Method and system for power generation optimization control of multiple strings of independent inverter for heterojunction photovoltaic module

Abstract

The application discloses a heterogeneous photovoltaic module multi-string independent inverter power generation optimization control method and system, relates to the photovoltaic utilization technical field, and the method comprises the following steps: collecting real-time operation data of a photovoltaic system cell block, and constructing a plurality of cell characteristic vectors; K isomorphism cell block clusters are generated through cross-component electrical topology clustering based on the characteristic vectors; the electromagnetic field balance of each cluster is optimized to obtain K alternative branch combinations, K target series branch combinations are selected through global impedance balance checking, the electrical architecture of the photovoltaic system is reconstructed, K independent direct-current routing channels are formed, and the MPPT interface dedicated to the inverter is connected respectively. The application solves the technical problems of low power generation efficiency and high safety hazards of the existing photovoltaic system caused by inconsistent performance of the cell block and the centralized inverter, and achieves the technical effects of improving the power generation efficiency and reducing the safety hazards of the system by optimizing the series connection path of the cell block and the independent inverter configuration.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic technology, specifically to a power generation optimization control method and system for multi-string independent inverters of heterojunction photovoltaic modules. Background Technology

[0002] With the continuous growth of global demand for renewable energy, photovoltaic (PV) power generation technology has developed rapidly. Heterojunction PV modules, as a highly efficient solar energy conversion technology, have attracted widespread attention due to their high efficiency and low temperature coefficient. However, traditional PV systems typically employ centralized inverters, which may result in some cell blocks not being fully utilized, affecting overall power generation efficiency. Furthermore, the high current and high voltage output characteristics also increase system heat loss and safety hazards. Summary of the Invention

[0003] This application provides a power generation optimization control method and system for multi-string independent inverters of heterojunction photovoltaic modules, which is used to solve the technical problems of low power generation efficiency and high safety hazards caused by inconsistent performance of battery cells and centralized inverters in existing photovoltaic systems.

[0004] The first aspect of this application provides a power generation optimization control method for multi-string independent inverters of heterojunction photovoltaic modules. The method includes: collecting real-time operating data of photovoltaic system cell blocks and constructing multiple cell feature vectors, wherein the cell feature vectors include open-circuit voltage, short-circuit current, annual degradation rate, location height, and geographical coordinates; clustering the in-service cell blocks across module electrical topology based on the multiple cell feature vectors to generate K homogeneous cell block clusters; optimizing the series path of the K homogeneous cell block clusters based on electromagnetic field equilibrium to output K candidate series branch combinations; performing global impedance equalization verification on the K candidate series branch combinations to filter and output K target series branch combinations; and using the K target series branch combinations to reconstruct the electrical architecture of K groups of in-service cell blocks in the photovoltaic system to obtain K independent DC routing channels, wherein each independent DC routing channel is connected to a dedicated inverter.

[0005] A second aspect of this application provides a power generation optimization control system for multi-string independent inverters of heterojunction photovoltaic modules. The system includes: a cell feature vector acquisition module, used to collect real-time operating data of the photovoltaic system's cell blocks and construct multiple cell feature vectors, wherein the cell feature vectors include open-circuit voltage, short-circuit current, annual degradation rate, location altitude, and geographical coordinates; an electrical topology clustering module, used to cluster the in-service cell blocks across modules based on the multiple cell feature vectors, generating K homogeneous cell block clusters; a series path optimization module, used to optimize the series paths of the K homogeneous cell block clusters based on electromagnetic field equilibrium, outputting K candidate series branch combinations; a global impedance equalization verification module, used to perform global impedance equalization verification on the K candidate series branch combinations, filtering and outputting K target series branch combinations; and an electrical architecture reconfiguration module, used to reconfigure the electrical architecture of the K groups of in-service cell blocks in the photovoltaic system using the K target series branch combinations, obtaining K independent DC routing channels, wherein each independent DC routing channel corresponds to a dedicated inverter.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application provides a power generation optimization control method and system for multi-string independent inverters of heterojunction photovoltaic modules, which relates to the field of photovoltaic utilization technology. It forms a homogeneous battery cell cluster by performing cross-module electrical clustering based on the multi-dimensional feature vectors of in-service battery cells, and optimizes the series path through electromagnetic field equalization and global impedance verification to achieve dynamic electrical reconfiguration across modules. This constructs multiple high-voltage, low-current DC routing channels that can be independently inverted, solving the technical problems of low power generation efficiency and high safety hazards caused by inconsistent battery cell performance and centralized inverters in existing photovoltaic systems. It achieves the technical effect of improving power generation efficiency and reducing system safety hazards by optimizing the series path of battery cells and the configuration of independent inverters. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a schematic diagram of the power generation optimization control method for multi-string independent inverter of heterojunction photovoltaic modules provided in an embodiment of this application. Figure 2 This is a schematic diagram of the power generation optimization control system structure for multiple independent inverters of a heterojunction photovoltaic module provided in an embodiment of this application.

[0009] Figure labeling: Battery feature vector acquisition module 11, electrical topology clustering module 12, series path optimization module 13, global impedance equalization verification module 14, electrical architecture reconstruction module 15. Detailed Implementation

[0010] This application provides a power generation optimization control method and system for multi-string independent inverters of heterojunction photovoltaic modules, which is used to solve the technical problems of low power generation efficiency and high safety hazards caused by inconsistent performance of battery cells and centralized inverters in existing photovoltaic systems.

[0011] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0012] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices.

[0013] Example 1, as Figure 1 As shown, this application provides a power generation optimization control method for multi-string independent inverters of heterojunction photovoltaic modules, the method comprising: P10: Collect real-time operating data of photovoltaic system cell blocks and construct multiple cell feature vectors, wherein the cell feature vectors include open-circuit voltage, short-circuit current, annual degradation rate, location altitude, and geographical coordinates.

