efficiency photoelectric conversion through the layered absorption of different energy photons by the WBG and NBG, and the carrier collection and conduction of each transport layer and electrode.

[0078] The current mainstream band gap optimization method of stacked solar cells in the industry is generally based on a single static standard solar spectrum (such as the internationally recognized AM1.5G spectrum, which simulates the solar radiation under standard test conditions). Performance simulation and band gap combination screening are carried out. However, in actual application scenarios, the working environment of solar cells is dynamically affected by multiple factors: geographical location differences (such as the solar altitude angle and spectral distribution differences between high-latitude and low-latitude regions); seasonal alternation (such as the spectral characteristics change of strong ultraviolet radiation in summer and weak infrared radiation in winter); weather condition fluctuations (such as the spectral intensity and wavelength distribution differences under sunny, overcast, and cloudy weather); diurnal time variation (such as the solar incidence angle and spectral energy proportion differences in the morning, noon, and evening). These factors result in dynamic and complex changes in the solar spectrum actually incident on the cell surface. The design method based on a single static spectrum cannot reflect the actual power generation performance of the cell under dynamic working conditions throughout the year (such as some band gap combinations that perform best under the AM1.5G spectrum may experience a significant efficiency decline in actual cloudy weather or high-latitude regions), ultimately leading to a significant deviation between the theoretical design efficiency and the actual installed power generation efficiency. In some scenarios, the deviation can be as high as 5%-10%, seriously affecting the commercial application value of stacked solar cells.

[0079] To solve the problem of idealized simulation scenarios, some studies attempt to introduce real spectral data at the hourly or even minute level throughout the year (such as based on observation data from meteorological stations in different regions around the world, satellite remote sensing spectral data), and conduct scanning optimization in combination with a wide range of band gap combinations (typically, the top cell band gap range is 1.5-2.2eV, and the bottom cell band gap range is 0.9-1.4eV, covering thousands to tens of thousands of combinations). However, the traditional simulation method uses a serialized nested loop calculation architecture: first fix the top cell band gap, then iterate through the bottom cell band gaps for performance calculation; then adjust the top cell band gap and repeat the above process until all combinations are traversed. In the face of massive dynamic spectral data and complex photoelectric conversion models (which need to include light absorption, carrier transport, and recombination loss), the calculation process of the traditional method is extremely time-consuming: it usually takes several hours to several days to complete the band gap optimization simulation for a single region. If the optimization needs of different climate zones and application scenarios around the world need to be covered, the calculation period can be extended to several weeks or even months, severely restricting the optimization iteration speed of the band gap parameters of stacked solar cells, and unable to meet the needs of rapid technological research and development and industrialization.

[0080] Based on the above reasons, such as Figure 2As shown, one aspect of the present application provides a method for predicting the performance of a stacked solar cell and optimizing the band gap parameters, comprising the following steps:

[0081] S1: automatically scan and load standard format annual hourly solar spectrum time series data files containing at least one geographic location; simultaneously parse the metadata in the solar spectrum time series data file, and convert the hourly solar energy spectrum of each geographic location into a photon energy domain flux spectrum;

[0082] S2: based on the SQ (Shockley-Queisser limit) theory, pre-calculate and generate key basic parameter maps covering a wide range of single-junction band gap energies for each hourly photon energy domain flux spectrum processed in step S1;

[0083] S3: for any stacked band gap combination, quickly query the key performance parameters of each sub-cell from the key basic parameter maps through interpolation, and simulate and calculate the instantaneous performance under different band gap combinations through a hybrid parallel computing architecture combining multi-process parallel processing and just-in-time compilation technology;

[0084] S4: group and aggregate the annual instantaneous performance data of each stacked solar cell band gap combination to obtain the annual total power generation of each band gap combination, and determine the optimal band gap combination according to the annual total power generation;

[0085] S5: automatically generate performance heat maps containing annual power generation and the relationship between top and bottom cell band gaps in different regions and years, multi-dimensional analysis reports of stacked mismatch current density statistical data at each time under different band gap combinations; the performance heat maps visually display the influence of different regions, times, and band gap combinations on power generation; the stacked mismatch current density statistical data assist in analyzing the reasons for the performance differences of each band gap combination, and provide comprehensive and intuitive reference for researchers and industry decision-makers.

