Evaluation method of photovoltaic module service life
By arranging soundprint sensors on photovoltaic modules, collecting and analyzing soundprint data in real time, predicting the aging degree and life, the problem of difficulty in detecting internal microstructure and non-real-time monitoring in the existing technology is solved, and the comprehensive life of photovoltaic modules is achieved , real-time evaluation and low-cost maintenance.
Patent Information
- Application Number
- CN202411126932.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-16
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-08-16
AI Technical Summary
The existing photovoltaic module life evaluation methods are difficult to detect internal microstructure problems, non-real-time monitoring, limited detection capabilities of hidden defects, and cannot directly reflect the internal dynamic changes of materials and installation process problems.
The soundprint sensor is used to arrange it on the photovoltaic module, collect voiceprint data from different aging stages, predict the aging degree and remaining life of the component through data processing and aging model, and output early warning signals in real time.
Real-time life assessment without affecting the normal working state of the photovoltaic panel, timely discover potential problems, and take maintenance measures in advance, ensuring the comprehensiveness and low cost of life assessment.
Smart Images

Figure CN118659736B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of photovoltaic modules, and in particular to a method for evaluating the service life of a photovoltaic module. Background Art
[0002] With the rapid development of the photovoltaic industry, solar photovoltaic power generation technology has been widely used in people's daily lives. For a product based on solar power generation, photovoltaic modules are its core components. They are combined with the energy demand entity, making the energy demand entity an independent green power generation entity. The energy entity is charged through sunlight, which solves the various electricity needs of the energy demand entity and improves the efficiency of energy production, recycling and sharing. In order to ensure the normal use of photovoltaic modules, it is necessary to evaluate their lifespan to facilitate timely maintenance and replacement by staff.
[0003] After searching, the applicant found the following prior art:
[0004] In a method for evaluating the service life of a photovoltaic module with application number 201810404539.1, a technical solution is provided, including conducting at least one set of environmental weathering resistance experiments on the photovoltaic module according to preset experimental standards; when the photovoltaic module passes at least one set of environmental weathering resistance experiments, the service life of the photovoltaic module is determined according to the evaluation life corresponding to the experimental standard. Through this application solution, an effective and feasible evaluation method for the service life of photovoltaic modules is provided.
[0005] The following is a summary of the photovoltaic panel life detection of the existing technology, which mainly includes: 1. Initial performance evaluation: conduct a comprehensive electrical performance test on the photovoltaic panel, including the measurement of parameters such as output power, open circuit voltage, short circuit current, and record the initial value; 2. Appearance inspection: carefully check whether there are obvious defects such as damage, cracks, bubbles, delamination, etc. on the surface of the photovoltaic panel; 3. Environmental monitoring: continuously monitor key parameters such as temperature, humidity, and light intensity of the installation environment; 4. Regular electrical performance test: repeat the electrical performance test at a certain time interval (such as every year), and compare it with the initial value to calculate the performance attenuation; 5. EL detection (optional): use electroluminescence detection technology to check whether there are hidden cracks, impurities and other problems inside the photovoltaic panel; 6. Thermal imaging detection: use a thermal imager to detect the temperature distribution of the photovoltaic panel when it is working, and find possible abnormal hot spots; 7. Material analysis (optional): perform composition analysis and microstructure observation on the materials of the photovoltaic panel to evaluate the aging condition of the material; 8. Data collation and analysis: organize and analyze all test data, draw performance change curves, etc.; 9. Comprehensive evaluation and judgment: Based on the test results and analysis, comprehensively judge the life status and remaining service life of the photovoltaic panels.
[0006] The above prior art mainly has the following defects:
[0007] 1. It is difficult to detect subtle internal structural problems: It is difficult to accurately detect some early structural changes such as smaller internal cracks and delamination at an early stage.
[0008] 2. Non-real-time continuous monitoring: Except for environmental monitoring, other tests are mostly conducted periodically, which may miss some rapid changes that occur during the interval;
[0009] 3. The ability to detect some hidden defects is limited, such as some problems that do not affect electrical performance but pose a potential threat to the structure;
[0010] 4. Unable to directly reflect the dynamic changes inside the material;
[0011] 5. The detection of problems related to the installation process is not direct, and problems such as loose connections that may occur during the installation process cannot be detected.