[0014] Furthermore, step P10 in this embodiment of the application also includes: P11: Parse the real-time operating data to obtain multiple multi-dimensional status logs of multiple in-service battery cells in the photovoltaic system; P12: Construct multiple initial feature vectors based on the multiple multi-dimensional status logs; P13: Predefine the baseline parameters of the photovoltaic system, perform data standardization processing on the multiple initial feature vectors, and output multiple battery feature vectors, wherein the battery feature vectors include open-circuit voltage, short-circuit current, annual degradation rate, location altitude, and geographical coordinates.

[0015] It should be understood that to achieve efficient power generation control of heterojunction photovoltaic modules in cross-module series connection, electrical architecture reconfiguration, and multi-string independent inverter modes, it is first necessary to accurately collect real-time operating data of each in-service cell block within the photovoltaic system. This data includes key parameters such as open-circuit voltage (Voc), short-circuit current (Isc), annual degradation rate, location height, and geographic coordinates. Open-circuit voltage refers to the cell voltage when no load is connected, while short-circuit current refers to the current when the two ends of the cell are directly connected. The annual degradation rate reflects the degree of performance degradation of the cell over time. Location height indicates the vertical position of the cell block during installation. Geographic coordinates, especially Universal Transverse Mercator (UTM) coordinates, are used to accurately locate the position of the cell block.

[0016] Next, this real-time operational data is analyzed to obtain multi-dimensional status logs for multiple in-service solar cells in the photovoltaic system. These logs record the operating status of the cells at different points in time, including voltage, current, and temperature, providing a rich data source for constructing the initial feature vector. By analyzing this data, a more comprehensive understanding of the cell's operating status can be achieved, laying the foundation for subsequent analysis and optimization.

[0017] Next, based on multi-dimensional status logs, the key electrical parameters of the battery cells are abstracted and described, constructing multiple initial feature vectors. These initial feature vectors include key parameters such as open-circuit voltage, short-circuit current, annual degradation rate, location altitude, and geographic coordinates. The geographic coordinates can be expressed using UTM coordinates, enabling precise positioning of each battery string within the array at the meter-scale, providing a precise geometric basis for subsequent cross-module series connection and electromagnetic field equalization calculations.

[0018] Finally, due to differences in factory parameters, installation orientation, operational lifespan, and environmental conditions among battery strings from different components, it is necessary to predefine benchmark parameters for the photovoltaic system and standardize the initial feature vectors to ensure comparability between different battery blocks. Benchmark parameters include standardized rated voltage ranges, short-circuit current ranges, typical degradation curve models, and unified coordinate reference surfaces. Data standardization involves scaling the data proportionally to fall within a specific range, such as [0,1], thereby restoring the features of different battery blocks to a comparable, unified scale, ensuring consistent scaling of the feature vectors in subsequent clustering analyses. After standardization, multiple battery feature vectors are output, including open-circuit voltage, short-circuit current, annual degradation rate, location altitude, and geographical coordinates, providing a solid data foundation for subsequent electrical architecture reconfiguration and power generation optimization control.

[0019] P20: Based on the multiple battery feature vectors, cluster the in-service battery blocks across the component electrical topology to generate K homogeneous battery block clusters.

[0020] Furthermore, step P20 in this embodiment of the application also includes: P21: Predefine battery block similarity constraints and battery block distance constraints; P22: Use the battery block similarity constraints to enumerate and compare the feature vectors of the multiple batteries to construct multiple electrically connected subgraphs; P23: Obtain multiple candidate isomorphic clusters by aligning and overlapping the multiple electrically connected subgraphs; P24: Extract multiple geographic coordinates from the multiple battery feature vectors, and use the battery block distance constraints to traverse the multiple geographic coordinates to perform spatial density verification and splitting of the multiple candidate isomorphic clusters to obtain the K isomorphic battery block clusters.

[0021] Optionally, to achieve electrical architecture reconfiguration for cross-component in-sequence series connection and independent inverter of multiple branches, it is necessary to identify in-service battery blocks with similar electrical performance, consistent attenuation characteristics, and suitable geometric spatial locations at the system level, so as to construct homogeneous series branches that can support high-voltage, low-current output. In this application, each component output is only a single string, that is, the output of a single cell string in each component is interconnected with the single cell string of the next component, and the cell strings within the same component are neither connected in series nor in parallel. Components are connected to each other through single-string interconnection, and the configuration of the series path is flexible and has no fixed restrictions. This interconnection method is not affected by traditional parallel constraints and can be flexibly adjusted as needed to ultimately form a series branch that meets the requirements for optimized electrical performance. Therefore, based on the multiple battery feature vectors obtained above, cross-component electrical topology clustering can be performed on the battery blocks within the photovoltaic array range to generate K homogeneous battery block clusters. Each homogeneous cluster consists of battery blocks with similar electrical performance, consistent operating states, and suitable for forming cross-component series paths, and is the basic unit for subsequent electromagnetic field equalization optimization and impedance verification.

[0022] Next, to ensure the effectiveness of the clustering results in terms of electrical compatibility and spatial layout, it is necessary to predefine battery block similarity constraints and battery block distance constraints. Battery block similarity constraints limit the compatibility range of two battery blocks in terms of electrical performance; that is, they define which parameters are similar enough to group two battery blocks into the same cluster, including similarity thresholds for parameters such as open-circuit voltage, short-circuit current, and annual degradation rate. Battery block distance constraints refer to the maximum allowable distance between battery blocks in spatial distribution, ensuring the spatial concentration of battery blocks within a cluster.

[0023] Next, using battery block similarity constraints, the collected feature vectors of multiple batteries are enumerated and compared. This involves comparing each battery block's feature vector pairwise and calculating their similarity score under predefined similarity constraints. Based on these scores, multiple electrically connected subgraphs can be constructed, each representing a group of battery blocks similar in electrical characteristics. Specifically, with each battery block as a graph node, when it satisfies the electrical consistency condition with other battery blocks under the aforementioned similarity constraints, an electrical connection edge can be established between them, thus forming a local graph structure reflecting their electrical performance interoperability. Multiple connected subgraphs can cover different components and battery strings with different sequence numbers within the array, reflecting the reconfigurability of battery blocks across component dimensions.