[0086] Further, the metadata in step S1 includes the solar energy spectrum in the wavelength domain at the minute or hourly level under actual regional conditions, the geographic location (such as latitude and longitude) of the stacked solar cell application, and the time zone;

[0087] In step S1, the input solar energy spectrum S λ (λ) is converted into a photon energy domain flux spectrum Φ E (E), and the conversion relationship is:

[0088]

[0089] In the formula, Φ λ (λ) is the photon wavelength domain flux spectrum, with a unit of s -1 ·m-2 · nm -1 , λ is the wavelength of light, E is the photon energy, h is the Planck constant, c is the speed of light in vacuum;

[0090] Further, the step S2 comprises:

[0091] The short-circuit current density Jsc is obtained by integrating the photon energy domain flux spectrum in the interval where the energy is greater than the band gap E g , and a short-circuit current density spectrum Jscis generated; sc sc_map ;

[0092] The reverse saturation current density J0 is calculated according to the Planck blackbody radiation spectrum, and a reverse saturation current density spectrum J0is generated; 0_map ;

[0093] Based on the short-circuit current density and the reverse saturation current density, the open-circuit voltage V oc , the theoretical maximum efficiency are further calculated, and an open-circuit voltage spectrum and a theoretical maximum efficiency spectrum are generated;

[0094] The calculation formula of the short-circuit current density is:

[0095]

[0096] Wherein, q is the elementary charge; E is the photon energy;

[0097] The calculation formula of the reverse saturation current density is:

[0098]

[0099] Wherein, is the Planck blackbody radiation spectrum at the battery operating temperature T; EQE EL is the external quantum efficiency of electroluminescence (set to 1 in ideal case); is the Planck distribution of blackbody radiation (reflecting the distribution law of different energy photons with temperature T); k is the Boltzmann constant;

[0100] The calculation formula of the open-circuit voltage V oc is:

[0101]

[0102] Wherein, is the thermal voltage;

[0103] Further, the preset single-junction cell band gap energy range in the step S2 is 0.5eV-4.5eV;

[0104] ​Further, in step S3, the key performance parameters of each sub-cell are obtained from the key parameter atlas by interpolation fast query for any stacked band gap combination, including:

[0105] Determine the target band gap combination: determine the current stacked cell band gap combination to be calculated (such as the top cell band gap and the bottom cell band gap), and the band gap range needs to cover the preset optimization interval (top cell 1.5-2.2eV, bottom cell 0.9-1.4eV);

[0106] Match the corresponding hourly spectrum atlas: according to the current time node, the short-circuit current density atlas and the reverse saturation current density atlas corresponding to the hourly spectrum pre-constructed in step S2 are called;

[0107] Interpolation calculation of key parameters: if the target band gap value is not on the preset discrete band gap point of the atlas, linear interpolation or cubic interpolation method is used to obtain the short-circuit current density of the top and bottom sub-cells from the short-circuit current density atlas, and the reverse saturation current density is obtained from the reverse saturation current density atlas; if the band gap value matches the preset point, the parameters are directly read;

[0108] Derivation of extended performance parameter calculation: based on the short-circuit current density and the reverse saturation current density obtained by query, the open circuit voltage of the top and bottom sub-cells is quickly calculated through the open circuit voltage formula; complete parameters are provided for subsequent instantaneous power calculation;

[0109] The present application first pre-calculates the annual hourly spectrum data by the parameter calculation unit, generates a key atlas, reduces the number of integral operations from N*M times to N times, and then quickly queries the parameters from the pre-calculated atlas by the performance simulation unit for the band gap combination, and performs simple operations, replaces a large number of repeated integrals with N*M times of table lookup, greatly improves the calculation efficiency, and realizes efficient processing of the performance simulation of the stacked solar cell;

[0110] Through step S3, a highly coupled calculation problem with a calculation amount of N spectra*M band gap combinations can be decoupled into an independent pre-processing problem with a calculation amount of N spectra and a fast table lookup problem with a calculation amount of N spectra*M band gap combinations; through the "pre-calculation" step, the number of time-consuming integral operations is reduced from N*M times to N times; subsequently, when evaluating the performance of any band gap combination, the required parameters are quickly "queried" from the constructed atlas by interpolation or other methods, and then simple algebraic operations are performed to obtain the final result, thereby avoiding a large number of repeated integral operations and greatly improving the calculation efficiency. Through the "pre-calculation-requery" mode, the present application aims to solve the problem of huge calculation amount and low efficiency caused by massive dynamic spectrum data and wide range of band gap combination scanning in the performance simulation of stacked solar cells;

[0111] Further, the simulation of the instantaneous performance of the different bandgap combinations in step S3 is performed by combining the multi-process parallel processing and the hybrid parallel computing architecture simulation of the instant compilation technology; including:

