Solar photovoltaic sunshade device efficiency intelligent monitoring system

Abstract

The invention relates to the technical field of photovoltaic sun-shading efficiency management, and relates to an intelligent monitoring system for efficiency of a solar photovoltaic sun-shading device, which synchronously acquires real-time thermoelectric coupling data during daytime operation and electrochemical impedance spectroscopy data in a silent state at night. Based on the multi-dimensional data, the central processing module does not perform simple threshold comparison, but constructs a dynamic characteristic model through fusion analysis of hot spot characteristics and impedance fingerprints, and intelligently calculates an optimal matching current value under the condition of considering the power generation efficiency and thermal safety constraints. By accurately distinguishing the aging types, misjudgment caused by temporary shadows or stains can be effectively avoided, the accuracy of fault positioning and aging state evaluation is greatly improved, a reliable basis is provided for subsequent refined maintenance decisions, and the accuracy of fault positioning and aging state evaluation is improved. The hot spot fire risk possibly caused by the fact that local serious degradation of the assembly is not found in time is prevented from the source, and long-term operation safety of the building attached photovoltaic system is guaranteed.

Description

Technical Field

[0001] This invention belongs to the field of photovoltaic shading efficiency management technology, and relates to an intelligent monitoring system for the efficiency of solar photovoltaic shading devices. Background Technology

[0002] Building-integrated photovoltaics (BIPV) systems, as key energy-generating components of modern green buildings, deeply integrate power generation with building materials, playing an irreplaceable role in achieving building-specific energy supply and reducing carbon emissions. Meanwhile, the photovoltaic modules, as the core power generation units of BIPV systems, directly determine the system's reliability and economic benefits over decades through their long-term operational health and performance stability, serving as the physical foundation for the continuous realization of asset value.

[0003] It is worth noting that BIPV modules are typically integrated into the building facade, and their service environment is more demanding and complex than that of ground-mounted power plants. Long-term exposure to the building's local microclimate (such as uneven shading, concentrated sunlight reflected by glass curtain walls, and heat buildup due to poor ventilation) not only directly accelerates the aging of encapsulation materials and induces microcracks in the cells, but may also cause distinct thermo-electrical stresses in each power generation unit. Therefore, meticulous management of the module's health status is particularly urgent in maintaining the overall performance and safety of the BIPV system. Especially for BIPV systems that have been operating for many years and integrated into the facades or shading components of high-rise buildings, there is a dynamic and close coupling relationship between the discreteness of module performance and its output power, system stability, and potential thermal safety risks.

[0004] However, current operation and maintenance management of such aging BIPV arrays has significant shortcomings. Traditional system-level monitoring can only detect the overall power decline of the string, but cannot locate the internal "short-circuit" unit. Faced with unevenly aged components, existing methods are caught in a dilemma: if the failed unit is simply bypassed, the operating voltage of the entire string may drop below the inverter's start-up threshold, causing the entire string to shut down and generating zero power; if it is allowed to continue operating, the severely aged short-circuit component will act as a blockage point, not only limiting the current of the entire string, but also converting precious solar energy into dangerous Joule heat due to the surge in its internal resistance, creating a risk of hot spot burn-out or even fire. Summary of the Invention

[0005] In view of the problems existing in the prior art, the present invention provides an intelligent monitoring system for the efficiency of solar photovoltaic shading devices to solve the above-mentioned technical problems.

[0006] To achieve the above and other objectives, the technical solution adopted by the present invention is as follows:

[0007] This invention provides an intelligent monitoring system for the efficiency of a solar photovoltaic shading device, the system comprising:

[0008] The data acquisition module is used to acquire daytime operating thermoelectric data and nighttime static impedance data of each sub-module in the photovoltaic shading device;

[0009] The central processing module is used to construct aging characteristic models of each sub-module based on daytime thermoelectric data and nighttime static impedance data, and to calculate the optimal matching current value of each sub-module based on the aging characteristic models.

[0010] The central processing module also performs clustering and hierarchical operations based on the optimal matching current value of each sub-module, classifying sub-modules with optimal matching current values ​​within a predetermined similar range into the same health level and generating corresponding topology instructions.

[0011] The matrix topology execution module is configured to respond to the topology command, adjust the state of the switch array connected between each sub-module, construct the sub-modules divided into the same health level into a first series string, and connect the first series string to the inverter input terminal.

[0012] Another aspect of the present invention provides an intelligent monitoring device for the efficiency of a solar photovoltaic shading device, characterized in that it includes a processor, a memory, and a communication bus;

[0013] The memory stores a computer-readable program that can be executed by the processor;

[0014] The communication bus enables communication between the processor and the memory;

[0015] When the processor executes the computer-readable program, it performs a module to implement a solar photovoltaic shading device efficiency intelligent monitoring system as described in any one of the present invention.

[0016] As described above, the intelligent monitoring system for the efficiency of a solar photovoltaic shading device provided by the present invention has at least the following beneficial effects:

[0017] This invention provides an intelligent monitoring system for the performance of a solar photovoltaic shading device. It acquires real-time thermoelectric coupling data during daytime operation and electrochemical impedance spectroscopy data during nighttime quiescent conditions. Based on this multi-dimensional data, the central processing module does not perform simple threshold comparisons. Instead, it integrates and analyzes hotspot characteristics and impedance fingerprints to construct a dynamic characteristic model that accurately reflects the internal physical aging mechanism of each sub-module. This model intelligently calculates the "optimal matching current value" while balancing power generation efficiency and thermal safety constraints. By accurately distinguishing aging types, misjudgments caused by temporary shading or stains can be effectively avoided, greatly improving the accuracy of fault location and aging status assessment. This provides a reliable basis for subsequent refined maintenance decisions, preventing the risk of hotspot fires that may arise from undetected severe localized degradation of components, and ensuring the long-term operational safety of building-attached photovoltaic systems.