[0024] Then, by aligning and overlapping these electrically connected subgraphs, multiple candidate isomorphic clusters can be obtained. The alignment process forms candidate battery block clusters with potential cross-component homogeneity by merging sets of battery nodes that recur in different connected subgraphs and reconstructing their common connectivity structure. The candidate clusters already possess isomorphism at the electrical performance level, but still need to meet the geometric requirements of spatial density.

[0025] To this end, the geographical coordinates of each battery cell are extracted from multiple battery feature vectors, and a distance-constrained traversal verification is applied to the candidate isomorphic clusters. By calculating the spatial distribution, mutual distance, projection position, and component arrangement direction of the nodes within the cluster, clusters that are too spatially scattered or lack sufficient feasibility for series connection can be identified and split or eliminated. After spatial density verification, K isomorphic battery cell clusters that satisfy electrical consistency and geometric feasibility are finally formed. These clusters are consistent in electrical characteristics and spatially concentrated, which will be used for the electrical architecture reconstruction of photovoltaic systems, laying a structured data foundation for building independent DC routing channels and connecting dedicated inverters.

[0026] P30: Based on electromagnetic field equilibrium, optimize the series path of the K isomorphic battery cell clusters and output K alternative series branch combinations.

[0027] Furthermore, step P30 in this embodiment of the application also includes: P31: Construct the shortest path topology of the first isomorphic battery cell cluster based on geographical coordinates to obtain the first initial series path; P32: Calculate the first equivalent DC impedance spectrum of the first initial series path; P33: Set quantum annealing segmentation parameters according to the first equivalent DC impedance spectrum, divide the first initial series path, and obtain multi-group parallel branches corresponding to multiple quantized breakpoint sequences; P34: Perform electromagnetic field equalization iteration on the multi-group parallel branches to output the first candidate series branch combination.

[0028] It should be understood that, in order to ensure that the cross-component in-sequence series path achieves optimal performance in terms of electrical consistency, spatial rationality, and DC-side magnetic coupling balance, the electromagnetic field balance principle is further used to optimize the series path of the K isomorphic battery clusters, so as to output K alternative series branch combinations.

[0029] First, the geometric relationships of each battery cell within the homogeneous cluster are modeled based on their geographical coordinates. Then, shortest path algorithms, such as Dijkstra's and Floyd's, are used to construct the shortest path topology of the target cluster, generating the first initial series path covering all battery cell nodes in the cluster. Since the homogeneous cluster has undergone spatial density verification, this type of shortest path can accurately reflect the optimal geometric routing in cross-component series connection scenarios, laying the foundation for reducing additional line losses caused by path length.

[0030] After obtaining the initial series path, the electrical characteristics of this path are further modeled, and its first equivalent DC impedance spectrum is calculated. The impedance spectrum comprehensively considers parameters such as the internal resistance of the battery pack, lead resistance, connector contact resistance, and distributed inductance and parasitic capacitance that may be introduced by cross-component wiring. It can be solved using a transmission line model, a lumped parameter model, or a frequency domain equivalent model. This impedance spectrum reflects the comprehensive electrical characteristics of the series path under different operating frequency bands and is an important input basis for subsequent electromagnetic field equalization optimization.

[0031] To further eliminate magnetic coupling inhomogeneities, local impedance abrupt changes, and potential current saturation points at the entire path scale, quantum annealing segmentation parameters are further set based on the first equivalent DC impedance spectrum, and the initial series path is segmented accordingly. Quantum annealing is an optimization algorithm that finds the optimal solution to a problem by simulating the annealing process of a quantum system. In this step, the initial series path can be segmented using the quantum annealing algorithm to obtain multiple sets of parallel branches corresponding to multiple quantized breakpoint sequences. For example, by mapping the impedance spectrum to the energy potential well function in the quantum annealing model, multiple quantized breakpoint sequences are automatically set according to the gradient changes in different impedance intervals, and the initial series path is then segmented using these breakpoint sequences, thereby forming multiple sets of electrically parallel branches. These branch structures serve as the optional state space of the quantum annealing algorithm during the path solving process, enabling the system to search simultaneously in multiple path energy states, which helps to escape local optima and obtain a series configuration with better overall electromagnetic balance.

[0032] Finally, electromagnetic field equilibrium iterative calculations were performed on the multi-group parallel branches formed by the quantized circuit breakers. Based on the multiphysics coupling model, the branch structure was gradually adjusted by repeatedly solving for path current distribution, magnetic flux coupling, parasitic parameter perturbations, and interactions between branches until the results converged to a series branch combination that met the requirements of minimum energy criterion, magnetic coupling equilibrium, and DC impedance minimization, thus obtaining the first candidate series branch combination. The first candidate series branch combination has significant characteristics such as compact wiring, low electromagnetic interference, and high impedance equilibrium, which can support the subsequent construction of high-voltage, low-current cross-component series independent inverter branches.

[0033] Similarly, a candidate series branch combination can be generated for each homogeneous battery cell cluster, outputting K candidate series branch combinations. These combinations achieve an equilibrium state in electromagnetic field distribution, which helps to improve the power generation efficiency and stability of the photovoltaic system.

[0034] Furthermore, step P33 in this embodiment of the application also includes: P33-1: Calculate the first Joule heat accumulation by combining the first carrier migration current and the first equivalent DC impedance spectrum of the first initial series path; P33-2: Dynamically set the quantum annealing segmentation parameters according to the entropy increase deviation of the first Joule heat accumulation from the preset component power threshold; P33-3: Perform short-circuit point perturbation on the first initial series path according to the quantum annealing segmentation parameters to locate multiple quantized disconnection point sequences; P33-4: Use the multiple quantized disconnection point sequences to perform discontinuous topological segmentation of the first initial series path to obtain the multi-group parallel branch.

[0035] Specifically, to further improve the global balance of cross-component series paths in terms of electromagnetic coupling, impedance distribution and heat generation behavior, more refined physical quantity-driven settings can be made for the quantum annealing segmentation process.