[0112] The total bandgap combination task queue containing all cities and all hourly spectral data is constructed, the total number of bandgap combinations to be calculated is read, and the range of the top cell bandgap E g1 and the bottom cell bandgap E g2 is set, the total bandgap combination task is dynamically allocated to all available CPU cores of the computer through the multi-process executor (ProcessPoolExecutor) combined with the number of CPU cores, coarse-grained parallel processing is achieved, and the annual hourly spectral data is split by time period (such as monthly, quarterly) to ensure that each CPU core can independently process the spectral-bandgap combination calculation within a certain time period, avoiding data competition;

[0113] In each parallel process, all bandgap combinations (E g1 , E g2 ) to be evaluated are traversed, and for each bandgap combination, the required basic parameters are directly obtained by performing fast one-dimensional linear interpolation on the key basic parameter maps generated in step S2;

[0114] The core numerical calculation function for solving the maximum output power is compiled into highly optimized local machine code using the just-in-time (JIT) technology, and fine-grained calculation acceleration is achieved; for a double-sided stacked device, the relationship between the total voltage and the sum of the sub-cell voltages is determined, and the optimal working point is efficiently solved by an iterative search algorithm to determine the maximum power density, and the instantaneous output power and the stack mismatch current density parameters at the optimal working point are calculated;

[0115] Each CPU core stores the current bandgap combination, time period, current density corresponding to the optimal working point, total voltage, instantaneous output power, and stack mismatch current density parameters in the local cache, and after completing all assigned tasks, the parameters are summarized into the instantaneous performance database of the main process;

[0116] Further, the optimal working point is efficiently solved by an iterative search algorithm to determine the maximum power density, and the instantaneous output power and the stack mismatch current density parameters at the optimal working point are calculated; including:

[0117] The top cell and the bottom cell are layered to absorb the solar spectrum according to the bandgap energy difference, and the top cell short-circuit current density and the bottom cell short-circuit current density are generated, respectively;

[0118] The power density is calculated by the product of the total voltage and the current density; the total voltage of the double-sided tandem cell is calculated by the sum of the sub-cell voltages, and the instantaneous output power of the double-sided tandem cell under the spectrum at that moment is obtained by solving the current density and the total voltage corresponding to the maximum power;

[0119] The tandem mismatch current density is calculated by the absolute value of the difference between the top cell short-circuit current density and the bottom cell short-circuit current density;

[0120] Specifically, for any tandem solar cell bandgap combination (E g1 ,E g2 ), E g1 is the top cell bandgap (unit: eV), which determines the lower limit of the photon energy that can be absorbed by the top cell (only the photon energy ≥ E g1 can be absorbed by the top cell); E g2 is the bottom cell bandgap (unit: eV), which absorbs the lower energy photons that are not absorbed by the top cell;

[0121] The short-circuit current density of the top cell is:

[0122] J sc,top = J sc (E g1 )

[0123] The J-V characteristic curve of the top cell is:

[0124] J top = J 0,top (e qV / kT -1)-J sc,top

[0125] Where J is the working current density; V is the voltage; J sc (E g1 ) is the short-circuit current absorbed by the top cell; J top is the current density of the top cell under voltage V (unit: A / m 2 ); J 0,top is the reverse saturation current density of the top cell (unit: A / m 2 );

[0126] The short-circuit current density of the bottom cell is:

[0127] J sc,bottom = J sc (E g2 )-J sc (E g1 )

[0128] The J-V characteristic curve of the bottom cell is:

[0129] J bottom = J 0,bottom(e qV / kT -1)-J sc,bottom

[0130] where J sc (E g2 ) is the total short-circuit current corresponding to the bandgap of the bottom cell; J bottom is the current density of the bottom cell at voltage V; J 0,bottom is the reverse saturation current density of the bottom cell (unit: A / m 2 );

[0131] For a double-terminal tandem device with a specific bandgap combination, under the spectrum at time t i , its total voltage V total (J) is the sum of the voltages of each sub-cell (the top cell voltage V top (J) and the bottom cell voltage V bottom (J)), i.e.:

[0132] V total (J) = V top (J) + V bottom (J)

[0133] The working point (J mpp , V mpp ) that maximizes the power density P(J) is determined by an iterative search algorithm, and the instantaneous output power P mpp of the tandem cell under the spectrum at this time is determined;

[0134] Power density:

[0135] P(J) = J x V total (J)

[0136] Instantaneous output power:

[0137] P mpp = J mpp · V mpp

[0138] where J mpp is the current density corresponding to the maximum power density; V mpp is the total voltage corresponding to the maximum power density;

[0139] The voltage of each sub-cell is:

[0140] where J sc,sub is the short-circuit current density of the sub-cell; J 0,sub is the reverse saturation current density of the sub-cell;

[0141] The stack mismatch current density AJ reflects the degree of mismatch between the top and bottom sub-cell current, and is expressed as: AJ = |J sc,top -J sc,bottom .