[0018] This invention further utilizes a central processing module to perform intelligent clustering and hierarchical calculations based on the "optimal matching current value." This automatically groups sub-modules with highly similar performance characteristics into the same health level and generates dynamic topology reconfiguration instructions accordingly. The matrix-style topology execution module responds to these instructions by adjusting the conduction state of the high-reliability switch array, reconstructing the electrical connections in real time at the physical level. This flexibly assembles sub-modules of the same level into independent, "homogenized" power generation branches. This not only releases the overall performance of the string being dragged down by aging components, effectively overcoming the "weakest link" effect and improving the overall power generation revenue of the system; more importantly, by connecting components with similar characteristics in series, it ensures that each component operates under optimal conditions matching its own capabilities. This avoids the unfavorable situation of high-performance groups being forcibly derated and low-performance groups overheating due to overload. Thus, at the system level, it achieves synergistic optimization of maximizing power generation and minimizing operational thermal risks, significantly enhancing the robustness and economy of photovoltaic arrays under complex aging scenarios. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the 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.

[0020] Figure 1 This is a schematic diagram showing the connections of the various modules in the system of the present invention.

[0021] Figure 2 This is a schematic diagram of the module operation logic connection of the present invention.

[0022] Figure 3 This is a schematic diagram of the device logic connection of the present invention. Detailed Implementation

[0023] The following description, in conjunction with the implementation of this invention, is merely an example and illustration of the concept of this invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the inventive concept or exceed the scope defined in these claims, all of which should fall within the protection scope of this invention.

[0024] This invention addresses the technical challenge of extremely uneven component aging in BIPV shading devices applied to complex environments such as high-rise building facades after long-term operation. Traditional monitoring and maintenance methods have fundamental flaws: simply bypassing aging components for safety will cause a drop in the entire string's operating voltage, potentially failing to meet the inverter's minimum input requirements and paralyzing the entire string's power generation function; maintaining the original series topology will cause severely aged "short-board" components to act as bottlenecks, not only limiting the entire string's output current but also generating excessive Joule heat due to a sharp increase in internal resistance, leading to hot spot risks. Existing solutions cannot achieve a dynamic balance between system power generation, operational safety, and topology feasibility, thus limiting the full lifecycle value of BIPV assets.

[0025] To address these issues, the study found that the temperature-power coupling characteristics during daytime operation and the broadband impedance spectrum characteristics under nighttime static conditions together constitute a unique "fingerprint" characterizing the health status of the module. Specifically, daytime thermoelectric data is sensitive to efficiency losses and thermal runaway tendencies caused by aging, while nighttime electrochemical impedance spectroscopy can penetrate the surface and distinguish between different aging mechanisms such as connection corrosion (high-frequency response) and cell-level damage (low-frequency diffusion). Based on this correlation, a precise aging model is proposed by fusing daytime and nighttime dual-mode data, from which the core indicator of "optimal matching current value" is derived. This serves as the optimal operating point of the module within the safety and efficiency boundary, laying a quantitative foundation for the dynamic reconfiguration of the system.

[0026] Specifically, the technical implementation of this solution begins with the synchronous acquisition of day and night dual-mode data for each sub-module: during the day, the coupling relationship between its operating temperature and output power is captured in real time; at night, multi-frequency AC disturbance signals are injected into the array to obtain its electrochemical impedance spectrum. Based on this, the central processing system constructs and continuously updates a refined aging characteristic model for each component, eliminating reversible interferences such as bird droppings and temporary shadows through cross-validation, accurately calculating its equivalent series internal resistance, parallel resistance, and thermal safety boundary, and then calculating a personalized "optimal matching current value". The system then dynamically clusters and classifies all components based on this current value (e.g., healthy, sub-healthy, faulty), and intelligently generates the optimal reconfiguration strategy: grouping components with similar performance into homogeneous series branches, while ensuring that the voltage of each branch meets the inverter requirements. Finally, the matrix topology execution module responds to commands, and through the rapid switching of the solid-state switch array, dynamically reconfigures the circuit connection physically, and can implement proactive "electrically controlled heat" management for branches with overheating risks.

[0027] After introducing the basic concept of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0028] Example 1

[0029] Please see Figures 1-2 As shown, a solar photovoltaic shading device efficiency intelligent monitoring system includes a data acquisition module, a central processing module, and a matrix topology execution module.

[0030] The various modules mentioned above are connected via wired and / or wireless means to enable data transmission between them;

[0031] The data acquisition module is used to acquire daytime operating thermoelectric data and nighttime static impedance data of each sub-module in the photovoltaic shading device.

[0032] Preferably, the operation logic of the data acquisition module includes:

[0033] The surface temperature and output power of each sub-module are monitored in real time during the illumination period. The difference between the surface temperature and the preset ambient temperature reference value is calculated to generate the temperature rise deviation index. At the same time, the ratio between the output power and the theoretical power under the current irradiance is calculated to generate the power efficiency index. The temperature rise deviation index and the power efficiency index are marked as daytime operation thermoelectric data.

[0034] During non-lighting periods, the control signal generator injects test signals in a predetermined high-frequency range and test signals in a predetermined low-frequency range into each sub-module, and collects the first voltage fluctuation data fed back by each sub-module for the test signal in the predetermined high-frequency range, and the second voltage fluctuation data fed back by each sub-module for the test signal in the predetermined low-frequency range.

[0035] The equivalent series internal resistance value characterizing the electrical connection state of the component is obtained by analyzing the first voltage fluctuation data, and the leakage current parallel internal resistance value characterizing the lattice defect state of the component is obtained by analyzing the second voltage fluctuation data. The equivalent series internal resistance value and the leakage current parallel internal resistance value are combined to form the static impedance data at night.

