Polycrystalline silicon solar panel intelligent fault detection early warning system and method

Abstract

The invention discloses a polycrystalline silicon solar panel intelligent fault detection early warning system and method, and relates to the technical field of intelligent fault detection. The polycrystalline silicon solar panel intelligent fault detection early warning system comprises a power attenuation fault detection module; a hot spot effect fault detection module; and a short circuit fault detection module. According to the invention, power attenuation early warning timeliness analysis is carried out, whether sine wave signal extraction optimization determination and early warning are carried out is determined, then hot spot effect early warning timeliness analysis is carried out, whether hot spot effect optimization determination and early warning are carried out is determined, and finally short circuit early warning timeliness analysis is carried out. According to the method, whether temperature-waveform extraction optimization judgment and early warning are carried out or not is judged, accurate and timely early warning of faults such as power attenuation, hot spot effect and short circuit is achieved, and the problem that in the prior art, fault detection and early warning of the polycrystalline silicon solar panel are not timely due to interference in the fault feature extraction process is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of intelligent fault detection technology, and in particular to an intelligent fault detection and early warning system and method for polycrystalline silicon solar panels. Background Technology

[0002] As a core component of photovoltaic power generation, the operational stability of polycrystalline silicon solar panels is crucial to power generation efficiency and system lifespan. To address the problems of low efficiency and difficulty in fault location associated with traditional manual inspections, an intelligent fault detection and early warning system has been introduced into polycrystalline silicon solar panels. This system relies on a variety of advanced technologies to provide comprehensive protection for polycrystalline silicon solar panels. During the intelligent fault detection process, sensors are installed at key nodes of the solar panel (such as cells, junction boxes, and inverter interfaces) to collect parameters such as voltage, current, temperature, and light intensity in real time. Signal processing algorithms transform the raw data into quantifiable fault characteristics, thereby quickly locating the fault. Finally, based on the data processing and analysis results, the system can issue timely fault warnings. When the parameters of the solar panel deviate from the normal range or potential fault characteristics are detected, the early warning layer will quickly issue corresponding warning signals.

[0003] For example, the Chinese invention patent with announcement number CN115940809B discloses a method for detecting solar panel faults based on power data and visual analysis, which includes: acquiring real-time operating power data of the solar panel; obtaining a power anomaly feature set; acquiring a real-time visual image set of the solar panel; performing fusion analysis and diagnosis based on a fault fusion diagnosis model to obtain the fault diagnosis result of the solar panel.

[0004] For example, the invention disclosed in Chinese Invention Patent No. CN110224673B is a method for fault detection of solar photovoltaic panels based on deep learning. The method includes: cleaning the collected time-series current data of photovoltaic panels, extracting horizontal and vertical features from the processed current data, generating a detection dataset by combining the extracted features with the original current values ​​of the panels, then using a convolutional neural network to obtain deep features of the current data, and then using a long short-term memory model to learn the features in the time dimension of the training data, ultimately completing the fault detection of the photovoltaic module.

[0005] In existing technologies, during intelligent fault detection of polycrystalline silicon solar panels, the operation of power transmission and transformation equipment and inverters around the photovoltaic power generation system generates strong electromagnetic radiation. Poor coupling and contact of cable wiring inside the photovoltaic array can also cause conducted interference. The superposition of electromagnetic interference is a key factor leading to a decrease in the signal-to-noise ratio of current and voltage signals. Therefore, sinusoidal signal extraction is performed on polycrystalline silicon solar panels to detect fault features that are difficult to accurately capture due to noise. The sinusoidal extraction results can achieve early identification of hot spot effects and accurately identify hot spot areas through temperature data. The high temperature in the hot spot area will accelerate the aging and degradation of the internal insulation material of the module, causing internal short circuits. However, the electrical signal distortion caused by hot spots (such as excessive harmonics) partially overlaps with the signal characteristics of short circuit faults, which will mask the real short circuit signal, making it difficult to distinguish fault features. Ultimately, this results in a lag in fault detection and early warning, and there is a problem of untimely fault detection and early warning of polycrystalline silicon solar panels due to interference in the fault feature extraction process. Summary of the Invention

[0006] To address the technical problem of untimely fault detection and early warning in polycrystalline silicon solar panels due to interference during the fault feature extraction process, this invention provides an intelligent fault detection and early warning system and method for polycrystalline silicon solar panels. The technical solution is as follows:

[0007] On one hand, a smart fault detection and early warning system for polycrystalline silicon solar panels is provided. This system includes: a power attenuation fault detection module, a hot spot effect fault detection module, and a short circuit fault detection module. The power attenuation fault detection module monitors the operating status of the polycrystalline silicon solar panel in real time, performs timeliness analysis of the power attenuation early warning during the conversion control process of the sine wave signal, and simultaneously performs sine wave signal extraction optimization judgment and early warning. Sine wave signal extraction optimization means reducing harmonic distortion during signal conversion by adjusting the carrier frequency and dead time of the SPWM. The hot spot effect fault detection module, after the sine wave signal extraction is qualified, uses the data obtained from the corresponding hot spot area of ​​the photovoltaic module to detect the power attenuation fault. The hot spot edge temperature gradient ratio is used to analyze the timeliness of hot spot effect early warning in the temperature dissipation process of photovoltaic modules. At the same time, it performs hot spot effect optimization judgment and early warning. Hot spot effect optimization means adjusting the sampling frequency and filter window size of the infrared thermal imager to enhance the acquisition accuracy and anti-interference capability of temperature data in the hot spot area. The short circuit fault detection module is used to analyze the timeliness of short circuit early warning in the temperature-waveform extraction process on the surface of photovoltaic modules based on the results of hot spot effect optimization judgment and early warning. It determines whether to perform temperature-waveform extraction optimization judgment and early warning. Temperature-waveform extraction optimization means adjusting the passband bandwidth of the bandpass filter to accurately filter out noise signals unrelated to short circuit faults.

[0008] On the other hand, a smart fault detection and early warning method for polycrystalline silicon solar panels is provided. The method includes: S1, real-time monitoring of the operating status of the polycrystalline silicon solar panel, performing power attenuation early warning timeliness analysis on the conversion control process of the sine wave signal, and simultaneously performing sine wave signal extraction optimization judgment and early warning; S2, after the sine wave signal extraction is qualified, based on the hot spot edge temperature gradient ratio obtained in the corresponding hot spot region of the photovoltaic module, performing hot spot effect early warning timeliness analysis on the temperature heat dissipation process of the photovoltaic module, and simultaneously performing hot spot effect optimization judgment and early warning; S3, based on the results of hot spot effect optimization judgment and early warning, performing short-circuit early warning timeliness analysis on the temperature-waveform extraction process on the surface of the photovoltaic module to determine whether to perform temperature-waveform extraction optimization judgment and early warning.