[0012] In summary, it is necessary to provide a new technical solution to solve the above technical problems. Summary of the invention
[0013] The present application provides a method for evaluating the service life of a photovoltaic module, comprising the following steps:
[0014] S1: Arrange the soundprint sensor on the photovoltaic module;
[0015] S2: The voiceprint sensor collects the voiceprints generated by the photovoltaic modules at different aging stages, converts the collected voiceprints into electrical signals, and sends them to the data processing unit;
[0016] S3: The data processing unit analyzes the received signal and extracts characteristic parameters related to the aging of the photovoltaic components;
[0017] S4: Establish an aging model by training the voiceprint data generated by a large number of aging photovoltaic modules;
[0018] S5: The voiceprint data collected in real time is input into the aging model, and the aging model predicts the aging degree and remaining life of the component based on the voiceprint characteristics.
[0019] As a preferred solution, it also includes S6, outputting the aging degree and remaining life predicted in S5 to the user in an intuitive manner.
[0020] As a preferred solution, it also includes S7, when the aging degree of the photovoltaic module exceeds the set threshold, the system sends out a warning signal to remind the user to take corresponding measures.
[0021] The S1 includes:
[0022] S11: Collect images of the entire photovoltaic module, analyze the flatness of the collected images, set the locations with poor flatness as voiceprint collection modules, and preset multiple voiceprint sensors in each voiceprint collection module to form a sensor network;
[0023] S12: Setting the initial sampling frequency and the sound wave characteristic information triggering threshold of the voiceprint sensor in each voiceprint collection module;
[0024] S13: The voiceprint sensor collects the voiceprint information of the photovoltaic module at a regular interval and transmits the collected information;
[0025] S14: when the voiceprint sensors in one or more voiceprint collection modules detect sound wave feature information, the sound wave feature information in each voiceprint collection module is collected to form a sound wave data set; the sound wave data set in each voiceprint collection module is transmitted to the comparison module;
[0026] S15: The comparison module compares the sound wave data set in the voiceprint collection module with the previous data set of the corresponding voiceprint collection module;
[0027] S16: Obtaining the acoustic wave feature deviation of the voiceprint collection module according to the difference between the acoustic wave data set and the previous data set;
[0028] S17: setting the optimal interval between adjacent voiceprint sensors and the optimal sampling frequency interval of each voiceprint acquisition module according to the difference between the acoustic wave feature deviation and the acoustic wave feature information trigger threshold;
[0029] S18: Locate the sound wave feature points collected by the voiceprint collection module at the sampling frequency in S17; S19: Input the difference between the sound wave feature deviation and the sound wave feature information trigger threshold into the neural network model to calculate and output the optimal placement position of the voiceprint sensor.
[0030] As a preferred solution, the sound wave characteristic information in S12 includes vibration information, abnormal sound, sound intensity and other information, wherein the vibration information includes the frequency and amplitude of the vibration.
[0031] As a preferred solution, S17 includes:
[0032] A1: Classify the differences into small difference interval, medium difference interval, medium-large difference interval, and large difference interval;
[0033] A2: Set the corresponding sampling frequency and the distance between adjacent voiceprint sensors according to the classification of the difference.
[0034] As a preferred solution, it also includes A3, which adjusts the distance between adjacent voiceprint sensors and the optimal sampling frequency range in the voiceprint collection module in real time.
[0035] As a preferred solution, the coordinates of the best placement of the voiceprint sensor of the output layer of the neural network model in S19 are G (X, Y):
[0036] G(X,Y) = f(∑(w 1i * (sound wave characteristic deviation - sound wave characteristic trigger threshold) + w 2i * X coordinate of the acoustic wave feature point + w 3i * Y coordinate of the acoustic wave feature point) + b 1i );
[0037] Among them, i represents the node of the hidden layer, i = 1-n; f is the activation function, w 1i 、w 2i 、w 3i is the weight of different neurons, b 1i is the bias term.
[0038] As a preferred solution, the aging stages in S2 include an initial aging stage, a mid-term aging stage, a severe aging stage, and a near-failure stage.
[0039] As a preferred solution, in S3, the characteristic parameters include sound frequency, sound intensity, sound duration, abnormal sound, sound attenuation, sound resonance, and sound amplitude.
[0040] As a preferred solution, the voiceprint data generated by the photovoltaic components in S4 include: vibration frequency, vibration amplitude, vibration duration, vibration periodicity, vibration pattern, sound frequency distribution, main frequency characteristics, frequency spectrum characteristics, amplitude change pattern, sound rhythm and rhythm, abnormal sound characteristics, timbre characteristics, and complexity of the voiceprint.