[0142] The core logic of step S3 is that the top cell and the bottom cell absorb the solar spectrum according to the band gap difference (E g1 ,E g2 ), respectively generating the top cell short-circuit current density J sc,top and the bottom cell short-circuit current density J sc,bottom ;

[0143] The J-V characteristics of the sub-cell are determined by the short-circuit current, the reverse saturation current and the temperature, which reflect the combined action of photo-generated carriers and thermally excited carriers; the total voltage of the double-end stacked cell is the sum of the sub-cell voltages, and the power density is determined by the product of the total voltage and the current density; by solving the maximum power density, the instantaneous output power of the cell under a certain spectrum at a certain time can be obtained, which provides basic data for subsequent annual total power generation calculation;

[0144] Further, step S4 includes:

[0145] The annual total power generation of each band gap combination is obtained by accumulating the instantaneous output power density at all discrete time points, and the annual total power generation table Y (E g1 ,E g2 ) is shown as:

[0146]

[0147] Where E g1 is the band gap of the top cell, E g2 is the band gap of the bottom cell, Q is the total number of time points (for example, for hourly data, Q = 8760), and Δt i is the time step (for example, 1 hour);

[0148] The entire band gap parameter space (all combinations of top cell 1.5-2.2eV and bottom cell 0.9-1.4eV) is traversed, the annual total power generation of different band gap combinations is compared, and the globally optimal band gap combination that can achieve maximum energy output is selected.

[0149] Further, step S5 further includes core data calling and data cleaning and standardization;

[0150] From the metadata of step S1, the "basic information of each geographic location" (such as latitude, longitude, and climate type) is called to provide labels for the regional dimension analysis of the report;

[0151] Extracting the "each bandgap combination - per-time stack mismatch current density" dataset (containing bandgap combination, timestamp, mismatch current density field) from the process data of step S3;

[0152] Extracting the "each region, each year bandgap combination - annual total power generation" correspondence table (containing top cell bandgap, bottom cell bandgap, geographical location, year, annual total power generation core field) from the calculation results of step S4;

[0153] Removing abnormal data, standardizing and sorting bandgap combinations, and unifying data units;

[0154] Further, the generation of the performance heat map of the annual power generation of different regions and different years and the relationship with the top and bottom cell bandgaps in step S5 includes:

[0155] Heat map data matrix construction: grouping by "region-year", and constructing a two-dimensional data matrix of "top cell bandgap x bottom cell bandgap" for each group, with the numerical value of each cell in the matrix being the annual total power generation of the bandgap combination;

[0156] Heat map visualization rendering: using a color mapping scheme (such as "blue → green → red" corresponding to "low power generation → medium power generation → high power generation"), drawing a heat map for each "region-year" data matrix; adding annotations in the heat map: marking the bandgap combination with the highest power generation in each "region-year" group (i.e. the optimal bandgap in this scenario determined in step S4) with a red box, and annotating the specific power generation value; adding a uniform color scale bar to the heat map for multiple regions and years to ensure that the power generation ranges corresponding to different colors are consistent across different heat maps, facilitating horizontal comparison;

[0157] Further, the generation of the per-time stack mismatch current density statistical data of different bandgap combinations in step S5 includes mismatch current density statistics and statistical result visualization;

[0158] Mismatch current density statistics include: grouping by "bandgap combination", and for each group, the following indicators are calculated: ① the average, maximum, and minimum of the annual mismatch current density; ② the duration (in hours) and proportion (as a percentage of the total 8760 hours per year) of mismatch current density > 5 A / m 2 (low mismatch threshold), > 10 A / m 2 (medium mismatch threshold), and > 15 A / m 2 (high mismatch threshold);

[0159] Statistical result visualization includes drawing a "bandgap combination - mismatch current duration proportion" column chart: the horizontal axis is the bandgap combination (sorted in ascending order of top cell bandgap), and the vertical axis is the "high mismatch duration proportion". Different bandgap combinations are distinguished by different colored columns, which intuitively displays the bandgap combinations with high mismatch risk;

[0160] Generate the "mismatch current density time series chart": select typical bandgap combinations (such as the optimal bandgap combination, the high mismatch bandgap combination), draw the hourly mismatch current density change curve throughout the year, mark the time point of peak value occurrence, and analyze the time distribution law of the mismatch current;

[0161] Organize the "mismatch current statistical data table": summarize the statistical indicators (average value, extreme value, and each threshold duration) of each bandgap combination into a table for researchers to quickly refer to specific data.