[0009] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0010] 1. In fault detection of polycrystalline silicon solar panels, the fault feature extraction process is often interfered with, leading to untimely early warnings. To address this issue, three modules are optimized collaboratively: For power attenuation fault detection, the early warning timeliness analysis of the sine wave signal conversion control process is performed, optimizing signal extraction and early warning, providing high-purity, high-stability electrical signals to avoid early warning delays caused by signal distortion; for hot spot effect fault detection, the early warning timeliness is analyzed based on the ratio of the temperature gradient between the hot spot area and its edge, optimizing judgment and early warning, enhancing the accuracy of temperature data acquisition and anti-interference capabilities, and clearly capturing temperature differences even in complex environments; for short-circuit fault detection, the surface temperature-waveform extraction process of the photovoltaic module is analyzed and optimized to highlight characteristic signals before the short circuit, improving identification sensitivity and ensuring protection is triggered before danger. The synergy of these three modules improves the timeliness of fault detection and early warning, ensuring stable application.

[0011] 2. Implement tiered early warning by monitoring interference indicators: When the indicator is within the set allowable range, it is determined to be a recoverable anomaly, triggering the first early warning and automatically optimizing the signal extraction status to avoid signal anomalies affecting the accuracy of power attenuation fault detection; when the indicator is less than the historical minimum value, it is determined to be fault-free, directly transitioning to hot spot effect detection, reducing invalid early warnings and resource consumption, and improving detection consistency and efficiency; when the indicator is greater than the historical maximum value, it is determined to be an unadjustable anomaly, triggering the second early warning and prompting manual intervention to address complex problems that the system cannot resolve automatically. This tiered mechanism ensures that the system automatically handles recoverable faults while also introducing timely manual intervention in cases of severe anomalies, comprehensively and accurately monitoring the signal extraction status, and mitigating the problem of untimely early warnings caused by interference.

[0012] 3. By comparing extracted interference indicators with three threshold categories in the database—interference allowable range, preset interference indicators, and historical maximum values—the system accurately determines three states: recoverable anomalies, no faults, and unadjustable faults. This avoids ambiguous judgments and provides data support for timely handling. Simultaneously, a two-tiered early warning system is employed to differentiate the handling of anomalies of varying degrees: minor anomalies are intervened promptly to prevent interference from escalating and affecting the early warning; major faults are responded to rapidly to avoid delays in handling. This mechanism, through refined evaluation and tiered response, effectively reduces early warning lag caused by interference during the extraction process, improving the timeliness of fault detection and early warning. Attached Figure Description

[0013] 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.

[0014] Figure 1 This is a schematic diagram of the structure of the intelligent fault detection and early warning system for polycrystalline silicon solar panels provided in an embodiment of the present invention;

[0015] Figure 2 A flowchart for optimizing sine wave signal extraction provided in an embodiment of the present invention;

[0016] Figure 3 A flowchart for optimizing the hot spot effect provided in an embodiment of the present invention;

[0017] Figure 4 A flowchart for temperature-waveform extraction optimization provided in an embodiment of the present invention;

[0018] Figure 5 The flowchart illustrates the intelligent fault detection and early warning method for polycrystalline silicon solar panels provided in this embodiment of the invention. Detailed Implementation

[0019] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0020] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0021] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent.

[0022] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0023] like Figure 1 The diagram shown is a structural schematic of the intelligent fault detection and early warning system for polycrystalline silicon solar panels provided in an embodiment of the present invention. The system includes: a power attenuation fault detection module, a hot spot effect fault detection module, and a short circuit fault detection module.

[0024] The power attenuation fault detection module is used to monitor the operating status of polycrystalline silicon solar panels in real time. It performs power attenuation early warning timeliness analysis on the conversion control process of sine wave signals (achieving DC to AC conversion through SPWM technology and ensuring that the output waveform is a pure sine wave). At the same time, it performs sine wave signal extraction optimization judgment and early warning. Sine wave signal extraction optimization means adjusting the carrier frequency and dead time of SPWM to reduce harmonic distortion in the signal conversion process, improve the purity and stability of the sine wave signal, ensure the accuracy of power monitoring data, and provide a reliable basis for accurate early warning of power attenuation faults.

[0025] The hot spot effect fault detection module is used to analyze the timeliness of hot spot effect early warning based on the ratio of the temperature gradient at the edge of the hot spot in the corresponding hot spot region of the (polycrystalline silicon solar panel) photovoltaic module after the sinusoidal signal is extracted and qualified. At the same time, it performs hot spot effect optimization judgment and early warning. Hot spot effect optimization means that by adjusting the sampling frequency and filter window size of the infrared thermal imager, the acquisition accuracy and anti-interference ability of the temperature data in the hot spot region are enhanced, the temperature change characteristics at the edge of the hot spot are captured more clearly, and the timeliness and accuracy of hot spot effect fault early warning are improved.

[0026] The short-circuit fault detection module is used to perform short-circuit warning timeliness analysis on the temperature-waveform extraction process of the photovoltaic module surface based on the results of hot spot effect optimization judgment and warning. This is to determine whether to perform temperature-waveform extraction optimization judgment and warning. Temperature-waveform extraction optimization means adjusting the passband bandwidth of the bandpass filter to accurately filter out noise signals that are not related to short-circuit faults, highlight the abnormal characteristics of temperature and current waveforms, improve the sensitivity of short-circuit fault identification, and ensure the timeliness and reliability of short-circuit warning.

[0027] In this embodiment, the power attenuation fault detection module is activated first. Utilizing Sinusoidal Pulse Width Modulation (SPWM) technology, it achieves DC-to-AC conversion, monitors the operating status in real time, and analyzes the timeliness of power attenuation early warnings. By adjusting the carrier frequency and dead time, it optimizes the extraction of the sine wave signal, reducing harmonic distortion and providing accurate data support for power attenuation fault early warnings. Once the sine wave signal extraction is satisfactory, the hot spot effect fault detection module takes over. Based on the edge temperature gradient ratio of the hot spot region of the photovoltaic module, it analyzes the timeliness of hot spot effect early warnings. Simultaneously, it adjusts the sampling frequency and filter window size of the infrared thermal imager to enhance the accuracy of temperature data acquisition, ensuring the accuracy of hot spot effect fault early warnings. Finally, the short-circuit fault detection module analyzes the timeliness of short-circuit early warnings based on the hot spot effect optimization results. By adjusting the passband bandwidth of the bandpass filter to filter noise, it highlights the abnormal correlation between temperature and current waveforms, improving the sensitivity of short-circuit fault identification. These three modules are progressively integrated and interconnected, comprehensively covering the main fault types and effectively solving the problem of untimely early warnings caused by interference in fault feature extraction.

[0028] Further, a timeliness analysis of power attenuation warning is conducted. The specific steps are as follows: During the conversion control process of the sine wave signal, based on the obtained useful signal power ratio and interference signal frequency ratio, the difference is compared with the useful signal power ratio and interference signal frequency ratio set in the database to obtain the useful signal power ratio score and interference signal frequency ratio score; the result of the inverse proportional processing of the obtained useful signal power ratio score is summed and averaged with the interference signal frequency ratio score to obtain the interference index extracted from the sine wave signal. The interference index extracted from the sine wave signal represents the quantitative data of the influence of the sine wave signal data on the timeliness of power attenuation warning. The sine wave signal data includes the useful signal power ratio and interference signal frequency ratio.