[0036] First, the first Joule heat accumulation is calculated by combining the carrier migration current and the first equivalent DC impedance spectrum of the first initial series path. The carrier migration current refers to the current of electrons or holes flowing in the battery cell, while the Joule heat accumulation refers to the heat accumulation caused by current flowing through resistance. By integrating the product of the carrier migration current and the path's equivalent impedance, the total Joule heat generation rate of the path over a typical operating cycle can be obtained. The Joule heat accumulation not only reflects the thermal sensitivity of local nodes or cross-module connection sections but also reveals potential hotspot sections. It is an important input parameter for subsequent path perturbation and segmentation optimization, and can be used to assess potential uneven heat distribution problems in the initial series path.

[0037] Next, based on the deviation between the first Joule heat accumulation and the preset component power threshold, the entropy increase deviation is calculated, and the quantum annealing segmentation parameters are dynamically set accordingly. The entropy increase deviation measures the deviation between the actual heating behavior and the ideal steady-state power distribution, and its calculation can be based on a statistical entropy model of the thermal distribution of path nodes. If the entropy increase deviation is large, it indicates significant local hotspots and a high degree of impedance imbalance in the path. In this case, the system will automatically increase the sensitivity of the quantum annealing segmentation, making the annealing model more inclined to generate more candidate breakpoints, thereby increasing the search space. Conversely, the segmentation density can be reduced to avoid excessive perturbation of paths that already have good balance.

[0038] Then, based on the dynamically updated quantum annealing segmentation parameters, short-circuit point perturbations are performed in the first initial series path to locate multiple quantized breakpoint sequences. This perturbation process, by applying virtual perturbations simulating short circuits or potential collapses to candidate locations on the path, enables the quantum annealing model to identify the node positions in the path most sensitive to changes in impedance gradients, and defines these positions as quantized breakpoints. The quantum breakpoint sequence is essentially a set of discrete nodes that satisfy the minimum transition condition of the energy potential well, representing the physically possible optimal tangent points of the series path.

[0039] Finally, a discontinuous topological partitioning of the first initial series path is performed using multiple quantized breakpoint sequences, resulting in multiple parallel branch paths. This topological partitioning does not change the battery cell set within the cluster, but by structurally decomposing the path at multiple quantum breakpoints, the originally single initial series path evolves into multiple candidate parallel branch structures capable of participating in the quantum annealing iterative solution. Each parallel branch represents a potential physically feasible state, which can compete in subsequent electromagnetic field equilibrium calculations through multi-state parallel solutions, ultimately leading to a series path with minimum impedance fluctuation, lowest Joule heat accumulation, and optimal magnetic coupling equilibrium.

[0040] Furthermore, step P34 in this embodiment of the application also includes: P34-1: Generate a magnetic field intensity distribution cloud map based on the spatial distribution coordinates of the Nth group of conductive cables in the Nth group of parallel branches and the direction of the Nth group of carrier migration currents; P34-2: Identify the magnetic field exceeding interference region in the magnetic field intensity distribution cloud map; P34-3: Locate multiple conductor structure adjustment points in the Nth group of parallel branches according to the magnetic field exceeding interference region; P34-4: After updating the spatial distribution coordinates of the Nth group of conductive cables according to the multiple conductor structure adjustment points, calculate the eddy current heat accumulation; P34-5: Iteratively update the multiple conductor structure adjustment points according to the entropy increase gradient of the eddy current heat accumulation deviating from the thermal stability margin, until the Nth alternative series branch is output.

[0041] In one possible embodiment of this application, in order to ensure that the multi-group parallel branches generated after quantum annealing achieve an engineering-grade balanced state in terms of electromagnetic coupling, conductor wiring and thermal characteristics that can be used for cross-component series electrical reconfiguration, it is necessary to further perform refined balanced iterative optimization driven by electromagnetic field on each group of parallel branches.

[0042] First, a magnetic field strength distribution cloud map is generated based on the spatial distribution coordinates of the conductive cables in the Nth group of parallel branches and the direction of carrier migration current. This step involves using electromagnetic principles and computational models to simulate the magnetic field distribution generated when current flows through the conductive cables. The magnetic field strength distribution cloud map can intuitively display the spatial distribution characteristics of the magnetic field, providing a basis for subsequent analysis and adjustments.

[0043] Next, the magnetic field strength distribution cloud map is analyzed to identify the magnetic field interference regions exceeding the standard. These regions refer to areas where the magnetic induction intensity exceeds the preset magnetic field safety threshold, or where there is significant magnetic coupling superposition, high eddy current density, or locally dense magnetic flux. These regions typically correspond to path turning points, cross-component conductor band confluence sections, closely spaced double-line segments, or neighborhoods with parasitic metal structures, and are key sources affecting series path stability and polarization interference.

[0044] After identifying the aforementioned regions, multiple conductor structure adjustment points are located in the Nth group of parallel branches based on the areas where the magnetic field exceeds the interference limit. These conductor structure adjustment points are locations of conductive cables that may require adjustment, typically conductor nodes, inflection points, cable tray gaps, bridging sections, or wiring sections where structural fine-tuning is possible. By precisely locating these adjustment points, the electromagnetic field distribution can be optimized in a targeted manner.

[0045] Then, after updating the spatial distribution coordinates of the Nth group of conductive cables based on multiple conductor structure adjustment points, the eddy current heat accumulation is calculated. Eddy current heat accumulation refers to the heat accumulation caused by eddy currents generated in the conductor due to electromagnetic induction. It can be calculated based on the integral method of the eddy current loss model, by using the electromagnetic field change rate, conductor cross-sectional parameters, and magnetic flux density gradient to obtain the local heat generation intensity, thereby assessing whether the conductor still has excessive energy accumulation and thermal instability risk after correction.

[0046] Finally, based on the entropy gradient of the eddy current heat accumulation deviating from the thermal stability margin, the aforementioned adjustment points of the conduction band structure are iteratively updated. The entropy gradient reflects the deviation between the current eddy current heat distribution and the system's target thermal stability margin. By using it as the direction of iterative updates, multiple rounds of perturbation corrections can be performed on the conduction band structure, including conduction band translation, rotation, increase or decrease in line spacing, and adjustment of winding sequence. This iterative process terminates with eddy current heat convergence, magnetic field strength peak elimination, and maximization of structural stability, ultimately outputting the Nth alternative series branch that satisfies the electromagnetic field equilibrium constraint. This branch achieves an equilibrium state in electromagnetic field distribution, satisfying the thermal stability requirements.