[0041] This application uses soundprints to evaluate the life of photovoltaic panels, which has the following advantages:
[0042] 1. No additional detection components are required to contact the photovoltaic panels, and the life span can be evaluated without affecting the normal working state of the photovoltaic panels, thus reducing interference to the photovoltaic panels;
[0043] 2. Real-time monitoring of the sound pattern changes of photovoltaic panels during operation to promptly detect potential problems and abnormal conditions;
[0044] 3. It can detect early signs through subtle changes in the soundprint before the photovoltaic panels show obvious performance degradation or failure, thus buying time for taking maintenance measures in advance.
[0045] 4. The entire photovoltaic panel array can be tested at the same time to obtain more comprehensive life information, rather than being limited to local testing, to ensure the comprehensiveness of photovoltaic panel life assessment;
[0046] 5. The detection device is simple, easy to install and operate, thus ensuring a low cost of use and having certain advantages in operating costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0047] Figure 1 It is the logic block diagram of this application; DETAILED DESCRIPTION
[0048] The following is combined with Figure 1 The specific implementation of the present invention is described in detail. It should be noted that the specific implementation described here is only used to illustrate and explain the present invention, and is not used to limit the present invention. Embodiment 1
[0049] This embodiment provides a method for evaluating the service life of a photovoltaic module, comprising the following steps:
[0050] S1: Arrange the soundprint sensor on the photovoltaic module;
[0051] S2: The voiceprint sensor collects the voiceprints generated by the photovoltaic components at different aging stages, converts the collected voiceprints into electrical signals, and sends them to the data processing unit, which amplifies, filters, digitizes, and processes the electrical signals, extracts and analyzes the characteristics of the voiceprints, so as to perform subsequent evaluation, identification, and other operations; preferably, the aging stages include the initial aging stage, the mid-aging stage, the severe aging stage, and the near-failure stage;
[0052] Among them, in the initial aging stage, photovoltaic modules will have some subtle structural changes and performance degradation, and the soundprints emitted at this time will have slight changes. The subtle structural changes include but are not limited to: changes in the internal lattice structure of the battery cell: resulting in changes in the physical properties of the battery cell, affecting the soundprint characteristics; slight separation between the packaging material and the battery cell interface: producing slight vibration differences; loosening of the junction box or connection parts: causing specific sound changes; performance degradation includes but is not limited to: unstable output power: changes in the soundprint can reflect the situation where the power output is sometimes high and sometimes low; the energy conversion efficiency gradually decreases: reflected in the change of the overall soundprint characteristics; abnormal heating: the component causes local overheating due to structural problems, which is reflected in the soundprint.
[0053] In the mid-term aging stage, there will be more obvious performance degradation, such as material fatigue, loose connections, etc. The sound pattern will become relatively complex, with some abnormal sound characteristics or frequency changes;
[0054] In the severe aging stage, components have greater potential for failure, such as cracks and partial failures. At this time, the soundprint will show characteristics such as large fluctuations and strong abnormal sounds.
[0055] As the voiceprint approaches failure, it will become chaotic, irregular, and of high intensity, and may even be accompanied by intermittent loud noises.
[0056] S3: The data processing unit analyzes the received signal and extracts characteristic parameters related to the aging of the photovoltaic components; the characteristic parameters include sound frequency, sound intensity, sound duration, abnormal sound, sound attenuation, sound resonance, sound amplitude, etc.
[0057] Among them, sound frequency: the aging of photovoltaic modules causes the appearance or change of certain specific frequency sounds. For example, when the packaging materials of photovoltaic modules age, tiny cracks or yellowing will appear, which will lead to higher sound pressure levels at specific frequencies; when hidden cracks appear in the battery cells, abnormal sounds will appear at specific frequencies;
[0058] Changes in sound intensity: The aging process is accompanied by changes in sound intensity. For example, when cracks or looseness appear in the packaging material, the sound intensity will increase at certain vibration frequencies. Changes in the internal lattice structure of the battery cell or tiny cracks may produce some abnormal friction or vibration sounds during operation, causing changes in the sound intensity at specific frequencies. Aging or poor contact of connecting lines lead to unstable current, causing intermittent strong noise and fluctuations in sound intensity. Deformation or loosening of the photovoltaic panel frame structure may cause a more obvious shaking sound than before under the action of external forces such as wind, and the sound intensity may also increase.
[0059] Sound duration characteristics: Compared with normal situations, aging-related sounds may have different duration characteristics. For example, when the packaging material ages and cracks or peels off, it will cause some intermittent "crackling" sounds, which may last for a short time and irregularly. When the battery cell has problems such as hidden cracks, it may occasionally emit a short "buzzing" sound during operation, which lasts for a short time. If the connecting parts are aged and loose, intermittent "creaking" sounds may be produced under certain conditions, such as wind and vibration, and the duration of each sound is uncertain. When the cooling system of the photovoltaic panel ages and its efficiency decreases, it may emit a long "buzzing" sound after running for a long time. This may be due to continuous abnormal vibration of components caused by overheating.