[0036] Specifically, the surface temperature values ​​of each sub-module are synchronously collected every n minutes using a surface-mount thermistor and a DC Hall sensor. and output power values Where n is a user-defined parameter, balancing data real-time performance and storage load, and is preferably set to 20; The unit is Celsius. The unit is watt;

[0037] The collected surface temperature values ​​were substituted into the temperature rise deviation index calculation formula. The solution is obtained in the following process: The ambient temperature is obtained from the ambient temperature sensor, and G is the current irradiance, measured in watts per square meter. The preset thermal radiation temperature rise coefficient, with units of degrees Celsius per square meter per watt, preferably ranging from 0.02 to 0.05, is used to correct for normal temperature rise caused by direct solar radiation. To normalize the reference temperature, it can be set to 25 degrees Celsius, thus obtaining the dimensionless temperature rise deviation index. This index can isolate the environmental impact and characterize the abnormal heat generated by the component due to internal defects.

[0038] Simultaneously calculate the power efficiency index The calculation formula is: ,in This refers to the rated power of the component under standard test conditions, measured in watts, which can be found in the manufacturer's report. The power temperature coefficient can be set to 0.004 / ℃, with units of per degree Celsius. 1000 is the irradiance constant under standard test conditions, with units of W / m². This formula generates a power efficiency index that reflects the degree of photoelectric conversion efficiency degradation by comparing the measured power with the theoretical maximum power under the current operating conditions. Ultimately and Encapsulated as daytime operating thermoelectric data.

[0039] Subsequently, during non-sunlight periods when the irradiance is below the activation threshold, the system switches to a nighttime static impedance data acquisition mode. This mode, based on the principle of electrochemical impedance spectroscopy, controls the signal generator to inject a micro-amplitude AC sinusoidal current signal into the photovoltaic string. The injection logic is divided into two frequency band stages: first, a test signal in a predetermined high-frequency range is injected. Preferably, the frequency is between 1kHz and 10kHz. At this frequency, due to the high-frequency bypass effect of the photovoltaic cell junction capacitance, the parallel branch is short-circuited, and the first voltage fluctuation data collected is... With injected current The ratio mainly reflects the circuit and contact resistance, and is used to formulate... The equivalent series internal resistance value was calculated. Where rms represents the valid value, The voltage and current phase difference at high frequencies; then, a test signal in a predetermined low-frequency range is injected, with a frequency... The preferred frequency is 0.1Hz to 1Hz, at which point the junction capacitance is approximately open, current flows through the parallel leakage current channel, and the second voltage fluctuation data is collected. And based on the formula The leakage current and parallel internal resistance values ​​were calculated in reverse. ,in The voltage and current phase difference at low frequencies will eventually be calculated. and The data are combined to form the static impedance data for nighttime.

[0040] The central processing module is used to construct aging characteristic models of each sub-module based on daytime operating thermoelectric data and nighttime static impedance data, and to calculate the optimal matching current value of each sub-module based on the aging characteristic models.

[0041] Preferably, the aging characteristic model of each sub-module is constructed, including:

[0042] The temperature rise deviation index and power efficiency index are extracted from the daytime operation thermoelectric data. The temperature rise deviation index is then correlated and aligned with the equivalent series internal resistance value and leakage current parallel internal resistance value contained in the nighttime static impedance data in the spatiotemporal dimension to construct a multidimensional health feature vector.

[0043] The multidimensional health feature vector is input into the preset attenuation judgment logic unit. When the temperature rise deviation index is higher than the preset normal threshold, the physical aging coefficient characterizing the degree of irreversible damage to the component material itself is generated by calculating the weighted sum of the rate of change of the equivalent series internal resistance value and the reciprocal rate of change of the leakage current parallel internal resistance value.

[0044] The environmental shielding coefficient, which characterizes the effects of external shielding or pollution, is calculated by using the numerical difference between the power performance index and the physical aging coefficient, thereby separating the recoverable performance loss component caused solely by environmental factors.

[0045] Based on the physical aging coefficient, the series resistance parameters and parallel resistance parameters in the preset standard photovoltaic cell output model are corrected. At the same time, based on the environmental shading coefficient, the photocurrent parameters in the standard photovoltaic cell output model are corrected. A corrected current-voltage characteristic curve that can fit the actual output capability of the current sub-module is generated, and the corrected current-voltage characteristic curve is established as the aging characteristic model of the sub-module.

[0046] Specifically, feature decoupling is performed on daytime thermal power data to extract the dimensionless temperature rise deviation index. With power efficiency index And according to the nearest neighbor matching principle, the corresponding nighttime static impedance data for that period of time—that is, the equivalent series internal resistance value—is used. The value of the internal resistance in parallel with the leakage current By performing spatiotemporal alignment, a multidimensional health feature vector that can comprehensively represent the current state of the component is constructed. The units for both the equivalent series internal resistance and the leakage current parallel internal resistance are ohms.

[0047] Next, the system inputs the vector to the attenuation determination logic unit, first determining whether the temperature rise deviates from the exponential. Whether the temperature exceeds the preset normal thermal fluctuation threshold, which is usually set between 0.05 and 0.1, is determined based on the heat capacity characteristics of the component material. If the temperature exceeds this threshold, abnormal heating is determined, and the physical aging coefficient calculation formula is used. Irreversible damage inside the quantization component, among which and This is the reference impedance value calibrated during component manufacturing or system initialization. and The weighting coefficients are preset, ensuring that the sum of the two is 1. The focus is on characterizing the increase in cross-resistance caused by solder strip corrosion. Focusing on characterizing the increase in leakage current caused by PID effect or microcracks, a dimensionless physical aging coefficient is generated by calculating the weighted sum of the rate of change of internal resistance and the rate of decay of leakage resistance. This coefficient directly reflects the degree of degradation at the hardware level of the component.

[0048] Subsequently, the system uses the formula Calculate the ambient shading coefficient, where Represents the total power loss rate. The aging-power coupling factor is a dimensionless factor used to map the physical aging coefficient to the corresponding theoretical power loss component. The physical meaning of this calculation logic is: the total power loss minus the loss caused by internal aging results in the environmental shading coefficient caused by recoverable factors such as external shading and dust accumulation. This allows for the precise separation of the fault source.