[0047] Furthermore, steps P34-5 in the embodiments of this application also include: P34-51: Determine the displacement direction of the conductive cable based on the entropy increase gradient of the eddy current heat accumulation deviating from the thermal stability margin; P34-52: Using the branch safety distance as a constraint, randomly change the multiple conductor structure adjustment points according to the displacement direction of the conductive cable to obtain multiple corrected structure adjustment points; P34-53: Adjust the spatial distribution coordinates of the Nth group of conductive cables according to the multiple corrected structure adjustment points to obtain the spatial coordinates of the Nth group of corrected conductive cables, and then calculate the updated heat accumulation; P34-54: If the updated heat accumulation meets the thermal stability margin, then perform a thermal stability test; P34-55: If the thermal stability test passes, then use the series branch topology under the spatial coordinates of the Nth group of corrected conductive cables as the Nth alternative series branch output.

[0048] Specifically, in order to further improve the thermal stability and magnetic interference suppression capability of the Nth group of parallel branches at the conductor strip routing level, the entropy increase gradient-driven update process can be more precisely controlled, so that the iterative adjustment of the conductor strip geometry path is both physically reasonable and engineering feasible.

[0049] First, the displacement direction of the conductive cable is determined based on the entropy gradient of the eddy current heat accumulation deviating from the thermal stability margin. The entropy gradient refers to the rate at which the system's entropy (i.e., disorder) changes with the amount of heat accumulation, indicating the uniformity of the system's heat distribution. By analyzing the entropy gradient, it can be determined which direction the conductive cable should move to reduce eddy current heat accumulation and improve the system's thermal stability.

[0050] Next, using the branch safety spacing as a constraint, multiple conductor strip structure adjustment points are randomly changed according to the displacement direction of the conductive cable, resulting in multiple modified structure adjustment points. The branch safety spacing refers to the minimum geometric distance required between different conductor strips to avoid insulation risks, excessive magnetic field coupling, and ensure construction accessibility. By using this spacing as a boundary condition and performing strip-constrained random perturbations in the displacement direction, the neighborhood feasible solution space can be effectively explored, preventing the conductor strip movement from getting trapped in local optima.

[0051] Subsequently, the spatial distribution coordinates of the Nth group of conductive cables are updated based on multiple correction structure adjustment points to form the spatial coordinates of the Nth group of corrected conductive cables, and the updated eddy current heat accumulation is calculated based on this. The updated heat accumulation is used to re-evaluate the magnetic field diffusion, eddy current loss distribution, and whether the hot spot section has been effectively reduced after the conductor space adjustment, and is a key indicator for measuring the effectiveness of this iteration.

[0052] Once the updated eddy current heat accumulation meets the thermal stability margin, further thermal stability testing is conducted. This test simulates the thermal behavior of the conductor under multiple operating conditions, including different irradiance levels, different temperature ranges, and different load conditions, to verify whether it can maintain thermal equilibrium and avoid local overheating in a dynamic working environment.

[0053] If the thermal stability test passes, the series branch topology in the Nth group of corrected conductive cable spatial coordinates will be used as the Nth alternative series branch output. At this point, the series branch has met a series of multi-physics constraints such as magnetic field interference suppression, eddy current heat minimization, impedance consistency, and safety spacing, and is the final alternative path structure that can be used for cross-component series reconfiguration and independent inverter access.

[0054] Furthermore, steps P34-5 in the embodiments of this application also include: P34-56: If the entropy increase gradient of the accumulated eddy current heat deviating from the thermal stability margin exceeds the preset cable temperature change threshold, then the Nth group of parallel branches will be designated as forbidden branches.

[0055] Optionally, if the calculated eddy current heat accumulation deviates from the entropy increase gradient of the thermal stability margin by more than the preset cable temperature change threshold, this means that the heat generated when current passes through the conductive cable may be too high, potentially causing the cable to overheat and affecting the safety and reliability of the entire photovoltaic system. In this case, to prevent potential safety risks, a conservative measure will be taken, and the Nth group of parallel branches will be marked as forbidden branches.

[0056] Taboo branches are those that are temporarily excluded from the current optimization process and will not be used to construct the final alternative series branch combinations. By marking branches that exceed the safety threshold as taboo branches, it can be ensured that these potentially problematic branches will not be used in subsequent optimization processes. This ensures that the final output series branches are engineering implementable and have long-term operational safety, preventing any paths with thermal instability risks from entering the system electrical architecture refactoring phase.

[0057] P40: By performing global impedance equalization verification on the K candidate series branch combinations, K target series branch combinations are selected and output.

[0058] Furthermore, step P40 in this embodiment of the application also includes: P41: Combine and enumerate the K candidate series branch combinations to obtain multiple structural topology schemes; P42: Perform branch DC resistance deviation verification on the multiple structural topology schemes and output multiple impedance equalization parameter groups; P43: Based on the descending order of multiple temperature rise data in the multiple impedance equalization parameter groups, select and locate the target topology scheme from the multiple structural topology schemes; P44: Decompose the target topology scheme to obtain the K target series branch combinations.

[0059] It should be understood that, in order to ensure that the K candidate series branch combinations formed after quantum annealing, electromagnetic field equalization and thermal stability verification can achieve a consistent, stable and low-loss cross-component series operation mode at the system level, it is necessary to further perform global impedance equalization analysis on the overall path combination, and screen out the K target series branch combinations that can be used for electrical architecture reconfiguration at the system level.

[0060] First, the K candidate series branch combinations are enumerated to obtain multiple possible structural topologies. This step involves systematically listing all possible branch combinations to explore different electrical architecture configurations. Each structural topology represents a potential electrical connection method that will affect the electrical performance of the entire photovoltaic system.

[0061] Next, the DC resistance deviation of these multiple structural topologies is verified. During this verification process, the DC resistance, equivalent voltage loss, parasitic parameters, and impedance consistency deviation of each series branch in each structural topology are calculated to form a corresponding impedance balancing parameter set. The magnitude of the impedance deviation is used to measure the consistency of electrical performance of different branches under independent inverter operation, and is an important basis for avoiding voltage mismatch, current deviation, and inverter MPPT disturbance between branches.