[0029] Among them, the useful signal power ratio score represents the ratio of the acquired useful signal power ratio to the set useful signal power ratio; the interference signal frequency ratio score represents the ratio of the absolute value of the difference between the acquired interference signal frequency ratio and 1 to the set interference signal frequency ratio; the useful signal power ratio represents the ratio of the power of the sinusoidal signal component of the solar panel power attenuation state in the sinusoidal signal (analyzed by Fourier transform to identify specific frequency components related to attenuation characteristics) to the total power of the total interference signal (including electromagnetic noise, interference signals generated by ambient light / temperature fluctuations, and fluctuation signals caused by changes in conductor contact resistance, etc.). The higher this ratio, the greater the proportion of useful power attenuation characteristics in the sinusoidal signal, the less affected by interference, and the stronger the timeliness of power attenuation warning. Conversely, it indicates that the interference signal overwhelms the useful characteristics, and the warning is prone to lag; the interference signal frequency ratio represents the ratio of the frequency of the high-frequency electromagnetic noise signal generated by the inverter operation to the inherent frequency of the power attenuation characteristic signal.

[0030] In this embodiment, the useful signal power ratio fraction is first inversely proportionalized, and then summed and averaged with the interference signal frequency ratio fraction. This is to reasonably balance the weighting of the two factors on the timeliness of power attenuation warnings. Useful signal power is positively correlated with warning accuracy; inverse proportionalization maps it to a quantified value in the same direction as warning timeliness. This process avoids evaluation bias caused by differences in the dimensions or dynamic ranges of the two factors, and ensures that the positive contribution of the useful signal is accurately quantified through inverse proportionalization. Finally, the summation and averaging are integrated into a comprehensive index, making the analysis results more consistent with the impact logic of actual warning scenarios and improving the accuracy of timeliness assessment.

[0031] The set useful signal power ratio is represented by the sum and average of historical useful signal power ratios during the conversion control process of sinusoidal signals in the database. Similarly, the set interference signal frequency ratio is represented by the sum and average of historical interference signal frequency ratios during the conversion control process of sinusoidal signals in the database. This example quantifies the proportion of useful features to interference by using the useful signal power ratio, which can accurately determine whether power attenuation features are overwhelmed by noise. The introduction of the interference signal frequency ratio reveals the degree of overlap between interference and feature signals at the frequency domain level, which helps to more effectively identify and address solar panel power attenuation problems in complex signal environments.

[0032] In this embodiment, the useful signal power ratio and the interference signal frequency ratio are two core dimensions for characterizing signal interference, and the interference index extracted from the sine wave signal is a comprehensive quantification of the two. The useful signal power ratio reflects the strength relationship between the useful feature (power attenuation signal) and the interference from the perspective of "power proportion"; the interference signal frequency ratio measures the interference from the perspective of "frequency domain overlap". The closer the ratio is to 1, the more serious the frequency overlap between the interference and the feature signal, and the stronger the interference; the farther it deviates from 1, the higher the frequency separation and the weaker the interference.

[0033] The interference index extracted from the sine wave signal integrates the influence of "power ratio" and "frequency domain overlap (based on 1)" by processing the former two (such as the inverse proportional conversion of the useful signal power ratio fraction and the summation and averaging of the interference signal frequency ratio fraction), and finally forms a unified index that intuitively reflects the comprehensive effect of the overall interference on power attenuation early warning.

[0034] The useful signal power ratio reflects the proportion of useful features (related to power attenuation) in the total signal and is the basis for the accuracy of early warning. The interference signal frequency ratio reflects the frequency overlap between interference and feature signals and determines the degree of confusion caused by interference in feature identification. The interference index is extracted by the sinusoidal signal obtained by summing and averaging the two after processing. It is the quantitative result of the combined effect of the two, integrating the dual influence of power ratio and frequency overlap. It comprehensively describes the total effect of signal interference on early warning timeliness from the two dimensions of amplitude and frequency, forming the correlation logic of "sub-item characterization - comprehensive quantification".

[0035] like Figure 2 The diagram shows a flowchart of the sine wave signal extraction optimization provided in an embodiment of the present invention. The specific design logic is as follows: After obtaining the sine wave signal interference index, a judgment is made according to the database range. If it is within the allowable range, it is determined to be recoverable and a first sine wave signal extraction status warning is issued. If it is less than the historical minimum value, there is no sine wave signal extraction status fault and hot spot effect detection is performed. If it is greater than the historical maximum value, it is determined to be an abnormal fault within the unadjustable range and a second sine wave signal extraction status warning is issued to prompt the preset personnel to take corresponding secondary warning measures, so as to realize the sine wave signal extraction optimization judgment and warning.

[0036] It is further important to understand that the optimization judgment and early warning of sine wave signal extraction specifically includes: sine wave signal extraction status judgment and sine wave signal extraction status early warning (including first sine wave signal extraction status early warning and second sine wave signal extraction status early warning).

[0037] The specific steps for determining the sine wave signal extraction status include: if the acquired sine wave signal extraction interference index is within the allowable range of sine wave signal extraction interference in the database, then the abnormal fault in the sine wave signal extraction status is determined to be within the recoverable range, and a first sine wave signal extraction status warning is issued. The allowable range of sine wave signal extraction interference represents the closed interval corresponding to the maximum and minimum values ​​of historical sine wave signal extraction interference during the conversion control process of historical sine wave signals in the database; if the acquired sine wave signal extraction interference index is less than the minimum value of historical sine wave signal extraction interference index in the database, then a fault of no sine wave signal extraction status is determined, and hot spot effect detection is performed. The minimum value of historical sine wave signal extraction interference index is verified through the database. The result is represented by the sum of the minimum values ​​of the historical sinusoidal signal extraction interference indexes during the conversion control process of each historical sinusoidal signal. If the obtained sinusoidal signal extraction interference index is greater than the maximum value of the historical sinusoidal signal extraction interference indexes in the database, it is determined that the abnormal fault of the sinusoidal signal extraction state is within the unadjustable range (i.e., the sinusoidal signal extraction state is unqualified), and a second sinusoidal signal extraction state warning is issued to prompt the preset personnel to take corresponding secondary warning measures (such as connecting an active filter to suppress high-frequency / low-frequency strong interference). The maximum value of the historical sinusoidal signal extraction interference index is represented by the sum of the maximum values ​​of the historical sinusoidal signal extraction interference indexes during the conversion control process of each historical sinusoidal signal in the database.

[0038] In this embodiment, the sine wave signal extraction status is divided into three cases. This fine distinction ensures that the system can take adaptive measures for interference of different magnitudes. When the interference index triggers the first sine wave signal extraction status warning, it provides guidance for timely recovery measures. When the interference index is less than the historical minimum value, unnecessary warnings are avoided, improving system operating efficiency. When the interference index is greater than the historical maximum value, the second sine wave signal extraction status warning is issued in a timely manner, ensuring stable system operation. This achieves comprehensive and accurate monitoring and warning of the sine wave signal extraction status, providing a strong guarantee for the reliable operation of the system.