[0049] It should be added that the core logic of the environmental shading coefficient calculation formula lies in "fault source decoupling". Among other things, This characterizes the total power loss rate of the sub-module relative to its theoretical maximum output capacity. This loss is a result of the combined effects of "internal physical aging" and "external environmental shading." To isolate the loss caused solely by external factors (such as dust, bird droppings, and tree shade), the formula introduces... This item, among which The physical significance lies in the dimensionless physical aging coefficient. This is mapped to the corresponding power attenuation component. That is, the estimated loss due to aging is subtracted from the total loss, and the remainder is attributed to environmental shading. The function Max[0,·] acts as a physical constraint to prevent negative values ​​in the calculation results due to measurement errors, thereby ensuring... This only represents the percentage decrease in photon capture due to light being blocked.

[0050] Finally, the system calls a preset standard single-diode photovoltaic cell output model and dynamically corrects the model parameters based on the above calculation results: adjusting the photocurrent parameters in the model. Revised to This is to reflect the reduction in photon capture caused by external shading; at the same time, the series resistance in the model is... and parallel resistors Update directly to actual test results and Numerical values, substituting the corrected parameters into the equation. In this process, a corrected current-voltage characteristic curve that can accurately fit the actual output capability of the current sub-module under various load conditions is generated, and this curve is established as the aging characteristic model of the sub-module.

[0051] After separating environmental and physical factors, the system dynamically corrects the parameters of the standard single-diode photovoltaic cell output model pre-installed in the control chip:

[0052] The first step is to correct the photocurrent parameters. Since environmental shading directly reduces the number of photons reaching the semiconductor PN junction, the correction formula calculates the actual photocurrent generated by subtracting the environmental shading ratio. The rated photocurrent under standard test conditions.

[0053] The second step is to correct the impedance parameters. This involves adjusting the series resistance in the model. and parallel resistors The actual values ​​obtained from the nighttime testing steps are directly replaced to reflect the ohmic contact loss and leakage path conditions inside the component.

[0054] Finally, the modified parameters are substituted into the transcendental equation—the Shockley diode equation—to generate the modified IV characteristic curve.

[0055] Where I and V are the current and voltage output by the model, respectively, the complete IV curve can be plotted by solving the equation using the numerical iteration method. It is represented by the calculated environmentally corrected photocurrent, which characterizes the "energy source" of the component.

[0056] This is the reverse saturation current of the diode. This parameter is mainly affected by temperature and is determined by the system based on the currently measured surface temperature T according to the material physics formula. Calculated in real time. q represents the electron charge constant, with a value of... Coulombs. n is the diode ideality factor, typically ranging from 1.0 to 1.5, depending on the cell manufacturing process, and is stored as a system preset constant. k is the Boltzmann constant, with a value of... Joules / Kelvin. T represents the absolute temperature of the current sub-module, converted from surface temperature data collected during the day.

[0057] Traditional maximum power point tracking can only scan current external characteristics and cannot predict "what will happen if the connection method is changed". By building this modified model that includes aging parameters, the system can virtually simulate the performance of the sub-module under different current conditions without actually switching the circuit, thereby accurately calculating the "optimal matching current value".

[0058] Preferably, the optimal matching current value for each sub-module is calculated based on the aging characteristic model. The specific calculation logic is as follows:

[0059] For the aging characteristic model of each sub-module, the numerical iterative algorithm is used to traverse and calculate the modified current-voltage characteristic curve within the preset voltage scanning range. The product of the voltage value and the current value at each scanning point is solved to obtain the power curve. The global maximum power point where the amplitude reaches its peak in the power curve is identified, and the current coordinate value corresponding to the global maximum power point is extracted as the theoretical electrical optimal current.

[0060] The equivalent series internal resistance value corresponding to the sub-module is stored in the static impedance data at night, and the maximum allowable heat dissipation power threshold preset by the system is read. The maximum thermal safety current that can limit the Joule heating effect from causing the component to overheat is generated by calculating the ratio between the maximum allowable heat dissipation power threshold and the equivalent series internal resistance value and taking the square root of the ratio.

[0061] The numerical comparison logic is executed to compare the theoretical electrical optimal current with the maximum thermal safety current, select the minimum value between the two and determine the minimum value as the best matching current value of the sub-module, thereby defining the upper limit of the safe operating current of the sub-module under the dual constraints of maximizing photoelectric conversion output and preventing the risk of thermal runaway caused by internal resistance aging.

[0062] Specifically, the system loads the modified current-voltage characteristic curve equation and uses the variable step-size hill-climbing method at open-circuit voltage. to short circuit current Numerical scanning is performed across the entire voltage range, with a step size set to 0.1V to ensure accuracy, to calculate each discrete voltage point. Corresponding output power By comparing the power peaks in the sequence The global maximum power point is located, and the current value corresponding to that point is extracted as the theoretical electrical optimal current. This parameter represents the ideal operating current at which the component can contribute the maximum electrical energy, without considering thermal safety limitations.

[0063] However, for severely aged components, their series internal resistance... Significantly increased power consumption, if blindly pursuing maximum electrical power, will result in excessive operating current generating intense Joule heating across the internal resistance, leading to hot spot effects and even burning out the backplane. Therefore, this invention introduces a thermal safety constraint mechanism: the system retrieves the measured equivalent series internal resistance value of the sub-module stored in the nighttime static impedance database. Simultaneously, it reads the maximum allowable heat dissipation power threshold preset in the system firmware. The unit is watts. This threshold is usually derived by working backwards from the temperature resistance rating of the module's encapsulation material, and it is recommended to set it to 2%-5% of the module's rated power. For example, for a 250W module, the allowable power consumption limit based on internal resistance is 5W to 12.5W. This is based on the inverse calculation formula of Joule's law. Calculate the maximum thermal safety current that can limit the internal heating of the component to no more than the safety limit. ,in The maximum permissible heat dissipation power threshold is defined as the maximum internal heat generation power that the component encapsulation material can safely withstand.