[0162] As shown in Figure 3 The second aspect of the present application provides a stacked solar cell performance prediction and bandgap parameter optimization system for realizing the above design method, comprising:

[0163] A data processing unit is used to automatically scan and load standard format, annual hourly solar spectrum time series data files containing at least one geographic location; In this process, the metadata in the data file is parsed, and each geographic location and each hourly solar energy spectrum is accurately converted into a photon energy domain flux spectrum, providing basic and key data support for subsequent calculation and analysis;

[0164] A parameter calculation unit is used to pre-calculate each hourly photon energy domain flux spectrum output by the data processing unit based on the SQ theory; Through complex and accurate algorithms, key basic parameter maps covering a wide range of single-junction bandgap energies are generated, such as short-circuit current density maps, reverse saturation current density maps, etc. These maps are important parameter references for subsequent performance simulation and bandgap optimization;

[0165] A performance simulation unit is used to quickly query and obtain the key performance parameters of each sub-cell from the key basic parameter maps generated by the parameter calculation unit for any given stacked bandgap combination. At the same time, through the hybrid parallel computing architecture combining multi-process parallel processing and just-in-time compilation technology, the instantaneous performance under different bandgap combinations is simulated and calculated, including instantaneous power, stacked mismatch current density parameters, providing data support for evaluating the real-time performance of different bandgap combinations;

[0166] A result aggregation and optimization unit is used to collect the annual instantaneous performance data of each stacked solar cell bandgap combination output by the performance simulation unit, and perform grouping aggregation operation, thereby obtaining the annual total power generation of each bandgap combination. According to the core indicator of annual total power generation, the optimal bandgap combination is determined among numerous bandgap combinations through optimization algorithms, realizing the global optimization of the performance of stacked solar cells;

[0167] The report generation unit is used for automatically generating a multi-dimensional analysis report according to the optimal band gap combination determined by the result aggregation and optimization unit and the process data provided by the performance simulation unit; the report content covers the performance heat map of the annual power generation and the relationship between the top and bottom cell band gaps in different regions and different years, and intuitively displays the influence law of the region, time and band gap combination on the power generation; meanwhile, the statistical data of the laminated mismatch current density at each moment under different band gap combinations are generated, which assist in analyzing the reasons for the performance difference of each band gap combination, and provide comprehensive and intuitive reference basis for the researchers and industry decision makers.

[0168] In some possible embodiments, the embodiment provides that the laminated solar cell performance prediction and band gap parameter optimization system can be realized in a software manner, which can be software in the form of programs and plug-ins, and includes a series of modules to realize the laminated solar cell performance prediction and band gap parameter optimization method provided by the embodiment.

[0169] In the specific embodiments of the present application, it is assumed that the representative cities of multiple different climate regions in the world, such as Urumqi in China (north latitude 43.83°, east longitude 87.61°) and Kuala Lumpur in Malaysia (north latitude 3.15°, east longitude 101.69°), need to be designed for optimal performance of the perovskite / silicon (PVK / Si), perovskite / perovskite (PVK / PVK) and perovskite / organic (PVK / OPV) double-junction laminated solar cells, so that the power generation is maximized in the actual operation of each region throughout the year; based on the above real-world dynamic spectral data, the laminated solar cell performance prediction and band gap parameter optimization method is further demonstrated, and the laminated solar cell performance prediction and band gap parameter optimization system is applied to realize, which includes the following steps in the specific embodiments:

[0170] S1: data loading, preprocessing and conversion

[0171] This step is executed by the data processing unit shown in Figure 3 The system first automatically scans and loads a specified data directory containing multiple city spectral data files in batch mode. In the embodiment, the annual hourly spectral data of Urumqi and Kuala Lumpur can be obtained in multiple ways, including but not limited to: using open-source atmospheric parameter data sets to drive atmospheric radiation transfer models to calculate local solar spectrum hourly; directly downloading publicly available hourly solar spectrum database; obtaining minute or hourly spectral data through ground monitoring network after quality control and interpolation processing. The system not only reads the annual 8760 hours of solar irradiance components in the wavelength domain and related meteorological parameters of each city, but also automatically parses the geographic location, time zone, altitude and other metadata in the file, and associates these information with the corresponding data set.

[0172] Subsequently, the hourly solar energy spectrum S for each city was analyzed. λ (λ)(Unit: W·m) -2 ·nm -1 Converted to photon energy domain flux spectrum Φ E (E)(Unit: s) -1 ·m -2 (eV-1). The conversion formula is as follows:

[0173]

[0174] This step provides preprocessed, standardized, and geolocation-related physical input for all subsequent calculations.