[0062] Next, based on the multiple temperature rise data points included in the aforementioned parameter set, they are sorted in descending order of temperature rise magnitude. By comparing the conduction band temperature rise characteristics of different structural topologies under typical operating conditions, the target topology is screened and located. The temperature rise data reflects the thermal response capability of the series branch under different resistance deviations, load conditions, and parasitic parameter disturbances. The lower the temperature rise, the better the thermal stability, impedance consistency, and energy loss of the path topology. Based on this ranking result, the target topology with the best thermal stability and most balanced impedance can be selected.

[0063] Finally, the target topology scheme is decomposed, and the corresponding K target series branch combinations are reconstructed and extracted from its overall combined structure. This step involves decomposing the selected target topology scheme into specific branch combinations, which will be directly used for the electrical architecture reconfiguration of the photovoltaic system. These K target series branch combinations represent the final series path structure that meets the requirements in terms of global impedance balance, thermal stability, and path consistency. They are the only effective set of paths that can be used for actual engineering deployment when performing cross-module electrical architecture reconfiguration.

[0064] P50: Using the K target series branch combinations, the electrical architecture of K groups of in-service battery cells in the photovoltaic system is reconfigured to obtain K groups of independent DC routing channels, wherein each independent DC routing channel is connected to a dedicated inverter.

[0065] Specifically, the K target series branch combinations selected earlier are used to construct K independent DC routing channels, and a dedicated inverter is configured for each channel. In the final electrical architecture, each series branch is a single-string output, and the electrical connections between components adopt a single-string interconnect structure. Each independent DC routing channel is connected to a dedicated inverter to achieve independent power conversion and maximum power point tracking control. All inverters control the power conversion on their respective paths in the system, and the branches do not interfere with each other, ensuring that each path can be optimized independently.

[0066] First, based on the arrangement and spatial position of the battery cells in each target series branch combination, the series and parallel connections within the original module are deconstructed and reconstructed. By reorganizing the battery strings across modules according to the target series relationship, battery cells from different modules but with isomorphic relationships in electrical performance are connected in series with a unified sequence number, forming a continuous high-voltage, low-current series path across modules. This process decouples the traditional fixed series and parallel arrangement within the module into a cross-module sequential series structure dynamically reconstructed based on the target branch.

[0067] After completing the cross-component connection reconstruction between battery strings, DC terminals are independently led out for each target series branch to construct K completely isolated independent DC routing channels. Each DC routing channel includes the conductor structure, line routing, insulation protection section and end output terminal of its corresponding series path, ensuring physical isolation between different channels in terms of space, electromagnetics and heat, and avoiding mutual interference between branches.

[0068] Finally, to achieve efficient electrical conversion and independent maximum power point tracking (MPPT) control, the aforementioned K independent DC routing channels are connected to the corresponding inverter's dedicated MPPT interface. Each inverter connected to a channel is equipped with an independent MPPT control unit, which can perform optimized control in real time based on the voltage, current, power, and operating status of the corresponding series path. Through this path-level independent inverter structure, each high-voltage, low-current series branch can maintain its optimal power generation point without being affected by other branches, thereby significantly improving the overall photovoltaic system's energy efficiency, reliability, and operational safety.

[0069] Furthermore, to ensure long-term stable operation in complex environments, the series branches are categorized according to the feature vectors of all their battery cells, and the inverters are precisely matched based on this categorization.

[0070] In summary, the embodiments of this application have at least the following technical effects: This application improves the electrical consistency of series path construction by achieving electrical isomorphic grouping across modules through real-time feature vectors of in-service battery cells; it optimizes the physical characteristics of cross-module series paths through electromagnetic field equalization and eddy current thermal confinement to suppress magnetic coupling and heat accumulation; it achieves system-level impedance equalization through multi-scheme impedance deviation verification and temperature rise assessment; it performs dynamic electrical architecture reconstruction of the photovoltaic array through target branch combination to overcome the fixed series and parallel limitations of modules; and it enhances the adaptability and scalability of the photovoltaic system by forming multiple DC routing channels with independent output and independent control capabilities.

[0071] The technology achieves the goal of improving power generation efficiency and reducing system safety risks by optimizing the series path of battery cells and the configuration of independent inverters.

[0072] Example 2, based on the same inventive concept as the power generation optimization control method for multi-string independent inverters of heterojunction photovoltaic modules in the foregoing examples, such as... Figure 2 As shown, this application provides a power generation optimization control system for multi-string independent inverters of heterojunction photovoltaic modules. The system and method embodiments in this application are based on the same inventive concept. The system includes: The battery feature vector acquisition module 11 is used to collect real-time operating data of photovoltaic system battery cells and construct multiple battery feature vectors, wherein the battery feature vectors include open circuit voltage, short circuit current, annual degradation rate, location altitude and geographical coordinates.

[0073] The electrical topology clustering module 12 is used to cluster the electrical topology of in-service battery blocks across components based on the multiple battery feature vectors, and generate K homogeneous battery block clusters.

[0074] The series path optimization module 13 is used to optimize the series path of the K isomorphic battery block clusters based on electromagnetic field equilibrium and output K alternative series branch combinations.

[0075] The global impedance equalization verification module 14 is used to perform global impedance equalization verification on the K candidate series branch combinations and filter out K target series branch combinations.

[0076] The electrical architecture reconfiguration module 15 is used to reconfigure the electrical architecture of K groups of in-service battery cells in the photovoltaic system using the K target series branch combinations to obtain K groups of independent DC routing channels, wherein each independent DC routing channel is connected to a dedicated inverter.

[0077] Furthermore, the battery feature vector acquisition module 11 is also used to perform the following steps: The real-time operating data is parsed to obtain multiple multi-dimensional status logs of multiple in-service battery cells in the photovoltaic system; multiple initial feature vectors are constructed based on the multiple multi-dimensional status logs; the baseline parameters of the photovoltaic system are predefined, and the multiple initial feature vectors are subjected to data standardization processing to output multiple battery feature vectors, wherein the battery feature vectors include open-circuit voltage, short-circuit current, annual degradation rate, location altitude, and geographical coordinates.