[0064] Finally, the system executes a "weakest link" decision-making logic, using formulas. Select the optimal matching current value The physical meaning of this logic is: if the component is healthy ( (small), then Very large, system selection To maximize power generation; if the components are severely aged ( big), It will drop sharply and may fall below At this point, the system forces the selection of a lower value. Although it sacrifices some of the theoretical power generation as the working current, it fundamentally eliminates the fire hazard and maximizes safety benefits.

[0065] The central processing module also performs clustering and hierarchical calculations based on the optimal matching current value of each sub-module, classifying sub-modules with optimal matching current values ​​within a predetermined similarity range into the same health level and generating corresponding topology instructions.

[0066] Preferably, the central processing module also performs clustering and hierarchical calculations based on the optimal matching current values ​​of each sub-module, classifying sub-modules with optimal matching current values ​​within a predetermined similarity range into the same health level, and generating corresponding topology instructions, including:

[0067] The optimal matching current values ​​calculated from all sub-modules are collected to construct a one-dimensional feature dataset. The one-dimensional feature dataset is then iteratively divided based on the K-Means clustering algorithm. By minimizing the intra-cluster variance and maximizing the inter-cluster distance, all sub-modules are automatically classified into the first health level set, the second health level set, and the third health level set. The cluster center current value corresponding to the first health level set is the highest, indicating the best performance. The cluster center current value corresponding to the second health level set is in the middle, indicating sub-healthy performance. The cluster center current value corresponding to the third health level set is the lowest, indicating severely degraded performance.

[0068] Identify and extract all sub-module numbers belonging to the first health level set, and generate a first topology instruction that connects all sub-modules in the first health level set in sequence to construct a main powerful string capable of outputting high current.

[0069] Identify and extract all sub-module numbers belonging to the second health level set, calculate the total voltage prediction value after all sub-modules in the second health level set are connected in series, and generate a second topology instruction to independently connect these sub-modules in series when the total voltage prediction value meets the minimum start-up voltage requirement of the inverter, so as to construct a secondary weak string that operates in low current mode.

[0070] Identify and extract the numbers of all sub-modules belonging to the third health level set, and generate a third topology instruction to disconnect all sub-modules of the third health level set from the main circuit and connect them to the isolation port, so as to prevent the bottleneck effect from reducing the overall system performance.

[0071] It should be added that the identification and extraction of all sub-module numbers belonging to the first health level set are performed, and a first topology instruction is generated to sequentially concatenate all sub-modules in the first health level set to construct the main strong string, including:

[0072] Access the first health level set generated by clustering operation, index and extract the physical location identifier of each sub-module contained therein, and verify whether the total number of sub-modules in the first health level set is higher than the preset minimum number of components required for the inverter to work normally.

[0073] After the verification is passed, based on the distribution of the physical location identifiers in the circuit matrix, a series path is planned that can sequentially conduct all the sub-modules belonging to the first health level set by connecting the positive and negative terminals.

[0074] The serial path is converted into a relay closure address code that can be recognized by the matrix switch array, generating the first topology instruction, which instructs the actuator to construct a main powerful string consisting only of high-performance sub-modules.

[0075] It should be added that the system identifies and extracts the sub-module numbers belonging to the third health level set, and generates a third topology instruction to disconnect all sub-modules of the third health level set from the main circuit and connect them to the isolation port, including:

[0076] Access the third health level set generated by clustering operation, index and extract the physical location identifier of each sub-module contained therein, and confirm that these sub-modules are severely aging nodes with the best matching current value in the lowest range and are prone to causing system bottleneck effect.

[0077] Based on the mapping relationship of physical location identifiers in the circuit matrix, a bypass control strategy for matrix switch arrays is formulated, and a third topology instruction is generated. This instruction is used to drive the switching elements to disconnect the electrical connection between these sub-modules and the external power generation circuit, and switch their positive and negative pins to a preset electrical isolation port.

[0078] By physically separating these sub-modules from the topology of the main strong string and the secondary weak string, the risk of hot spots and power clamping caused by the forced series connection of high internal resistance components is eliminated, thereby ensuring the safety of system operation while realizing the net value retention of the output performance of the remaining healthy components.

[0079] Specifically, after completing the clustering and classification of sub-modules, the central processing module first executes the main strong string construction logic for the first health level set. The system accesses the first health level set, indexes and extracts the physical location identifier code of each sub-module contained therein, and obtains the total number of sub-modules in the first health level set. Subsequently, the system performs string size verification, calculated by subtracting a preset minimum component number threshold from the total number of sub-modules in the first health level set to obtain the size margin value. The preset minimum component number threshold is obtained by dividing the inverter's minimum starting voltage by the average rated voltage of a single healthy component and rounding up. Its value range is usually set to 6 to 12 components, aiming to ensure that the total string voltage can cross the inverter's DC conversion threshold. After the size margin value is greater than or equal to zero and passes the verification, the system plans a conduction path based on the spatial distribution relationship of the physical location identifier codes in the circuit matrix, cascading all sub-modules belonging to the first health level set in the electrical order of positive to negative. The conduction path is then converted into a relay closure address code that can be recognized by the matrix switch array, generating the first topology instruction, which instructs the actuator to construct a low-loss main high-efficiency string consisting only of high-performance sub-modules in the electrical circuit.