[0039] Furthermore, a first sine wave signal extraction status warning is performed. The specific steps are as follows: The acquired sine wave signal extraction interference index deviation is input into the database. Based on the mapping relationship between the sine wave signal extraction interference index deviation and the carrier frequency adjustment amount in the database, the carrier frequency adjustment amount is obtained. Simultaneously, the acquired sine wave signal extraction interference index deviation is input into the database. Based on the mapping relationship between the sine wave signal extraction interference index deviation and the dead time adjustment amount in the database, the dead time adjustment amount is obtained. The sine wave signal extraction interference index deviation represents the difference between the sine wave signal extraction interference index during the sine wave signal conversion control process and the set sine wave signal extraction interference index. The proportion of high-frequency noise signal and the inverter switching delay before the first sine wave signal extraction status warning are obtained. The proportion of high-frequency noise signal reflects the degree of high-frequency interference in the sine wave signal; a higher proportion indicates more significant interference from high-frequency noise on signal extraction. The inverter switching delay reflects the response lag time of the switching devices; a larger delay means a larger time difference in switching action.

[0040] The acquired high-frequency noise signal proportion deviation is input into the database. Based on the mapping relationship between the high-frequency noise signal proportion deviation and the carrier frequency increase in the database, the carrier frequency increase is obtained. By increasing the carrier frequency of SPWM, the difference between it and the inherent frequency of the power attenuation characteristic signal is widened, thereby reducing the interference caused by frequency overlap. The high-frequency noise signal proportion deviation represents the difference between the proportion of high-frequency noise signal in the sine wave signal conversion control process and the set high-frequency noise signal proportion. The set high-frequency noise signal proportion deviation is represented by the sum and average of the historical high-frequency noise signal proportion deviations in the sine wave signal conversion control process in the database.

[0041] The acquired inverter switching delay deviation is input into the database. Based on the mapping relationship between the inverter switching delay deviation and the dead time reduction in the database, the dead time reduction is obtained. By reducing the dead time of SPWM, the glitches caused by switching action are reduced, the signal-to-noise ratio of the sine wave signal is improved, and a more stable signal source is provided for power attenuation warning. The inverter switching delay deviation represents the difference between the inverter switching delay during the sine wave signal conversion control process and the set inverter switching delay. The set inverter switching delay deviation is represented by the sum and average of the historical inverter switching delay deviations during the sine wave signal conversion control process in the database.

[0042] After adjusting the carrier frequency and dead time, the power attenuation warning timeliness analysis is performed again. If the interference index of the newly acquired sine wave signal is less than the minimum value of the historical sine wave signal interference index in the database, it is determined that the fault recovery of the no sine wave signal extraction state is completed, and hot spot effect detection is performed. Otherwise, the second sine wave signal extraction state warning is performed.

[0043] In this embodiment, the proportion of high-frequency noise signal is determined using an adaptive noise cancellation algorithm. High-frequency interference is separated by constructing a noise reference signal, and the proportion is calculated by energy integration. The inverter switching delay can be obtained by pulse trigger-response time difference analysis. A preset trigger pulse is input to the inverter switching device, and the actual turn-on / turn-off response signal of the switching device is collected using a high-speed oscilloscope or photoelectric sensor. The switching delay is obtained by comparing the difference between the trigger time and the response time using timestamps.

[0044] It is important to understand that when the proportion of high-frequency noise is reduced by increasing the carrier frequency, the increased switching frequency may increase switching losses, indirectly affecting the response speed of the switching devices and leading to increased switching delay. On the other hand, when reducing the dead time to reduce switching delay, it is necessary to avoid the risk of bridge arm shoot-through caused by insufficient dead time. In this case, it is necessary to pay attention to the changes in the proportion of high-frequency noise and ensure the coordinated optimization of the two. The proportion of high-frequency noise reflects the level of high-frequency interference in the signal, while the switching delay reflects the degree of dynamic interference of the switching action on the signal. The two together affect the signal-to-noise ratio and provide complementary basis for signal preprocessing for power attenuation warning.

[0045] This embodiment establishes a database mapping relationship between interference index deviation and carrier frequency and dead time adjustment, and adopts differentiated strategies to increase carrier frequency and reduce dead time for high-frequency noise signal proportion deviation and inverter switching delay deviation, respectively. It achieves precise control based on quantization parameters, avoids the subjectivity and blindness of traditional adjustment methods, can accurately suppress the impact of high-frequency interference and switching glitches on signal extraction, effectively improves the signal-to-noise ratio of sine wave signals, and provides a more stable and reliable signal source for subsequent power attenuation early warning.

[0046] like Figure 3 The flowchart shown is a hot spot effect optimization flowchart provided by an embodiment of the present invention. The specific design logic is as follows: after obtaining the temperature gradient ratio at the edge of the hot spot, the database determines whether the abnormal fault is within a preset range and a primary hot spot effect state warning is issued if it is within the preset range. If it is less than the historical minimum value, it is determined that there is no hot spot effect state fault and a short circuit fault is detected. If it is greater than the historical maximum value, the abnormal fault has exceeded the adjustable range and an advanced hot spot effect state warning is issued, prompting preset personnel to take corresponding advanced warning measures to achieve hot spot effect optimization judgment and warning.

[0047] It is further important to understand that hot spot effect optimization judgment and early warning specifically includes: hot spot effect state judgment and hot spot effect state early warning. Hot spot effect state early warning includes primary hot spot effect state early warning and advanced hot spot effect state early warning.

[0048] The hot spot effect status determination process includes the following steps: If the obtained hot spot edge temperature gradient ratio is within the preset range of hot spot edge temperature gradient ratios in the database, it indicates that the abnormal fault of the hot spot effect status is within the recoverable range, and a primary hot spot effect status warning is issued. The hot spot edge temperature gradient ratio represents the ratio of the hot spot edge temperature gradient in the hot spot region to the preset allowable hot spot edge temperature gradient. The preset allowable hot spot edge temperature gradient ratio is represented by the sum and average of the historical allowable hot spot edge temperature gradient ratios during the temperature dissipation process of historical photovoltaic modules in the database. The larger the hot spot edge temperature gradient ratio, the more drastic the temperature change at the hot spot edge, the clearer the hot spot boundary, reducing interference in feature point extraction, and resulting in high warning timeliness. Conversely, a smaller ratio indicates that the hot spot boundary is blurred, increasing interference in feature point extraction, and resulting in low warning timeliness.