[0175] S2: Constructing the basic parameter map (pre-calculation stage)

[0176] This step is by Figure 3 The parameter calculation unit shown executes the core algorithm principle as follows: Figure 4 As shown. For each hourly photon energy domain flux spectrum obtained in S1 (e.g. Figure 4 The system performs pre-calculations based on SQ theory (as shown in the input spectrum).

[0177] Reference Figure 4 The intermediate parameter calculation unit scans a preset wide range of single-junction bandgap energies (e.g., from 0.5 eV to 4.5 eV) and, through numerical integration, generates a unique spectrum of key fundamental parameters for that hour's spectrum, including:

[0178] Short-circuit current density spectrum J sc_map :

[0179]

[0180] Reverse saturation current density spectrum J 0_map :

[0181]

[0182] in, The Planck blackbody radiation spectrum at temperature T, EQE EL It is the external quantum efficiency of electroluminescence (ideally set to 1).

[0183] Through this step, for each of the 8760 spectra collected throughout the year for each city, a unique, high-resolution J-series dataset was generated for each spectrum. sc_map and J 0_map The most time-consuming integration calculations are completed in one go during this stage.

[0184] S3: High-performance parallel simulation (re-query and computation phase)

[0185] This step is performed by the performance simulation unit shown in Figure 3 The system sets the range of top cell bandgaps E g1 and bottom cell bandgaps E g2 that need to be scanned.

[0186] Parallel task distribution: The system constructs a total task queue containing all cities, all hourly spectral data, and utilizes multi-process parallel processing technology to dynamically distribute tasks to all available CPU cores of the computer.

[0187] Fast query and calculation: In each parallel process, the system traverses all bandgap combinations (E g1 , E g2 ) to be evaluated. For each combination, the required basic parameters are directly queried by performing fast one-dimensional linear interpolation on the J sc_map and J 0_map already generated in S2, and then the maximum output power under that hour is efficiently solved through an iterative search algorithm.

[0188] Just-in-time (JIT) acceleration: To further improve efficiency, the system uses just-in-time compilation technology to compile the above iterative search algorithm and other computationally intensive functions into local machine code, achieving extreme acceleration at the micro-algorithm level.

[0189] S4: Result aggregation and global optimization

[0190] This step is performed by the result aggregation and optimization unit shown in Figure 3 When all parallel computing tasks for a city are completed, the system groups and aggregates the results by city.

[0191] For each city, the system aggregates all hourly power density data for that city and calculates the annual total power generation of each bandgap combination (E g1 , E g2 ) in that city:

[0192]

[0193] Where Q is the total number of time points (for example, Q = 8760 for hourly data), and Δt i is the time step (e.g., 1 hour).

[0194] Finally, the system searches for the maximum value in the two-dimensional AEY matrix generated for each city, respectively, to obtain the globally optimal bandgap combination (E g1,opt , E g2,opt ) for each city. city

[0195] S5: Report generation and deep analysis​

[0196] This step is performed by the report generation unit shown in Figure 3 The system outputs a comprehensive analysis report:

[0197] Referring to Figure 5 , Figure 6 , which shows the annual power generation performance heat map for Urumqi and Kuala Lumpur respectively while changing the top and bottom cell bandgaps simultaneously. The X-axis represents the top cell bandgap, the Y-axis represents the bottom cell bandgap, and the color depth represents the level of annual total power generation. Through the figure, the optimal bandgap combination point can be clearly located.

[0198] Referring to Figure 7 , Figure 8 , which shows the annual power generation performance map for Urumqi and Kuala Lumpur respectively while changing only the top cell bandgap. Through the figure, the optimal wide bandgap value of the different tandem cells can be clearly located. Figure 7 and Figure 8 “organic” in the tandem solar cell refers to the narrow bandgap material being an organic material.

[0199] Referring to Table 1, which shows the optimal top cell bandgap value, the corresponding best bandgap efficiency, and the efficiency of the bandgap under the actual spectrum determined by the AM1.5G spectrum for three common bottom cells—silicon (1.12 eV), perovskite (1.22 eV), and organic tandem solar cell (1.33 eV) in different geographical locations around the world. It can be clearly observed that due to the differences in spectral resources and climate conditions, the optimal bandgap points in different regions are significantly different, which directly proves the necessity and accuracy of the optimization for specific locations by the present application.