[0080] Simultaneously, for degraded sub-modules classified into the third health level set, the system synchronously initiates safety isolation and bypass control logic. The system indexes the physical location identifier codes of all sub-modules in the third health level set, confirming that the optimal matching current values ​​of these sub-modules are below the preset aging warning threshold. Based on the logical mapping relationship of the physical location identifier codes in the circuit matrix, the system formulates a bypass switching strategy for the matrix switch array and generates a third topology instruction. This instruction is used to drive specific switching elements to disconnect the electrical connection between these sub-modules and the external power generation circuit, and switch their positive and negative pins to a preset electrical isolation port. The preset electrical isolation port is usually connected to a discharge dissipation load to absorb any residual charge that may be generated when the sub-module is disconnected. Its resistance setting logic needs to refer to the maximum open-circuit voltage of the sub-module to ensure that the charge discharge time is controlled within a preset time window, preferably 2 to 5 seconds. By physically disconnecting these sub-modules from the main circuit, the risk of hot spots and power clamping caused by forced series connection of high internal resistance components is fundamentally eliminated, realizing the net retention of the output performance of the remaining healthy components.

[0081] To verify the feasibility of the above reconfiguration logic, a reconfiguration task involving an array with 20 sub-modules was taken as an example: the clustering and grading results showed that the first health level set contained 15 sub-modules, and the third health level set contained 2 degraded sub-modules. The system read the preset minimum component quantity threshold as 8, performed a subtraction operation to obtain a scale margin value of 7. Since 7 is greater than zero, it was determined that the conditions for constructing the main strong string were met. The system then extracted the identification codes of the 15 strong sub-modules and generated the corresponding relay closing address codes, and issued the first topology instruction to complete the cascading. At the same time, for the 2 degraded sub-modules in the third health level set, the system generated a third topology instruction to switch their pins to the isolation port. Through actual operation testing, due to the elimination of the current suppression effect of the 2 degraded components, the output current of the main strong string rose from the originally clamped 2 amps to the theoretical optimal value of 8.5 amps. The overall power generation of the system was significantly improved, and the real-time monitoring value of the temperature rise at the degraded components dropped to the ambient temperature level, proving the logical rigor and technical effectiveness of the topology instruction in improving efficiency and ensuring safety.

[0082] Preferably, the optimal matching current values ​​calculated from all sub-modules are aggregated to construct a one-dimensional feature dataset, and the one-dimensional feature dataset is iteratively partitioned based on the K-Means clustering algorithm, including:

[0083] Traverse the entire photovoltaic shading device, read the best matching current value of each sub-module in sequence, and arrange all the best matching current values ​​in the order of the physical location number of the sub-module to form a one-dimensional feature dataset.

[0084] Initialize three cluster centers, including selecting the maximum value in the one-dimensional feature dataset as the first initial center, the minimum value as the third initial center, and the median as the second initial center;

[0085] Perform iterative allocation operations. For each optimal matching current value, calculate its Euclidean distance to the first initial center, the second initial center, and the third initial center, and assign the optimal matching current value to the temporary cluster set represented by the nearest initial center to form three temporary cluster sets.

[0086] The execution center updates the operation by calculating the arithmetic mean of all values ​​in the three temporary cluster sets and replacing the old cluster centers with these arithmetic means as the new cluster centers.

[0087] Repeatedly perform iterative allocation and center update operations until the deviation between the new cluster center and the old cluster center is less than the preset convergence threshold. At this point, stop the iteration and label the three temporary cluster sets formed as the first health level set representing the healthy state, the second health level set representing the sub-healthy state, and the third health level set representing the deteriorated state, respectively.

[0088] Specifically, the system first traverses the entire photovoltaic shading device, sequentially reading the optimal matching current value for each sub-module. Where n is the sub-module index, n=1...N, and N is the total number of components, these values ​​are arranged in order of physical location number to construct a one-dimensional feature dataset. This dataset intuitively reflects the current carrying capacity distribution of all components at the current moment.

[0089] Next, to avoid the slow convergence speed or getting trapped in local optima caused by traditional random initialization, the system adopts a deterministic initialization strategy, selecting the maximum value in the dataset D. As the first initial center (Representing the theoretically strongest performance point), select the minimum value in dataset D. As the third initial center (Representing severe aging points), and calculate the median of the dataset. As the second initial center (Representing the sub-health transition point); then it enters the iterative allocation calculation stage, targeting each current value in the dataset. Using the Euclidean distance formula Calculate the distances between each node and its three cluster centers (k=1,2,3), where d_{n,k} is the distance metric in amperes. Let the value of the k-th cluster center be given in the t-th iteration. The system calculates the value based on the minimum distance principle. This current value Join the set of the nearest temporary cluster middle.

[0090] After completing one round of allocation for all personnel, the system execution center updates the calculations using the formula. The geometric center of each cluster is recalculated by taking the arithmetic mean of all current values ​​within that cluster as the new cluster center. This step eliminates the influence of individual noisy data points. The system continuously repeats the above allocation and update process, and calculates the change in the old and new cluster centers at the end of each iteration. ,when The iteration stops when the current falls below a preset convergence threshold, which is recommended to be set between 0.01 amperes and 0.05 amperes. This threshold is set based on the measurement accuracy of the current sensor and is intended to avoid invalid calculations of tiny jitters.

[0091] Based on the physical characteristics of photovoltaic modules, a higher optimal matching current value means a lower equivalent series resistance, lower internal power loss, and less impact from external shading, indicating a higher level of health. Therefore, the system automatically designates the temporary cluster set with the largest final cluster center value as the first health level set. The sub-modules within this set possess the highest operating current carrying capacity and can serve as core units for constructing the main high-efficiency string. Temporary cluster sets with intermediate final cluster center values ​​are automatically designated as the second health level set, representing sub-healthy modules with slightly degraded performance but still capable of generating electricity. The temporary cluster set with the smallest final cluster center value is automatically designated as the third health level set, representing degraded modules that have undergone irreversible severe aging or heavy damage and require stripping.