[0049] If the obtained hot spot edge temperature gradient ratio is less than the minimum value of the historical hot spot edge temperature gradient ratio in the database, it is determined that there is no hot spot effect state fault (i.e., no hot spot effect state warning is issued) and short circuit fault detection is performed. The minimum value of the historical hot spot edge temperature gradient ratio is represented by the sum and average of the minimum values ​​of the historical hot spot edge temperature gradient ratios during the temperature dissipation process of each historical photovoltaic module in the database. If the obtained hot spot edge temperature gradient ratio is greater than the maximum value of the historical hot spot edge temperature gradient ratio in the database, it indicates that the abnormal fault of the hot spot effect state has exceeded the adjustable range (i.e., the hot spot effect state is not up to standard), and an advanced hot spot effect state warning is issued to prompt the preset personnel to take corresponding advanced warning measures. For example, maintenance personnel go to the hot spot monitoring area for on-site verification to confirm the hot spot situation and intervene and adjust it. The maximum value of the historical hot spot edge temperature gradient ratio is represented by the sum and average of the maximum values ​​of the historical hot spot edge temperature gradient ratios during the temperature dissipation process of each historical photovoltaic module in the database.

[0050] In this embodiment, the hot spot effect status is accurately determined by comparing the ratio of the temperature gradient at the edge of the hot spot with historical data in the database: a primary warning is triggered when the ratio is within a preset range, indicating that the fault is recoverable; a higher warning is activated when the ratio is greater than the maximum value, requiring manual intervention; and a hot spot is excluded and a short circuit is detected when the ratio is less than the minimum value. The advantage lies in combining historical data to achieve tiered warnings, which reduces misjudgments when the boundary is ambiguous, reflects the clarity of the hot spot boundary through the ratio, improves the accuracy of feature point extraction and the timeliness of warnings, and balances automated monitoring with precise intervention, ensuring the safe and efficient operation of photovoltaic modules.

[0051] Furthermore, a preliminary hot spot effect state early warning is performed. The specific steps are as follows: the obtained hot spot edge temperature gradient ratio deviation is input into the database, and the mapping relationship between it and the corresponding sampling frequency adjustment amount and filter window size adjustment amount is obtained to obtain the sampling frequency increase amount and filter window size increase amount. The hot spot edge temperature gradient ratio deviation represents the difference between the hot spot edge temperature gradient ratio during the detection process in the hot plate area and the preset hot spot edge temperature gradient ratio. The preset hot spot edge temperature gradient ratio is represented by the sum and average of the historical hot spot edge temperature gradient ratios during the temperature dissipation process of historical photovoltaic modules in the database.

[0052] It is important to understand that as the sampling frequency and filter window size increase, the monitoring process monitors the change in the boundary temperature fluctuation range corresponding to the hot spot edge region within a preset time period. The boundary temperature fluctuation range represents the absolute value of the difference between the maximum and minimum temperatures at the hot spot boundary in the hot spot edge region. If the monitored boundary temperature fluctuation range is within the preset allowable range of boundary temperature fluctuation in the database, the monitoring of the initial hot spot effect state continues until the obtained hot spot edge temperature gradient ratio is less than the preset hot spot edge temperature gradient ratio in the database, thus restoring the boundary temperature fluctuation range to the preset allowable range. If the monitored boundary temperature fluctuation range is not within the preset allowable range of boundary temperature fluctuation in the database, a boundary temperature fluctuation range warning is issued, prompting the personnel to take corresponding measures, namely, to investigate whether there are problems such as equipment response delay or instruction execution error during parameter adjustment. The preset allowable range of boundary temperature fluctuation represents the closed interval corresponding to the maximum and minimum values ​​of the historical preset boundary temperature fluctuations during the temperature dissipation process of historical photovoltaic modules in the database.

[0053] After adjusting the sampling frequency and filter window size, the timeliness analysis of hot spot effect warning is re-performed. If the newly acquired hot spot edge temperature gradient ratio is less than the minimum historical hot spot edge temperature gradient ratio in the database, it is determined that the hot spot effect state fault recovery is completed and short circuit fault detection is performed. Otherwise, advanced hot spot effect state warning is performed.

[0054] In this embodiment, the sampling frequency can be optimized using a particle swarm optimization algorithm. The sampling frequency is increased accordingly based on the deviation of the temperature gradient ratio at the edge of the hot spot obtained by the hot spot effect, so as to obtain temperature data changes in a timely and accurate manner, capture subtle temperature anomaly characteristic signals, and provide richer data information for fault early warning. The filter window size is determined and adjusted by a moving average algorithm. When filtering the temperature data of the hot spot effect, the moving average algorithm can increase the size of the filter window according to the noise level and temperature change trend of the data, so that the filtered data can reflect the real-time temperature changes more quickly.

[0055] This example establishes a mapping relationship between the deviation of the temperature gradient ratio at the hot spot edge and the adjustment of the sampling frequency and the adjustment of the filter window size, thereby achieving dynamic adaptation of the early warning parameters. It can accurately adjust the monitoring strategy according to the real-time changes in the hot spot state, improve the pertinence and sensitivity of the primary early warning, and thus enhance the fault tolerance and completeness of the overall early warning system. This helps to adopt appropriate response strategies at different stages of the hot spot effect development, minimizing the negative impact of the hot spot effect.

[0056] Furthermore, a short-circuit warning timeliness analysis is conducted. The specific steps are as follows: During the temperature-waveform extraction process on the photovoltaic module surface, the differences between the obtained harmonic component ratio of the module waveform and the preset harmonic component ratio of the module waveform in the database are compared. At the same time, the differences between the obtained peak current ratio of the module waveform and the preset peak current ratio of the module waveform in the database are also compared, resulting in the obtained harmonic component ratio parameter and peak current ratio parameter. The obtained harmonic component ratio parameter and peak current ratio parameter are summed and averaged to obtain the temperature-waveform extraction interference index. The temperature-waveform extraction interference index represents the quantitative data on the degree of influence of the harmonic component ratio and peak current ratio of the module waveform on the short-circuit warning timeliness during the temperature-waveform extraction process. The temperature-waveform extraction process typically refers to the simultaneous acquisition of real-time temperature data and output current waveform signals on the photovoltaic module surface, and the extraction of features such as harmonic components and peak current through signal processing, providing a dynamic monitoring link for raw parameters for short-circuit state identification.

[0057] The preset harmonic component ratio of the component waveform is represented by the summation and averaging of the harmonic component ratios of historical component waveforms extracted from the temperature-waveform extraction process of historical photovoltaic component surfaces in the database. The preset peak current ratio of the component waveform is represented by the summation and averaging of the peak current ratios of historical component waveforms extracted from the temperature-waveform extraction process of historical photovoltaic component surfaces in the database.