[0060] Abnormal sounds: specific noise patterns or intermittent abnormal sounds, such as: high-frequency whistling sounds indicate high-frequency oscillations in the circuit or abnormal friction in certain components; dull "bang bang" sounds indicate the sound of local expansion or contraction caused by changes in internal stress of the packaging material; continuous "rustling" sounds represent the sound produced by dust or debris rubbing against the surface of the photovoltaic panel; intermittent "crackling" sounds represent the sound produced by local discharge or slight arcing, etc.
[0061] Sound attenuation: The attenuation characteristics of sound during propagation may vary due to aging. For example, when the packaging material ages and cracks or becomes loose, the sound will be more easily scattered around, resulting in faster attenuation, and the sound that was originally audible at a certain distance will become more blurred; problems inside the battery cell will change the resonance and reflection characteristics of the sound, affecting the attenuation speed and degree of sound when it propagates inside and around the photovoltaic panel; aging connection lines will hinder the conduction of sound, causing the sound to decay faster and the propagation distance to be shorter; aging and deformation of the photovoltaic panel frame structure will change the path and method of sound propagation, causing the sound to decay faster or slower in certain directions; aging of the heat dissipation system will lead to abnormal temperature distribution, which will also indirectly affect the attenuation of sound, causing the sound attenuation law of certain frequencies to change, etc.
[0062] Sound resonance frequency: Aging of components may affect their resonance frequency. For example, as the packaging material ages and hardens or develops local defects, its elastic properties change, resulting in an increase in the resonance frequency or the appearance of multiple resonance frequency peaks. Changes in the lattice structure or cracks in the cell will change its inherent vibration characteristics, thereby causing the resonance frequency to shift. Aging and loosening of connecting parts will change the stiffness of the entire structure, causing the resonance frequency to decrease or fluctuate. Aging phenomena such as deformation or corrosion of the photovoltaic module frame will change its overall mechanical structure, and the resonance frequency will also change accordingly, and some new and abnormal resonance frequency points may appear.
[0063] Sound amplitude: When PV modules begin to age, their internal structures may become loose, connections may become poor, and material properties may degrade. These changes may cause abnormal vibrations and sounds under certain operating conditions, thereby increasing the sound amplitude. For example, aging may cause the modules to emit a louder "buzzing" or other abnormal sounds when operating, and the corresponding sound amplitude will be higher.
[0064] S4: Establish an aging model by training the soundprint data generated by a large number of aging photovoltaic modules; the aging model can associate the soundprint characteristics with the aging degree of the module. The establishment of the aging model requires a large amount of experimental data and precise algorithms to improve the accuracy of the assessment; the experimental data includes: vibration frequency, vibration amplitude, vibration duration, vibration periodicity, vibration mode, sound frequency distribution, main frequency characteristics, frequency spectrum characteristics, amplitude change mode, sound rhythm and rhythm, abnormal sound characteristics, timbre characteristics, complexity of the soundprint, etc.
[0065] The frequency of the above vibrations. Different faults or aging conditions cause the vibration of a specific frequency to appear or change.