[0092] Preferably, the process involves identifying and extracting the sub-module numbers belonging to the second health level set, calculating the total voltage prediction value after connecting all sub-modules in the second health level set in series, and generating a second topology instruction to independently connect all sub-modules in the second health level set, provided that the total voltage prediction value meets the inverter's minimum start-up voltage requirement, to construct a secondary weakly effective string, including:

[0093] Access the second health level set generated by clustering operations and extract the physical location identifier code of each sub-module contained therein;

[0094] The aging characteristic model corresponding to each sub-module is called, and the operating voltage value corresponding to the best matching current value of the sub-module is located in the corrected current-voltage characteristic curve. The operating voltage values ​​of all sub-modules in the second health level set are accumulated to generate a predicted total voltage value that characterizes the output capability of the entire string.

[0095] The predicted total voltage value is compared with the preset minimum inverter start-up voltage threshold. If the predicted total voltage value is greater than the preset minimum inverter start-up voltage threshold, a second topology instruction is generated, which includes the logic of independently cascading these sub-modules, to instruct the matrix switch array to construct a secondary weak string operating at a low current matching point.

[0096] Specifically, after completing the clustering and grading based on the K-Means algorithm, the system performs a topology reconstruction feasibility assessment for the sub-healthy sub-modules classified into the second health level set. First, the system accesses the second health level set and extracts the physical location identifiers of all sub-modules contained within it. Simultaneously, it obtains the central current value of this set after clustering convergence as the unified benchmark matching current for this string. Next, the system calls the aging characteristic model of each corresponding sub-module, i.e., the modified Shockley diode equation, and uses the numerical iteration method to substitute the reference matching current into the equation to solve for the corresponding operating voltage value in reverse. The calculation formula is as follows: ,in Let be the predicted output voltage of the j-th sub-module under the reference matching current. and These are the measured series and parallel resistances of this sub-module. This is the corrected photocurrent.

[0097] Subsequently, the system performs an accumulation operation on the predicted output voltages of all M sub-modules within the second health level set to generate the total predicted voltage value. The calculation formula is as follows: ,in A preset cable loss correction factor is used, which characterizes the ohmic loss of the connecting wires within the string. Its value is recommended to be between 0.95 and 0.98, and in this embodiment, 0.97 is preferred to ensure the rigor of the voltage assessment. Next, the system reads the inverter's minimum start-up voltage threshold preset in the non-volatile memory. This parameter is determined by the step-up transformer ratio and the withstand voltage characteristics of the power semiconductor devices in the inverter hardware circuit, and is usually set to 40% to 60% of the inverter's rated input voltage.

[0098] Finally, the system executes threshold comparison logic. If the predicted total voltage value is greater than or equal to the inverter's minimum start-up voltage threshold, it determines that the set has independent grid-connected power generation capability and generates a second topology instruction containing a relay closing sequence mapping table based on the logical topology relationship of the physical location identifier code. If the predicted total voltage value is less than the inverter's minimum start-up voltage threshold, it generates a remedial instruction to merge the set into the end of the first health level set or to execute isolation skip.

[0099] The matrix topology execution module is configured to respond to the topology command, adjust the state of the switch array connected between each sub-module, construct the sub-modules divided into the same health level into a first series string, and connect the first series string to the inverter input terminal;

[0100] The matrix topology execution module responds to the topology command by first parsing the physical location identifier code of the target sub-module and the corresponding relay closing address code contained in the command, and mapping the address code to the hardware drive port of the switch array. Subsequently, the control logic unit sends a level conversion signal to the target drive port, adjusting the switch contact state connected between the positive and negative pins of each sub-module. This ensures that sub-modules classified as having the same health level are sequentially turned on in the electrical path according to the physical identifier code sequence, thus constructing a first series string with uniform current carrying capacity. Finally, after detecting that the output voltage at the end of the first series string reaches a preset stability threshold, the module closes the main switch located between the end of the string and the busbar, connecting the first series string to the inverter input, completing the stable delivery of electrical energy from the reconfigured string to the energy conversion unit. The preset stability threshold is logically set to be more than 95% of the sum of the predicted voltage values ​​retrieved from all sub-modules in the first series string in the preceding steps, ensuring that there is no excessive voltage drop due to poor contact during circuit switching.

[0101] Example 2:

[0102] See Figure 3 As shown, a solar photovoltaic shading device efficiency intelligent monitoring device is characterized by including a processor, a memory and a communication bus.

[0103] The memory stores a computer-readable program that can be executed by the processor;

[0104] The communication bus enables communication between the processor and the memory;

[0105] When the processor executes the computer-readable program, it performs a module to implement a solar photovoltaic shading device efficiency intelligent monitoring system as described in any one of the present invention.

[0106] It should be noted that the interval and threshold sizes are set for ease of comparison. The size of the threshold depends on the amount of sample data and the base number set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless calculations, and the formulas are derived from software simulations using a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0107] It should be understood that, in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0108] It should be understood that determining B based on A does not mean determining B solely based on A; it also means determining B based on A and / or other information.

[0109] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0110] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart monitoring system for the efficiency of a solar photovoltaic shading device, characterized in that, The system includes: The data acquisition module is used to acquire daytime operating thermoelectric data and nighttime static impedance data of each sub-module in the photovoltaic shading device; The central processing module is used to construct aging characteristic models of each sub-module based on daytime thermoelectric data and nighttime static impedance data, and to calculate the optimal matching current value of each sub-module based on the aging characteristic models. The central processing module also performs clustering and hierarchical operations based on the optimal matching current value of each sub-module, classifying sub-modules with optimal matching current values ​​within a predetermined similar range into the same health level and generating corresponding topology instructions. The matrix topology execution module is configured to respond to the topology command, adjust the state of the switch array connected between each sub-module, construct the sub-modules divided into the same health level into a first series string, and connect the first series string to the inverter input terminal.

2. The intelligent monitoring system for the efficiency of a solar photovoltaic shading device according to claim 1, characterized in that, The daytime operating thermoelectric data includes the temperature rise deviation index and the power efficiency index; the nighttime static impedance data includes the equivalent series internal resistance value and the leakage current parallel internal resistance value.