[0200] Table 1-Statistical results of the original bandgap efficiency and the best bandgap efficiency of three common bottom cells in different geographical locations around the world

[0201]

[0202] Referring to Figures 9-11 , which shows the probability density distribution of current mismatch in the actual operation of the perovskite / silicon, perovskite / perovskite, and perovskite / organic tandem cell in Urumqi under the current matching bandgap combination determined by the standard spectrum (AM1.5G). Through the figure, the deviation between the theoretical design and the actual operation can be analyzed. Figure 11 “organic” in the tandem solar cell refers to the narrow bandgap material being an organic material.

[0203] Those skilled in the art can understand that the above description is only the preferred embodiment of the present application, and is not intended to limit the present application. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for predicting the performance and optimizing the bandgap parameters of tandem solar cells, characterized in that, Includes the following steps: S1: Automatically scan and load a standard format solar spectrum time series data file containing at least one geographical location, covering the entire year at the hourly level; simultaneously parse the metadata in the solar spectrum time series data file, converting the solar energy spectrum of each geographical location and each hourly level into a photon energy domain flux spectrum. S2: Based on SQ theory, pre-calculate the photon energy domain flux spectrum for each hour after processing in step S1 and generate a key fundamental parameter map covering a wide range of single-junction bandgap energies. S3: For any stacked bandgap combination, the key performance parameters of each sub-cell are quickly obtained from the key basic parameter map by interpolation. The instantaneous performance under different bandgap combinations is simulated by a hybrid parallel computing architecture that combines multi-process parallel processing and just-in-time compilation technology. S4: Group and aggregate the instantaneous performance data of each tandem solar cell bandgap combination throughout the year to obtain the total annual power generation of each bandgap combination, and determine the optimal bandgap combination based on the total annual power generation. S5: Automatically generates performance heatmaps showing the relationship between annual power generation and top and bottom cell bandgap in different regions and years, as well as multi-dimensional analysis reports on the statistical data of stacked mismatch current density at each moment under different bandgap combinations.

2. The method for predicting the performance and optimizing the bandgap parameters of a tandem solar cell according to claim 1, characterized in that: The metadata mentioned in step S1 includes the solar energy spectrum in the wavelength domain at the minute or hour level under actual geographical conditions, the geographical location of the tandem solar cell application, and the time zone.

3. The method for predicting the performance and optimizing the bandgap parameters of a tandem solar cell according to claim 2, characterized in that: In step S1, the input solar energy spectrum S λ (λ) is converted into the photon energy domain flux spectrum Φ E (E), the transformation relationship is: In the formula, Φ λ (λ) represents the photon wavelength-domain flux spectrum, with units of s. -1 ·m -2 ·nm -1 λ is the wavelength of light, E is the photon energy, h is Planck's constant, and c is the speed of light in a vacuum.

4. A method for predicting the performance and optimizing the bandgap parameters of a tandem solar cell according to any one of claims 1-3, characterized in that: The preset bandgap energy range for a single-junction cell in step S2 is 0.5eV to 4.5eV.

5. The method for predicting the performance and optimizing the bandgap parameters of a tandem solar cell according to claim 4, characterized in that, Step S2 includes: By analyzing the photon energy domain flux spectrum at energies greater than the band gap E... g The short-circuit current density is obtained by integral calculation over the interval, and a short-circuit current density spectrum is generated. The reverse saturation current density is calculated based on the Planck blackbody radiation spectrum, and a reverse saturation current density map is generated. Based on the short-circuit current density and reverse saturation current density, the open-circuit voltage and theoretical maximum efficiency are further calculated, and the open-circuit voltage spectrum and theoretical maximum efficiency spectrum are generated. Short-circuit current density J sc (E g The formula for calculating ) is: Where q is the elementary charge; E is the photon energy; Reverse saturation current density J0(E) g The formula for calculating ) is: in, The Planck blackbody radiation spectrum at the battery operating temperature T; EQE EL It is the external quantum efficiency of electroluminescence; is the Planck distribution of blackbody radiation; k is the Boltzmann constant; Open circuit voltage V oc The formula for calculation is: in, This is thermal voltage.