[0066] The amplitude of vibration, the embodiment of sound intensity, reflects the intensity of the vibration of the photovoltaic panel. A larger amplitude indicates a more serious problem. The duration of vibration can show the duration characteristics of certain vibration phenomena. The periodicity or regularity of vibration, such as whether there is a fixed period of vibration, is helpful for analyzing the status of the photovoltaic panel. The mode (characteristic) of vibration, such as whether there is a specific type of vibration mode, such as intermittent vibration, continuous vibration, etc., and the difference in vibration in different parts. This vibration information can be combined with other data and analysis methods to comprehensively judge the health status and aging degree of the photovoltaic panel. The sound frequency distribution, different vibration modes will produce sounds of specific frequencies. For example, when the lattice structure of the battery cell changes due to aging and other factors, it will produce Some low-frequency vibration sounds may occur, and the corresponding frequency may be tens to hundreds of hertz; local expansion and contraction or tiny cracks in the packaging material may cause medium- and high-frequency "hissing" sounds, and the frequency may be hundreds to thousands of hertz; the shaking of the photovoltaic panel frame under the action of external forces such as wind may produce a relatively low-frequency "buzzing" sound, and the frequency may be tens to one or two hundred hertz; poor contact or intermittent discharge of the internal circuit may cause a higher-frequency "pop" sound, and the frequency may be around thousands of hertz; some local friction or jamming caused by aging may produce specific medium- and high-frequency sounds with a wide frequency range; main frequency characteristics, the main and representative sound frequency, for example, photovoltaic panels usually have some relatively stable low-frequency vibration sounds during normal operation, and the main frequency may be around 50-200 Hz, which is related to the overall vibration of the battery cells and frames; in the event of certain specific faults or aging, some abnormal main frequencies will appear, such as when there are serious problems with the packaging materials, there may be a main frequency of 500-1000 The sound of about 100 Hz represents a certain damage mode of the packaging material; when intermittent arcs appear in the connecting line, a high-frequency sound with a main frequency of several thousand Hz will be generated; when the heat dissipation system of the photovoltaic panel is abnormal, resulting in obvious local thermal expansion and contraction, the sound with a main frequency of 300-500 Hz may reflect this situation, etc. The specific main frequency characteristics will also vary depending on the specific structure, aging type and degree of the photovoltaic panel, and are not specifically limited here; the frequency spectrum characteristics represent the distribution of sound energy in the entire frequency range. For example, under normal circumstances, there is a certain basic energy distribution in the lower frequency band (below 100 Hz), which is consistent with the overall The sound quality of the battery is related to the inherent vibration of the structure; there will be some relatively discrete energy peaks in the mid-frequency band (100Hz-1000Hz), which are related to the specific vibration modes of some components; when the packaging material ages and cracks, new energy accumulation areas will appear in the mid-high frequency band (500Hz-2000Hz), representing the specific sound characteristics caused by the packaging material problem; if the battery cell has problems such as hidden cracks, abnormal energy will appear in a specific mid-frequency band (such as 300Hz-800Hz); when the connection line has poor contact or slight discharge, the high frequency band (near several thousand Hertz) will show obvious energy increase, etc.As the aging of photovoltaic panels deepens, the energy distribution of the entire frequency spectrum may become more complex and uneven, and there may be more scattered energy peaks or abnormal increases or decreases in the energy of certain frequency bands; the amplitude change mode represents the fluctuation of the amplitude over time, such as the gradual increase mode. With the development of aging, the internal structure becomes loose or damaged gradually, and the amplitude may show a trend of continuous increase; the intermittent increase mode, under certain specific conditions (such as temperature changes, strong winds, etc.), there is an intermittent sudden increase in the amplitude; the periodic fluctuation mode, such as related to the temperature difference between day and night or certain periodic external forces, causes the amplitude to show regular periodic fluctuations; the random fluctuation mode, due to the combined influence of multiple uncertain factors, the amplitude is Random fluctuations with no obvious rules; a sudden increase mode after a period of stability, in which the amplitude is relatively stable at the beginning, but after a certain point in time, the amplitude increases sharply due to aging and failure of key parts; a gradual decrease mode, in some cases, such as slight aging in the early stage, the amplitude may gradually decrease first, and then change according to the specific situation; the rhythm and rhythm of the sound: regular or periodic sound characteristics, for example, if the connecting parts make intermittent sounds due to thermal expansion and contraction, it may present a "click" sound with a certain period; in certain specific vibration modes, a periodic and repetitive sound rhythm similar to "buzzing" may be produced; if the internal circuit has intermittent poor contact, a regular "pop" sound may be produced , the rhythm interval is relatively stable, etc.; abnormal sound characteristics: such as noise, sudden sound, etc.; timbre characteristics: the unique quality of the sound is related to the state of the photovoltaic panel, such as: when the photovoltaic panel is in good condition, the sound is relatively crisp, pure, and the timbre is bright; when the packaging material is aged and slight cracks appear, the sound may have some subtle "hissing" sound, and the timbre is slightly sharp; if the battery cell has problems such as hidden cracks, the sound emitted under certain circumstances may have some dull "buzzing" feeling, and the timbre is heavy; when the connection line is aged and causes poor contact, it may emit an intermittent "crackling" sound with a messy timbre; when the cooling system of the photovoltaic panel is aged and works abnormally, the sound emitted may have a "whooshing" sound