3. The intelligent monitoring system for the efficiency of a solar photovoltaic shading device according to claim 2, characterized in that, Construct aging characteristic models for each sub-module, including: Extract the temperature rise deviation index and power efficiency index from the daytime operation thermoelectric data, and align the temperature rise deviation index with the equivalent series internal resistance value and leakage current parallel internal resistance value contained in the nighttime static impedance data in the spatiotemporal dimension to construct a multidimensional health feature vector. The multidimensional health feature vector is input into the preset attenuation judgment logic unit. When the temperature rise deviation index is higher than the preset normal threshold, the physical aging coefficient is generated by calculating the weighted sum of the rate of change of the equivalent series internal resistance value and the reciprocal rate of change of the leakage current parallel internal resistance value. The environmental shielding coefficient is calculated by using the numerical difference between the power performance index and the physical aging coefficient, thereby separating the recoverable performance loss component. Based on the physical aging coefficient, the series resistance parameters and parallel resistance parameters in the preset standard photovoltaic cell output model are corrected. At the same time, based on the environmental shading coefficient, the photocurrent parameters in the standard photovoltaic cell output model are corrected. The corrected current-voltage characteristic curve of the actual output capability of the current sub-module is obtained by fitting, and the corrected current-voltage characteristic curve is established as the aging characteristic model of the sub-module.

4. The intelligent monitoring system for the efficiency of a solar photovoltaic shading device according to claim 3, characterized in that, The optimal matching current value for each sub-module is calculated based on the aging characteristic model. The specific calculation logic is as follows: For the aging characteristic model of each sub-module, the numerical iterative algorithm is used to traverse and calculate the modified current-voltage characteristic curve within the preset voltage scanning range. The product of the voltage value and the current value at each scanning point is solved to obtain the power curve. The global maximum power point where the amplitude reaches its peak in the power curve is identified, and the current coordinate value corresponding to the global maximum power point is extracted as the theoretical electrical optimal current. The equivalent series internal resistance value corresponding to the sub-module is retrieved from the static impedance data stored at night, and the maximum allowable heat dissipation power threshold preset by the system is read. The maximum thermal safety current is generated by calculating the ratio between the maximum allowable heat dissipation power threshold and the equivalent series internal resistance value and taking the square root of the ratio. The numerical comparison logic is executed to compare the theoretical electrical optimal current with the maximum thermal safety current, select the minimum value between the two, and determine the minimum value as the best matching current value for the sub-module.

5. The intelligent monitoring system for the efficiency of a solar photovoltaic shading device according to claim 1, characterized in that, The central processing module also includes: The optimal matching current values ​​calculated from all sub-modules are collected to construct a one-dimensional feature dataset. The one-dimensional feature dataset is then iteratively divided based on the K-Means clustering algorithm. By minimizing the intra-cluster variance and maximizing the inter-cluster distance, all sub-modules are automatically classified into the first health level set, the second health level set, and the third health level set. Identify and extract the numbers of all sub-modules belonging to the first health level set, and generate a first topology instruction that sequentially connects all sub-modules in the first health level set to construct the main strong string; Identify and extract all sub-module numbers belonging to the second health level set, calculate the total voltage prediction value after all sub-modules in the second health level set are connected in series, and generate a second topology instruction to independently connect all sub-modules in the second health level set to construct a secondary weak-effect string when the total voltage prediction value meets the minimum start-up voltage requirement of the inverter. Identify and extract the numbers of all sub-modules belonging to the third health level set, and generate a third topology instruction to disconnect all sub-modules of the third health level set from the main circuit and connect them to the isolation port.

6. The intelligent monitoring system for the efficiency of a solar photovoltaic shading device according to claim 5, characterized in that, The optimal matching current values ​​calculated from all sub-modules are aggregated to construct a one-dimensional feature dataset, which is then iteratively partitioned using the K-Means clustering algorithm, including: Traverse the entire photovoltaic shading device, read the best matching current value of each sub-module in sequence, and arrange all the best matching current values ​​in the order of the physical location number of the sub-module to form a one-dimensional feature dataset. Initialize three cluster centers by selecting the maximum value in the one-dimensional feature dataset as the first initial center, the minimum value as the third initial center, and the median as the second initial center. Perform iterative allocation operations. For each optimal matching current value, calculate its Euclidean distance to the first initial center, the second initial center, and the third initial center, and assign the optimal matching current value to the temporary cluster set represented by the nearest initial center to form three temporary cluster sets. The execution center updates the operation by calculating the arithmetic mean of all values ​​in the three temporary cluster sets and replacing the old cluster centers with these arithmetic means as the new cluster centers. Repeatedly perform iterative allocation and center update operations until the deviation between the new cluster center and the old cluster center is less than the preset convergence threshold. At this point, stop the iteration and label the three temporary cluster sets formed as the first health level set, the second health level set, and the third health level set, respectively.

7. The intelligent monitoring system for the efficiency of a solar photovoltaic shading device according to claim 5, characterized in that, The generation logic for the second topology instruction includes: Access the second health level set generated by clustering operations and extract the physical location identifier code of each sub-module contained therein; The aging characteristic model corresponding to each sub-module is called, and the operating voltage value corresponding to the best matching current value of the sub-module is located in the corrected current-voltage characteristic curve. The operating voltage values ​​of all sub-modules in the second health level set are accumulated to generate a predicted total voltage value that characterizes the output capability of the entire string. The predicted total voltage value is compared with the preset minimum inverter start-up voltage threshold. If the predicted total voltage value is greater than the preset minimum inverter start-up voltage threshold, a second topology instruction is generated.

8. A smart monitoring device for the efficiency of a solar photovoltaic shading device, characterized in that, Includes processor, memory, and communication bus; The memory stores a computer-readable program that can be executed by the processor; The communication bus enables communication between the processor and the memory; When the processor executes the computer-readable program, it performs a module to implement a solar photovoltaic shading device efficiency intelligent monitoring system as described in any one of claims 1 to 7.

Images