6. A method for predicting the performance and optimizing the bandgap parameters of a tandem solar cell according to any one of claims 1-3, characterized in that: Step S3 describes simulating the instantaneous performance under different bandgap combinations using a hybrid parallel computing architecture that combines multi-process parallel processing with just-in-time (JIT) compilation techniques; including: Construct a task queue containing total bandgap combinations of spectral data for all cities and all hours, read the total number of bandgap combinations to be calculated, and set the top cell bandgap E to be scanned. g1 and bottom cell bandgap E g2 The range, combined with the number of CPU cores, is used to dynamically allocate the total bandgap combined tasks to all available CPU cores of the computer through a multi-process executor, thereby achieving coarse-grained parallel processing. In each parallel process, all bandgap combinations to be evaluated (E) are traversed. g1 E g2 For each bandgap combination, the required basic parameters can be directly obtained by performing fast one-dimensional linear interpolation on the key basic parameter map generated in step S2. Just-in-time compilation technology is used to compile the core numerical calculation function for solving the maximum output power into highly optimized native machine code in just-in-time, thereby achieving fine-grained calculation acceleration. For double-ended stacked devices, the relationship between total voltage and the sum of sub-cell voltages is clarified. The optimal operating point is determined by solving for the maximum power density, and then the instantaneous output power and stacked mismatch current density parameters at the time of the optimal operating point are calculated. Each CPU core associates and stores the current bandgap combination, time period, current density, total voltage, instantaneous output power, and stack-up mismatch current density parameters corresponding to the optimal operating point in its local cache. After all assigned tasks are completed, these parameters are aggregated into the instantaneous performance database of the main process.

7. The method for predicting the performance and optimizing the bandgap parameters of a tandem solar cell according to claim 6, characterized in that: The optimal operating point is determined by efficiently solving for the maximum power density using an iterative search algorithm, and the instantaneous output power and stack mismatch current density parameters at the moment of the optimal operating point are calculated; including: The top and bottom cells absorb the solar spectrum in layers based on the difference in bandgap energy, generating short-circuit current densities for the top and bottom cells, respectively. The power density is calculated by multiplying the total voltage and the current density; the total voltage of the dual-ended stacked battery is calculated by summing the voltages of the sub-cells; by solving for the current density and total voltage corresponding to the maximum power, the instantaneous output power of the dual-ended stacked battery under the spectrum at that moment is obtained. The stack mismatch current density is calculated by the absolute value of the difference between the short-circuit current density of the top cell and the short-circuit current density of the bottom cell.

8. A method for predicting the performance and optimizing the bandgap parameters of a tandem solar cell according to any one of claims 1-3, characterized in that: Step S4 includes: Each bandgap combination (E) is obtained by summing the instantaneous output power density at all discrete time points. g1 E g2 The annual total power generation; By traversing the entire bandgap parameter space, comparing the annual total power generation of different bandgap combinations, the globally optimal bandgap combination that maximizes energy output is selected.

9. The method for predicting the performance and optimizing the bandgap parameters of a tandem solar cell according to claim 8, characterized in that, The annual total power generation table Y(E) g1 E g2 This is shown as: Among them, E g1 For the top cell band gap, E g2 The cell bandgap is the lowest, Q is the total number of time points, and Δt is the lowest bandgap. i It is the time step; t i t is the time; P is the power generation output.

10. A system for predicting the performance and optimizing the bandgap parameters of tandem solar cells, characterized in that, A method for predicting the performance and optimizing the bandgap parameters of tandem solar cells as described in any one of claims 1-9, comprising: The data processing unit is used to automatically scan and load a standard format hourly solar spectrum time series data file containing at least one geographical location throughout the year; in this process, the metadata in the data file is parsed to accurately convert the solar energy spectrum of each geographical location and each hour into a photon energy domain flux spectrum. The parameter calculation unit is used to pre-calculate the photon energy domain flux spectrum output by the data processing unit for each hour based on SQ theory; and to generate a key fundamental parameter map covering a wide range of single-junction bandgap energies. The performance simulation unit is used to quickly retrieve the key performance parameters of each sub-cell from the key basic parameter map generated by the parameter calculation unit for any given stacked bandgap combination through interpolation algorithm; at the same time, it simulates and calculates the instantaneous performance under different bandgap combinations, including instantaneous power and stacked mismatch current density parameters, through a hybrid parallel computing architecture that combines multi-process parallel processing and just-in-time compilation technology. The result aggregation and optimization unit is used to collect the instantaneous annual performance data of each bandgap combination of tandem solar cells output by the performance simulation unit, and perform group aggregation calculations to obtain the annual total power generation of each bandgap combination. Based on the annual total power generation, the optimal bandgap combination is determined from many bandgap combinations through optimization algorithms to achieve global optimization of the performance of tandem solar cells. The report generation unit automatically generates a multi-dimensional analysis report based on the optimal bandgap combination determined by the result aggregation and optimization unit, and the process data provided by the performance simulation unit. The report content includes performance heatmaps showing the relationship between annual power generation and top and bottom cell bandgap in different regions and years, intuitively demonstrating the impact of region, time, and bandgap combination on power generation. At the same time, it generates statistical data on the stacked mismatch current density at each moment under different bandgap combinations, assisting in the analysis of the reasons for the performance differences of each bandgap combination, and providing comprehensive and intuitive reference for R&D personnel and industry decision-makers.

Images