Noisy, muddy sound; if the photovoltaic panel's frame and other structural parts are corroded or deformed, the sound it emits may have a "creaking" and distorted feeling; Complexity of soundprint: reflects the complexity and diversity of sound signals. For example, when there are multiple aging problems of photovoltaic panels at the same time, such as cracking of packaging materials, hidden cracks in solar cells, and poor contact of connecting lines, the soundprint may be very complex, including multiple sounds with different characteristics appearing at the same time, such as "hissing", "buzzing", "crackling" and other sounds intertwined; or when photovoltaic panels are in a complex external environment, such as wind, temperature changes and other factors acting at the same time, the sound signal will become complex and diverse, including a mixture of various frequencies, amplitudes and rhythms;When different types of faults or abnormalities occur in different parts of the photovoltaic panel, its soundprint may have a combination of sounds with different rhythms and timbres, making the overall soundprint more complex; in addition, during the operation of the photovoltaic panel, some random interference factors may also increase the complexity of the soundprint, such as the irregular sounds produced by occasional object collisions or small animal touches. ;
[0067] The precise algorithm adopts mature algorithms in the existing technology, including but not limited to statistical analysis algorithms, machine learning algorithms, time series prediction algorithms, optimization algorithms, and reliability assessment algorithms; among them, statistical analysis algorithms such as variance analysis and regression analysis are used to find the correlation between photovoltaic panel performance and various factors; machine learning algorithms such as random forests and gradient boosting algorithms can handle complex nonlinear relationships; time series prediction algorithms are used to predict future aging trends based on historical data; optimization algorithms help determine the optimal parameters of the model; reliability assessment algorithms, such as Weibull distribution and other related algorithms, are used to evaluate the changes in the reliability of photovoltaic panels over time.
[0068] S5: input the voiceprint data collected in real time into the aging model, and the aging model predicts the aging degree and remaining life of the component based on the voiceprint characteristics;
[0069] S6, outputting the aging degree and remaining life predicted in S5 to the user in an intuitive manner; such as displaying on a device interface or generating a report, the device adopts a mobile phone, tablet, computer, monitor, or other device commonly used in the prior art that can display an interface;
[0070] S7, when the aging degree of the PV modules exceeds the set threshold, the system sends out a warning signal to remind the user to take corresponding measures, such as maintenance or replacement, to ensure that the PV modules can continue to work stably and efficiently, and avoid possible greater problems and losses; the setting of the warning signal always monitors the aging degree of the PV modules, and reminds the user in time to ensure the normal operation of the PV system. Embodiment 2
[0071] This embodiment specifically describes S1. Specifically, S1 includes:
[0072] S11: The photovoltaic module is imaged as a whole through image acquisition devices such as digital cameras, industrial cameras, scanners, and cameras, and the acquired images are analyzed for flatness. The position with poor flatness is set as a voiceprint acquisition module, and each voiceprint acquisition module is preset with multiple voiceprint sensors to form a sensor network; the standard definition of poor flatness in visual measurement is adopted in the following ways: 1. A fixed maximum allowable fluctuation height or depth difference is set, such as, if the height fluctuation of the surface exceeds a certain number of millimeters, it is considered to be poor flatness; 2. Judging based on the continuous flatness of the surface, if there are obvious discontinuous, wavy or sudden changes, it can be identified as poor; 3. Measured from the perspective of overall visual uniformity, when a certain area is visually obviously inconsistent and uneven with most of the surrounding areas, it can be used as a judgment standard; preferably, multiple tests can be performed on the position with poor flatness. If it is considered uneven after multiple tests, it can be confirmed that the position has poor flatness, so as to prevent misjudgment and improve the accuracy of the selection of the installation position of the voiceprint sensor.
[0073] S12: Setting the initial sampling frequency and the trigger threshold of the acoustic wave characteristic information of the soundprint sensor in each soundprint acquisition module; the initial sampling frequency can be selected according to actual needs and application scenarios, such as being set to 5000Hz, which can better capture most of the sound features; the acoustic wave characteristic information includes vibration information, abnormal sound, sound intensity and other information, wherein the vibration information includes the frequency and amplitude of vibration, etc., which is used to reflect the health status of the photovoltaic module; abnormal sound, when cracks, looseness and other conditions occur in the photovoltaic module, specific abnormal sound may be generated, which can timely discover potential problems; sound intensity, which is used to reflect the severity of the vibration of the photovoltaic module, etc.; the thresholds of the above-mentioned characteristics are specifically set, such as the trigger threshold of the amplitude is set to 0.5 volts, and the relevant data processing and analysis will be started when the detected sound amplitude exceeds this value, so as to avoid excessive processing of some small and insignificant sound fluctuations; the specific values need to be comprehensively considered and adjusted according to multiple factors such as the characteristics of the photovoltaic module, the monitoring environment and the requirements for monitoring accuracy, so as to achieve the best monitoring effect, and this application does not make specific restrictions here;
[0074] S13: The voiceprint sensor collects the voiceprint information of the photovoltaic module at regular intervals and transmits the collected information. The interval time is set according to the specific situation. The voiceprint sensor collects the voiceprint information of the photovoltaic module at a fixed interval time, which can detect problems in time and evaluate the life of the photovoltaic module at any time.