[0078] Furthermore, the electrical topology clustering module 12 is also used to perform the following steps: Predefined battery block similarity constraints and battery block distance constraints are used; the battery block similarity constraints are used to enumerate and compare the feature vectors of the multiple batteries to construct multiple electrically connected subgraphs; multiple candidate isomorphic clusters are obtained by aligning and overlapping the multiple electrically connected subgraphs; multiple geographic coordinates are extracted from the multiple battery feature vectors, and the battery block distance constraints are used to traverse the multiple geographic coordinates to perform spatial density verification and splitting of the multiple candidate isomorphic clusters to obtain the K isomorphic battery block clusters.

[0079] Furthermore, the serial path optimization module 13 is also used to perform the following steps: Based on geographical coordinates, the shortest path topology of the first isomorphic battery block cluster is constructed to obtain the first initial series path; the first equivalent DC impedance spectrum of the first initial series path is calculated; the quantum annealing segmentation parameters are set according to the first equivalent DC impedance spectrum to divide the first initial series path, thereby obtaining multi-group parallel branches corresponding to multiple quantized breakpoint sequences; electromagnetic field equalization iteration is performed on the multi-group parallel branches to output the first candidate series branch combination.

[0080] Furthermore, the serial path optimization module 13 is also used to perform the following steps: The first Joule heat accumulation is calculated by combining the first carrier migration current and the first equivalent DC impedance spectrum of the first initial series path; the quantum annealing segmentation parameters are dynamically set according to the entropy increase deviation of the first Joule heat accumulation from the preset component power threshold; short-circuit point perturbation is performed on the first initial series path according to the quantum annealing segmentation parameters to locate multiple quantized disconnection point sequences; the discontinuous topology segmentation of the first initial series path is performed using the multiple quantized disconnection point sequences to obtain the multi-group parallel branch.

[0081] Furthermore, the serial path optimization module 13 is also used to perform the following steps: Based on the spatial distribution coordinates of the Nth group of conductive cables in the Nth group of parallel branches and the direction of the Nth group of carrier migration currents, a magnetic field intensity distribution cloud map is generated; the magnetic field exceeding interference region in the magnetic field intensity distribution cloud map is identified; multiple conductor structure adjustment points are located in the Nth group of parallel branches according to the magnetic field exceeding interference region; after updating the spatial distribution coordinates of the Nth group of conductive cables according to the multiple conductor structure adjustment points, the eddy current heat accumulation is calculated; according to the entropy increase gradient of the eddy current heat accumulation deviating from the thermal stability margin, the multiple conductor structure adjustment points are iteratively updated until the Nth alternative series branch is output.

[0082] Furthermore, the serial path optimization module 13 is also used to perform the following steps: Based on the entropy increase gradient of the eddy current heat accumulation deviating from the thermal stability margin, the displacement direction of the conductive cable is determined; with the branch safety distance as a constraint, the multiple conductor structure adjustment points are randomly changed according to the displacement direction of the conductive cable to obtain multiple corrected structure adjustment points; the spatial distribution coordinates of the Nth group of conductive cables are adjusted according to the multiple corrected structure adjustment points to obtain the spatial coordinates of the Nth group of corrected conductive cables, and then the updated heat accumulation is calculated; if the updated heat accumulation meets the thermal stability margin, a thermal stability test is performed; if the thermal stability test passes, the series branch topology under the spatial coordinates of the Nth group of corrected conductive cables is used as the Nth alternative series branch output.

[0083] Furthermore, the serial path optimization module 13 is also used to perform the following steps: If the entropy increase gradient of the accumulated eddy current heat deviating from the thermal stability margin exceeds the preset cable temperature change threshold, then the Nth group of parallel branches will be designated as forbidden branches.

[0084] Furthermore, the global impedance equalization verification module 14 is also used to perform the following steps: The K candidate series branch combinations are enumerated to obtain multiple structural topology schemes; the DC resistance deviation of the branches of the multiple structural topology schemes is checked to output multiple impedance equalization parameter groups; the target topology scheme is selected from the multiple structural topology schemes based on the descending order of multiple temperature rise data in the multiple impedance equalization parameter groups; the target topology scheme is decomposed to obtain the K target series branch combinations.

[0085] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0086] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0087] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A method for optimizing power generation control of multi-string independent inverters in heterojunction photovoltaic modules, characterized in that, The method includes: Real-time operating data of photovoltaic system cell blocks are collected to construct multiple cell feature vectors, wherein the cell feature vectors include open-circuit voltage, short-circuit current, annual degradation rate, location height and geographic coordinates; Based on the multiple battery feature vectors, the in-service battery cells are clustered across the component electrical topology to generate K homogeneous battery cell clusters; Based on electromagnetic field equilibrium, the series path of the K isomorphic battery block clusters is optimized, and K alternative series branch combinations are output. By performing global impedance equalization verification on the K candidate series branch combinations, K target series branch combinations are selected and output. Using the K target series branch combinations, the electrical architecture of K groups of in-service battery cells in the photovoltaic system is reconfigured to obtain K groups of independent DC routing channels, wherein each independent DC routing channel is connected to a dedicated inverter; The method involves optimizing the series path of the K isomorphic battery cell clusters based on electromagnetic field equilibrium, and outputting K alternative series branch combinations. Based on geographical coordinates, the shortest path topology of the first isomorphic battery block cluster is constructed to obtain the first initial serial path; Calculate the first equivalent DC impedance spectrum of the first initial series path; Based on the first equivalent DC impedance spectrum, set the quantum annealing segmentation parameters, divide the first initial series path, and obtain multi-group parallel branches corresponding to multiple quantized circuit break sequence; Electromagnetic field equalization iteration is performed on the multi-group parallel branches to output the first candidate series branch combination; The method involves setting quantum annealing segmentation parameters based on the first equivalent DC impedance spectrum, dividing the first initial series path, and obtaining multi-group parallel branches corresponding to multiple quantized circuit breaker sequences. The method includes: The first Joule heat accumulation is calculated by combining the first carrier migration current of the first initial series path and the first equivalent DC impedance spectrum. The quantum annealing segmentation parameters are dynamically set based on the degree of entropy increase deviation of the first Joule heat accumulation from the preset component power threshold. Based on the quantum annealing segmentation parameters, short-circuit point perturbation is performed in the first initial serial path to locate multiple quantized breakpoint sequences; The discontinuous topological segmentation of the first initial serial path is performed using the multiple quantized circuit breaker sequences to obtain the multiple sets of parallel branch paths; The method involves performing global impedance equalization verification on the K candidate series branch combinations to select and output K target series branch combinations. By combining and enumerating the K candidate series branch combinations, multiple structural topology schemes are obtained; The branch DC resistance deviation of the multiple structural topology schemes is checked, and multiple impedance equalization parameter groups are output. Based on the descending order of multiple temperature rise data in the multiple impedance equalization parameter groups, the target topology scheme is selected from the multiple structural topology schemes. The target topology scheme is decomposed to obtain the K target series branch combinations.