[0058] The component waveform harmonic component ratio parameter represents the ratio of the acquired component waveform harmonic component ratio to the preset component waveform harmonic component ratio. The component waveform current peak value ratio parameter represents the ratio of the acquired component waveform current peak value to the preset component waveform current peak value. The component waveform harmonic component ratio represents the ratio of the average value of the effective value of each harmonic component in the current waveform output by the component to the effective value of the fundamental wave. This ratio directly reflects the degree of waveform distortion. The higher the distortion rate, the more obvious the short circuit characteristics. At this time, the influence of interference factors on the warning is smaller, and the short circuit warning response speed is faster and the warning timeliness is higher. Conversely, if the distortion rate is low and changes slowly, it means that the short circuit characteristics are not prominent. At this time, the warning is easily affected by other factors. If there is interference from normal fluctuations or measurement errors, the timeliness of the early warning will be reduced. The peak current ratio of the component waveform represents the ratio of the average value of each current peak in the current waveform output by the component to the corresponding rated current peak. It reflects the intensity of the sudden, high-intensity current spike that appears in the current waveform in a very short time. The higher the peak ratio, the more severe the current impact during a short circuit (a short circuit will cause the current to increase instantaneously, and the peak value will soar). The transient changes of the short circuit characteristics are more significant, making it easier to quickly capture abnormal currents, and thus the early warning timeliness is higher. Conversely, if the peak ratio is low and the fluctuation is gentle, the current peak value deviates little from the normal state and the change is not obvious. This may cause the short circuit characteristics to be masked by the waveform fluctuations during normal operation, resulting in a delayed early warning response.

[0059] In this embodiment, by comparing the differences between the component waveform harmonic component ratio and the current peak ratio and preset values, and forming corresponding ratio parameters, a multi-dimensional and accurate characterization of short-circuit characteristics is achieved. The temperature-waveform extraction interference index is obtained by summing and averaging the two ratio parameters, realizing the quantitative integration of factors affecting the timeliness of short-circuit warning. The summation and averaging are used to integrate the two dimensions of short-circuit characteristic parameters, namely the component waveform harmonic component ratio and the current peak ratio, into a unified quantitative index. Both affect the timeliness of short-circuit warning. Averaging can balance their respective weights, avoid the one-sidedness of a single parameter, and transform multi-dimensional information into intuitive values, which facilitates rapid judgment of the comprehensive interference level and improves analysis efficiency and objectivity.

[0060] like Figure 4 The diagram shows a flowchart of temperature-waveform extraction optimization provided in an embodiment of the present invention. The specific design logic is as follows: After obtaining the temperature-waveform extraction interference index, the database determines whether the abnormal fault is within the allowable range and a first-level temperature-waveform extraction status warning is issued. If the value is less than the preset value, it is determined that there is no temperature-waveform extraction status fault and the next stage of sine wave signal extraction is performed. If the value is greater than the historical maximum value, the abnormal fault is within the unadjustable range and a second-level temperature-waveform extraction status warning is issued to prompt the preset personnel to take corresponding second-level warning measures, so as to realize the temperature-waveform extraction optimization judgment and warning.

[0061] Further understanding is needed regarding temperature-waveform extraction optimization judgment and early warning, which specifically includes: temperature-waveform extraction status judgment and temperature-waveform extraction status early warning. Temperature-waveform extraction status early warning includes primary and secondary temperature-waveform extraction status early warnings. Specifically, the temperature-waveform extraction status judgment involves the following steps: if the acquired temperature-waveform extraction interference index is within the allowable range of temperature-waveform extraction interference in the database, then the abnormal fault in the temperature-waveform extraction status is determined to be within a recoverable range, and a primary temperature-waveform extraction status early warning is issued. The allowable range of temperature-waveform extraction interference represents the closed interval corresponding to the maximum and minimum values ​​of historical temperature-waveform extraction interference during the historical temperature-waveform extraction process on the surface of the photovoltaic module in the database. If the acquired temperature-waveform extraction interference index is less than the preset temperature-waveform extraction interference index in the database, then it is determined to be a fault without temperature-waveform extraction status (i.e., no temperature-waveform extraction status early warning is issued), and the next... In the first stage of sine wave signal extraction, the preset temperature-waveform extraction interference index is represented by the sum and average of the historical temperature-waveform extraction interference indices from the historical temperature-waveform extraction process of each photovoltaic module surface in the database. If the obtained temperature-waveform extraction interference index is greater than the maximum value of the historical temperature-waveform extraction interference index in the database, it is determined that the abnormal fault in the temperature-waveform extraction state is within the unadjustable range (i.e., the temperature-waveform extraction state is unqualified), and a second-level temperature-waveform extraction state warning is issued to prompt the preset personnel to take corresponding second-level warning measures. For example, maintenance personnel perform offline testing on the output waveform to determine whether the increase in harmonics or peak shift is caused by the module itself (such as internal short circuit, component aging, or switch device damage). If necessary, repair or replacement is carried out. The maximum value of the historical temperature-waveform extraction interference index is represented by the sum and average of the maximum values ​​of the historical temperature-waveform extraction interference indices from the historical temperature-waveform extraction process of each photovoltaic module surface in the database.

[0062] In this embodiment, the status is determined by comparing the temperature-waveform extraction interference index with the preset threshold in the database, which realizes accurate quantitative evaluation of the temperature-waveform extraction status, avoids fuzzy judgment of the extraction status, and provides a clear decision basis for subsequent processing. Secondly, a first-level and second-level hierarchical early warning mechanism is adopted to realize differentiated response to abnormal states of different degrees. This hierarchical response mode ensures timely intervention for minor abnormalities and efficient handling of serious faults, improving the practicality and accuracy of the early warning. In addition, for cases determined to be without faults, the process directly enters the next stage of sine wave signal extraction, optimizing the continuity of the process.

[0063] Furthermore, the first-level temperature-waveform extraction status early warning involves the following steps: The acquired temperature-waveform extraction interference index deviation is input into the database, and the corresponding passband bandwidth adjustment is mapped in the database; the fundamental frequency bandwidth of the distorted signal before the first-level temperature-waveform extraction status early warning is acquired, and this bandwidth reflects the change in the corresponding current peak value in the distorted signal; the acquired fundamental frequency bandwidth of the distorted signal is input into the database, and the increase in passband bandwidth is mapped in the database. By increasing the passband bandwidth of the bandpass filter, the risk of current peak frequency truncation caused by short circuit is reduced, thereby improving the accuracy and anti-interference capability of temperature-waveform extraction in short-circuit early warning; after the passband bandwidth is increased, the short-circuit early warning timeliness analysis is re-performed. If the re-acquired temperature-waveform extraction interference index is less than the preset temperature-waveform extraction interference index in the database, it is determined that the temperature-waveform extraction status fault has been recovered, and the next stage of sine wave signal extraction is performed; otherwise, the second-level temperature-waveform extraction status early warning is performed.

[0064] In this embodiment, the acquisition of the fundamental frequency bandwidth of the distorted signal usually relies on Fourier transform and its derived signal analysis algorithms. By performing Fourier transform on the distorted signal, it can be transformed from the time domain to the frequency domain to obtain the spectral distribution of the signal. The frequency corresponding to the peak in the spectrum is the frequency of each component, and the frequency range near the peak reflects the bandwidth characteristics of the component. Increasing the passband bandwidth of the bandpass filter effectively reduces the risk of current peak frequency cutoff caused by short circuit, and improves the anti-interference capability and data integrity of temperature-waveform extraction in short circuit warning.