[0075] The voiceprint sensor uses the initial sampling frequency to regularly collect information related to the sound wave characteristics of the photovoltaic module, such as vibration frequency, amplitude, sound intensity, abnormal sound, etc. according to the collection node of the voiceprint sensor. Then the voiceprint sensor transmits the collected sound wave characteristic information. When the collection node of the voiceprint sensor fails to collect information related to the sound wave characteristics for 4 times, the voiceprint sensor enters a dormant state and waits for the next collection cycle to start. The timely dormancy of the voiceprint sensor can reduce energy consumption and thus extend its working time.
[0076] S14: when the voiceprint sensors in one or more voiceprint collection modules detect sound wave feature information, the sound wave feature information in each voiceprint collection module is collected to form a sound wave data set; the sound wave data set in each voiceprint collection module is transmitted to the comparison module;
[0077] S15: The comparison module compares the sound wave data set in the voiceprint collection module with the previous data set of the corresponding voiceprint collection module;
[0078] S16: Obtaining the acoustic wave characteristic deviation of the voiceprint collection module according to the difference between the acoustic wave data set and the previous data set; the higher the acoustic wave characteristic deviation in the voiceprint collection module, the more sensitive the monitoring needs to be, and the lower the acoustic wave characteristic deviation in the voiceprint collection module, the slower or less obvious the change needs to be, and an overly dense monitoring layout is not required;
[0079] S17: According to the difference between the acoustic wave feature deviation and the acoustic wave feature trigger threshold, the optimal interval between adjacent voiceprint sensors and the optimal sampling frequency interval of each voiceprint acquisition module are set; specifically, regarding the setting of the optimal interval, when the difference between the acoustic wave feature deviation and the acoustic wave feature trigger threshold is large, it means that more sensitive and detailed monitoring is required. Setting the interval between adjacent voiceprint sensors to be smaller can accurately capture the changes and propagation of acoustic wave features, detect anomalies earlier, and avoid monitoring loopholes; and when the difference is small, it means that the changes are relatively slow or not obvious, and the interval between adjacent voiceprint sensors is appropriately increased, because at this time, an overly dense monitoring layout is not required, which can meet both basic monitoring needs and major monitoring needs, and avoid excessive monitoring loopholes. It saves costs and resources, and also reduces the possibility of misjudgment or omission due to oversensitivity, while ensuring a certain monitoring effect, thereby improving the efficiency and reliability of the system. Regarding the setting of the optimal sampling frequency, when the difference between the acoustic feature deviation and the acoustic feature trigger threshold is small, the acoustic signal is relatively stable and close to expectations, and a relatively low sampling frequency is set. A lower sampling frequency can reduce the amount of data and processing burden while meeting basic needs. When the difference is large, the acoustic signal changes more dramatically or differs greatly from expectations. In order to more accurately capture these changing details, the sampling frequency needs to be increased. A higher sampling frequency can obtain more information, so as to better analyze and deal with such large deviations.