2. The power generation optimization control method for multi-string independent inverters of heterojunction photovoltaic modules as described in claim 1, characterized in that, Electromagnetic field equalization iteration is performed on the multi-group parallel branches to output a first candidate series branch combination. The method includes: Based on the spatial distribution coordinates of the Nth group of conductive cables in the Nth group of parallel branches and the direction of the Nth group of carrier migration current, a magnetic field intensity distribution cloud map is generated. Identify the interference regions where the magnetic field exceeds the standard in the magnetic field intensity distribution cloud map; Based on the magnetic field exceeding interference region, multiple conduction band structure adjustment points are located in the Nth group of parallel branches; After updating the spatial distribution coordinates of the Nth group of conductive cables based on the multiple conductor structure adjustment points, the eddy current heat accumulation is calculated. Based on the entropy increase gradient of the eddy current heat accumulation deviating from the thermal stability margin, the multiple conductor structure adjustment points are iteratively updated until the Nth alternative series branch is output.

3. The power generation optimization control method for multi-string independent inverters of heterojunction photovoltaic modules as described in claim 2, characterized in that, Based on the entropy increase gradient of the eddy current heat accumulation deviating from the thermal stability margin, the multiple conduction band structure adjustment points are iteratively updated until the Nth alternative series branch is output. The method includes: The displacement direction of the conductive cable is determined based on the entropy increase gradient of the eddy current heat accumulation deviating from the thermal stability margin. Using the branch safety distance as a constraint, the multiple conductor structure adjustment points are randomly changed according to the displacement direction of the conductive cable to obtain multiple modified structure adjustment points; The spatial distribution coordinates of the Nth group of conductive cables are adjusted according to the multiple correction structure adjustment points to obtain the spatial coordinates of the Nth group of corrected conductive cables. Then, the updated heat accumulation is calculated. If the updated heat accumulation meets the thermal stability margin, then a thermal stability test is performed; If the thermal stability test is passed, the series branch topology under the Nth group of corrected conductive cable spatial coordinates will be used as the Nth alternative series branch output.

4. The power generation optimization control method for multi-string independent inverters of heterojunction photovoltaic modules as described in claim 2, characterized in that, If the entropy increase gradient of the accumulated eddy current heat deviating from the thermal stability margin exceeds the preset cable temperature change threshold, then the Nth group of parallel branches will be designated as forbidden branches.

5. The power generation optimization control method for multi-string independent inverters of heterojunction photovoltaic modules as described in claim 1, characterized in that, The method involves collecting real-time operating data of photovoltaic system cell blocks and constructing multiple cell feature vectors, including: The real-time operating data is analyzed to obtain multiple multi-dimensional status logs of multiple in-service battery cells in the photovoltaic system; Multiple initial feature vectors are constructed based on the aforementioned multi-dimensional state logs; The baseline parameters of the photovoltaic system are predefined, and the multiple initial feature vectors are processed by data standardization to output multiple cell feature vectors. The cell feature vectors include open circuit voltage, short circuit current, annual degradation rate, location altitude, and geographic coordinates.

6. The power generation optimization control method for multi-string independent inverters of heterojunction photovoltaic modules as described in claim 1, characterized in that, Based on the multiple battery feature vectors, the in-service battery cells are clustered across the module electrical topology to generate K homogeneous battery cell clusters. The method includes: Predefined battery block similarity constraints and battery block distance constraints; Using the battery block similarity constraint, the feature vectors of the multiple batteries are enumerated and compared to construct multiple electrically connected subgraphs; Multiple candidate isomorphic clusters are obtained by aligning and overlapping the multiple electrical connectivity subgraphs; Multiple geographic coordinates are extracted from the multiple battery feature vectors, and the multiple geographic coordinates are traversed using the battery block distance constraint to perform spatial density verification and splitting of the multiple candidate isomorphic clusters, thereby obtaining the K isomorphic battery block clusters.

7. A power generation optimization control system for multi-string independent inverters of heterojunction photovoltaic modules, characterized in that, The system is used to perform the power generation optimization control method for multi-string independent inverter of heterojunction photovoltaic modules according to any one of claims 1 to 6, the system comprising: The battery feature vector acquisition module is used to collect real-time operating data of photovoltaic system battery cells and construct multiple battery feature vectors, wherein the battery feature vectors include open circuit voltage, short circuit current, annual degradation rate, location altitude and geographical coordinates; The electrical topology clustering module is used to cluster the in-service battery blocks across the component electrical topology based on the multiple battery feature vectors, generating K homogeneous battery block clusters. The series path optimization module is used to optimize the series path of the K isomorphic battery block clusters based on electromagnetic field equilibrium and output K alternative series branch combinations. The global impedance equalization verification module is used to perform global impedance equalization verification on the K candidate series branch combinations and filter out K target series branch combinations. The electrical architecture reconfiguration module is used to reconfigure the electrical architecture of K groups of in-service battery cells in the photovoltaic system using the K target series branch combinations to obtain K groups of independent DC routing channels, wherein each independent DC routing channel is connected to a dedicated inverter.

Images