[0065] This example achieves precise adaptation and adjustment of bandpass filter parameters by mapping the temperature-waveform interference index deviation into a database to obtain the passband bandwidth adjustment amount, and combining it with the distortion signal fundamental frequency bandwidth mapping to obtain the passband bandwidth increase amount. Secondly, by increasing the passband bandwidth of the bandpass filter, the key features in the distortion signal can be captured more completely, reducing feature loss caused by insufficient bandwidth. In addition, after adjusting the passband bandwidth, the short-circuit warning timeliness analysis is re-performed, and the fault recovery is determined based on whether the interference index has recovered to the preset value, thus constructing a closed-loop optimization verification mechanism.

[0066] like Figure 5The flowchart shown is a process for a smart fault detection and early warning method for polycrystalline silicon solar panels provided in an embodiment of the present invention. The method includes the following steps: S1, real-time monitoring of the operating status of the polycrystalline silicon solar panel, performing power attenuation early warning timeliness analysis on the conversion control process of the sine wave signal, and simultaneously performing sine wave signal extraction optimization judgment and early warning; S2, after the sine wave signal extraction is qualified, based on the temperature gradient ratio of the hot spot edge obtained in the corresponding hot spot region of the photovoltaic module, performing hot spot effect early warning timeliness analysis on the temperature heat dissipation process of the photovoltaic module, and simultaneously performing hot spot effect optimization judgment and early warning; S3, based on the results of hot spot effect optimization judgment and early warning, performing short-circuit early warning timeliness analysis on the temperature-waveform extraction process on the surface of the photovoltaic module to determine whether to perform temperature-waveform extraction optimization judgment and early warning.

[0067] In this embodiment, the method performs correlation analysis on the signal characteristics, temperature characteristics and short-circuit characteristics of photovoltaic modules during the intelligent fault detection and early warning process of polycrystalline silicon solar panels, realizing the integrated application of multi-dimensional characteristics. Each stage incorporates a dual mechanism of optimization judgment and fault early warning, realizing dynamic optimization and accurate response of fault detection. Through targeted monitoring of characteristic parameters at different stages, the comprehensiveness of monitoring the operating status of solar panels is improved.

[0068] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0069] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0070] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers 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 the present invention.

[0071] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0072] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A smart fault detection and early warning system and method for polycrystalline silicon solar panels, characterized in that, include: Power attenuation fault detection module, hot spot effect fault detection module, and short circuit fault detection module; The power attenuation fault detection module is used to monitor the operating status of the polycrystalline silicon solar panel in real time, perform power attenuation early warning timeliness analysis on the conversion control process of the sine wave signal, and perform sine wave signal extraction optimization judgment and early warning. The sine wave signal extraction optimization means reducing harmonic distortion in the signal conversion process by adjusting the carrier frequency and dead time of the SPWM. The hot spot effect fault detection module is used to perform hot spot effect early warning timeliness analysis on the temperature heat dissipation process of the photovoltaic module based on the hot spot edge temperature gradient ratio obtained in the corresponding hot spot area of ​​the photovoltaic module after the sinusoidal signal is qualified. At the same time, it performs hot spot effect optimization judgment and early warning. The hot spot effect optimization means adjusting the sampling frequency and filter window size of the infrared thermal imager to enhance the acquisition accuracy and anti-interference ability of the temperature data in the hot spot area. The short-circuit fault detection module is used to perform short-circuit warning timeliness analysis on the temperature-waveform extraction process on the surface of the photovoltaic module based on the results of hot spot effect optimization judgment and warning, so as to determine whether to perform temperature-waveform extraction optimization judgment and warning. The temperature-waveform extraction optimization means adjusting the passband bandwidth of the bandpass filter to accurately filter out noise signals that are unrelated to short-circuit faults.

2. The intelligent fault detection and early warning system for polycrystalline silicon solar panels as described in claim 1, characterized in that, The specific steps for the power attenuation early warning timeliness analysis are as follows: In the conversion control process of sinusoidal signal, based on the obtained useful signal power ratio and interference signal frequency ratio, the difference is compared with the useful signal power ratio and interference signal frequency ratio set in the database to obtain the useful signal power ratio score and interference signal frequency ratio score. The result of inversely proportional processing of the obtained useful signal power ratio score is summed and averaged with the interference signal frequency ratio score to obtain the sinusoidal signal interference extraction index. The sinusoidal signal interference extraction index represents the quantitative data of the influence of sinusoidal signal data on the timeliness of power attenuation warning. The sinusoidal signal data includes the useful signal power ratio and the interference signal frequency ratio.

3. The intelligent fault detection and early warning system for polycrystalline silicon solar panels as described in claim 2, characterized in that, The sine wave signal extraction optimization judgment and early warning specifically includes: sine wave signal extraction status judgment and sine wave signal extraction status early warning; The specific steps for determining the extraction state of the sine wave signal include: If the obtained sine wave signal extraction interference index is within the allowable range of sine wave signal extraction interference in the database, it is determined that the abnormal fault of the sine wave signal extraction state is within the recoverable range and the first sine wave signal extraction state warning is issued. If the obtained sinusoidal signal extraction interference index is less than the minimum historical sinusoidal signal extraction interference index in the database, it is determined to be a sinusoidal signal extraction failure, and hot spot effect detection is performed. The sinusoidal signal extraction failure indicates that the sinusoidal signal extraction status is qualified, and no sinusoidal signal extraction status warning is issued at this time. If the obtained sine wave signal extraction interference index is greater than the maximum value of the historical sine wave signal extraction interference index in the database, it is determined that the abnormal fault in the sine wave signal extraction state is within the unadjustable range, and a second sine wave signal extraction state warning is issued to prompt the preset personnel to take corresponding secondary warning measures.

4. The intelligent fault detection and early warning system for polycrystalline silicon solar panels as described in claim 3, characterized in that, The first sine wave signal is used to extract the status warning. The specific steps are as follows: The interference index deviation of the acquired sinusoidal signal is extracted and input into the database for mapping matching to obtain the corresponding carrier frequency adjustment and dead time adjustment. The proportion of high-frequency noise signal and the inverter switching delay are obtained before the first sine wave signal extraction state warning. The proportion of high-frequency noise signal is used to reflect the degree of high-frequency interference in the sine wave signal, and the inverter switching delay is used to reflect the response lag time of the switching device on / off. Based on the deviation of the high-frequency noise signal ratio, the carrier frequency increase is matched in the database. By increasing the carrier frequency of SPWM, the gap between it and the inherent frequency of the power attenuation characteristic signal is widened, thereby reducing the interference caused by frequency overlap. Based on the obtained inverter switching delay deviation, the dead time reduction amount is matched in the database. By reducing the dead time of SPWM, the glitches caused by switching action are reduced, and the signal-to-noise ratio of the sine wave signal is improved. After adjusting the carrier frequency and dead time, the power attenuation warning timeliness analysis is performed again. If the interference index of the newly acquired sine wave signal is less than the minimum value of the historical sine wave signal extraction interference index in the database, it is determined that the sine wave signal extraction status fault has been recovered and hot spot effect detection is performed. Otherwise, a second sine wave signal extraction status warning is performed.

5. The intelligent fault detection and early warning system for polycrystalline silicon solar panels as described in claim 1, characterized in that, The hot spot effect optimization judgment and early warning specifically includes: hot spot effect state judgment and hot spot effect state early warning, wherein the hot spot effect state early warning includes primary hot spot effect state early warning and advanced hot spot effect state early warning. The determination of the hot spot effect state includes the following specific steps: If the obtained hot spot edge temperature gradient ratio is within the preset range of hot spot edge temperature gradient ratio in the database, it indicates that the abnormal fault of the hot spot effect state is within the recoverable range, and a primary hot spot effect state warning is issued. If the obtained hot spot edge temperature gradient ratio is less than the minimum value of the historical hot spot edge temperature gradient ratio in the database, it is determined that there is no hot spot effect state fault and short circuit fault detection is performed. If the obtained hot spot edge temperature gradient ratio is greater than the maximum value of the historical hot spot edge temperature gradient ratio in the database, it indicates that the abnormal fault of the hot spot effect state has exceeded the adjustable range, and an advanced hot spot effect state warning is issued to prompt the preset personnel to take corresponding advanced warning measures.

6. The intelligent fault detection and early warning system for polycrystalline silicon solar panels as described in claim 5, characterized in that, The specific steps for the early warning of the primary hot spot effect state are as follows: The obtained hot spot edge temperature gradient ratio deviation is input into the database. Based on the mapping relationship between the hot spot edge temperature gradient ratio deviation and the increase in sampling frequency in the database, the increase in sampling frequency is obtained. At the same time, the obtained hot spot edge temperature gradient ratio deviation is input into the database. Based on the mapping relationship between the hot spot edge temperature gradient ratio deviation and the increase in filter window size in the database, the increase in filter window size is obtained. As the sampling frequency and filter window size increase, monitor the change in the boundary temperature fluctuation range corresponding to the hot spot edge region within a preset increase period. If the monitored boundary temperature fluctuation range is within the preset allowable range of boundary temperature fluctuation in the database, the monitoring of the early warning process of the primary hot spot effect state will continue until the obtained hot spot edge temperature gradient ratio is less than the preset hot spot edge temperature gradient ratio in the database. If the monitored boundary temperature fluctuation range is not within the preset allowable range of boundary temperature fluctuation in the database, an early warning of the boundary temperature fluctuation range will be issued, and the preset personnel will be prompted to take corresponding measures. After adjusting the sampling frequency and filter window size, the timeliness analysis of hot spot effect warning is re-performed. If the newly acquired hot spot edge temperature gradient ratio is less than the minimum historical hot spot edge temperature gradient ratio in the database, it is determined that the hot spot effect state fault recovery is completed and short circuit fault detection is performed. Otherwise, advanced hot spot effect state warning is performed.

7. The intelligent fault detection and early warning system for polycrystalline silicon solar panels as described in claim 1, characterized in that, The specific steps for the short-circuit early warning timeliness analysis are as follows: In the process of temperature-waveform extraction on the surface of photovoltaic modules, the harmonic component ratio of the obtained module waveform is compared with the preset harmonic component ratio of the module waveform in the database. At the same time, the peak current ratio of the obtained module waveform is compared with the preset peak current ratio of the module waveform in the database to obtain the harmonic component ratio parameter and the peak current ratio parameter of the module waveform. The obtained component waveform harmonic component ratio parameter and component waveform current peak ratio parameter are summed and averaged to obtain the temperature-waveform extraction interference index. The temperature-waveform extraction interference index represents the quantitative data of the degree of influence of the component waveform harmonic component ratio and component waveform current peak ratio on the timeliness of short circuit warning during the temperature-waveform extraction process.

8. The intelligent fault detection and early warning system for polycrystalline silicon solar panels as described in claim 7, characterized in that, The temperature-waveform extraction optimization judgment and early warning specifically includes: temperature-waveform extraction status judgment and temperature-waveform extraction status early warning, wherein the temperature-waveform extraction status early warning includes a first-level temperature-waveform extraction status early warning and a second-level temperature-waveform extraction status early warning. The temperature-waveform extraction status determination process includes the following steps: If the obtained temperature-waveform extraction interference index is within the allowable range of temperature-waveform extraction interference in the database, it is determined that the abnormal fault of the temperature-waveform extraction status is within the recoverable range, and a first-level temperature-waveform extraction status warning is issued. If the obtained temperature-waveform extraction interference index is less than the preset temperature-waveform extraction interference index in the database, it is determined to be a fault of no temperature-waveform extraction state, and the next stage of sine wave signal extraction is performed. If the obtained temperature-waveform extraction interference index is greater than the maximum value of the historical temperature-waveform extraction interference index in the database, it is determined that the abnormal fault in the temperature-waveform extraction status is within the unadjustable range, and a secondary temperature-waveform extraction status warning is issued to prompt the preset personnel to take corresponding secondary warning measures.

9. The intelligent fault detection and early warning system for polycrystalline silicon solar panels as described in claim 8, characterized in that, The specific steps for the first-level temperature-waveform extraction status early warning are as follows: The obtained temperature-waveform interference index deviation is input into the database, and the corresponding passband bandwidth adjustment is obtained by mapping in the database. The fundamental frequency bandwidth of the distortion signal before the first-level temperature-waveform extraction state warning is obtained. The fundamental frequency bandwidth of the distortion signal is used to reflect the change of the corresponding current peak value in the distortion signal. The obtained fundamental frequency bandwidth of the distorted signal is input into the database, and the increase in passband bandwidth is obtained by mapping in the database. By increasing the passband bandwidth of the bandpass filter, the risk of peak current cutoff caused by short circuit is reduced. After the passband bandwidth is increased, the short-circuit warning timeliness analysis is performed again. If the temperature-waveform extraction interference index is less than the preset temperature-waveform extraction interference index in the database, it is determined that the temperature-waveform extraction state fault has been recovered and the next stage of sine wave signal extraction is carried out. Otherwise, the second-level temperature-waveform extraction state warning is carried out.

10. A method for intelligent fault detection and early warning of polycrystalline silicon solar panels, characterized in that, Includes the following steps: S1 monitors the operating status of polycrystalline silicon solar panels in real time, performs power attenuation early warning timeliness analysis on the conversion control process of sine wave signals, and performs sine wave signal extraction optimization judgment and early warning. S2, after the sinusoidal signal is extracted and qualified, based on the ratio of the temperature gradient at the edge of the hot spot obtained in the corresponding hot spot region of the photovoltaic module, the timeliness analysis of the hot spot effect early warning of the temperature heat dissipation process of the photovoltaic module is carried out, and the hot spot effect optimization judgment and early warning are carried out at the same time. S3. Based on the results of the hot spot effect optimization judgment and early warning, a short-circuit early warning timeliness analysis is performed on the temperature-waveform extraction process on the surface of the photovoltaic module to determine whether to perform temperature-waveform extraction optimization judgment and early warning.

Images