[0080] A more specific implementation plan is:
[0081] A1: Classify the difference and divide it into multiple intervals, such as small difference interval, medium difference interval, medium-large difference interval, large difference interval, etc.; for example, when the sound wave feature uses the frequency of vibration, the small difference interval is greater than or equal to 0 and less than 1000HZ, the medium difference interval is greater than or equal to 1000HZ and less than 2000HZ, the medium-large difference interval is greater than or equal to 2000HZ and less than 3000HZ, and the large difference interval is greater than or equal to 3000HZ; A2: Set the corresponding sampling frequency and / or the distance between adjacent voiceprint sensors according to the classification of the difference, and set the corresponding optimal sampling frequency interval and optimal interval distance for each difference interval category. In this embodiment, in the small difference interval, the sampling frequency is greater than or equal to 1000HZ and less than 2000HZ, and the spacing between adjacent voiceprint sensors is set to 1 meter, which can ensure basic coverage and monitoring accuracy when the deviation is small; the sampling frequency in the medium difference interval is greater than or equal to 2000HZ and less than 3000HZ, and the spacing between adjacent voiceprint sensors is set to 0.8 meters; the sampling frequency in the medium and large difference interval is greater than or equal to 3000HZ and less than 4000HZ, and the spacing between adjacent voiceprint sensors is set to 0.5 meters; the sampling frequency in the large difference interval is greater than or equal to 4000HZ, and the spacing between adjacent voiceprint sensors is set to 0.2 meters;
[0082] Preferably, it also includes A3: real-time adjustment of the distance between adjacent voiceprint sensors in the voiceprint acquisition module and the optimal sampling frequency interval, that is, constantly dynamically adjusting the spacing and sampling frequency between the voiceprint sensors according to the difference between the acoustic wave characteristic deviation calculated in real time and the acoustic wave characteristic trigger threshold. For example, if the current difference is 600, the sampling frequency is greater than or equal to 1000HZ and less than 2000HZ, and they are arranged at a spacing of 1 meter; if the difference becomes 2700, the sampling frequency is greater than or equal to 3000HZ and less than 4000HZ, the spacing between adjacent voiceprint sensors is set to 0.5 meters, and the position and sampling frequency interval of the voiceprint sensors are adjusted in time to facilitate better evaluation of the lifespan; when the voiceprint sensor is movable, an electric slide rail can be installed at the frame of the photovoltaic module, and the voiceprint sensor can be installed on the electric slide rail, thereby facilitating the adjustment of the distance between the voiceprint sensors. The installation position of the electric slide rail needs to consider whether it will reduce the effective illumination area to prevent the reduction of power generation efficiency. The technical personnel in this field can set and select according to the specific situation. It does not belong to the technical problem to be solved by the present application and will not be described in detail here.
[0083] S18: Locate the acoustic wave feature points collected by the collection node of the soundprint sensor at the sampling frequency in S17 to confirm the specific coordinates of the acoustic wave feature points; specifically, determine the position of the vibration point through the four adjacent beacon nodes, and the communication connection status and characteristics between the four beacon nodes; more specifically, when the voiceprint sensor maintains a network connection with the four adjacent beacon nodes, the position of the acoustic wave feature point can be calculated based on the time, strength, angle and other information of the signal transmission, combined with the location information of the known beacon nodes, through trilateral measurement, angle measurement, arrival time difference (TDOA) algorithm and the like;
[0084] S19: input the difference between the acoustic wave feature deviation and the acoustic wave feature trigger threshold, and the coordinates of the acoustic wave feature into the neural network model, calculate and output the optimal placement position of the voiceprint sensor, and adjust the position between the sensors according to the optimal placement position;
[0085] After providing the difference between the acoustic feature deviation and the acoustic feature information trigger threshold and the coordinates of the acoustic feature as input features to the neural network model, the neural network can learn the potential relationship pattern between these differences and the reasonable arrangement of the voiceprint sensor; the fully trained neural network can output the specific location of the voiceprint sensor based on the input difference information. Assuming that the neural network model has multiple hidden layers, the output of the output layer is the optimal placement position G (X, Y);
[0086] in:
[0087] G(X,Y) = f(∑(w 1i * (sound wave characteristic deviation - sound wave characteristic trigger threshold) + w 2i * X coordinate of the acoustic wave feature point + w 3i * Y coordinate of the acoustic wave feature point) + b 1i );
[0088] Among them, i represents the node of the hidden layer, i = 1-n; f is the activation function, w 1i 、w 2i 、w 3i is the weight of different neurons, b 1i is the bias term.
[0089] Through the arrangement of the voiceprint sensors in this embodiment, it is possible to ensure that the voiceprint signals of the photovoltaic components are accurately captured.
[0090] This application uses soundprints to evaluate the life of photovoltaic panels, which has the following advantages:
[0091] 1. No additional detection components are required to contact the photovoltaic panels, and the life span can be evaluated without affecting the normal working state of the photovoltaic panels, thus reducing interference to the photovoltaic panels;
[0092] 2. Real-time monitoring of the sound pattern changes of photovoltaic panels during operation, timely detection of potential problems and abnormal conditions; facilitating timely adoption of corresponding maintenance and replacement measures, and improving the reliability and economic benefits of the photovoltaic system;
[0093] 3. It can detect early signs through subtle changes in the soundprint before the photovoltaic panels show obvious performance degradation or failure, thus buying time for taking maintenance measures in advance;
[0094] 4. The entire photovoltaic panel array can be tested at the same time to obtain more comprehensive life information, rather than being limited to local testing, to ensure the comprehensiveness of photovoltaic panel life assessment;
[0095] 5. The detection device is simple, easy to install and operate, thus ensuring a low cost of use and having certain advantages in operating costs.
[0096] The devices, connection relationships, etc. not specifically described above all belong to the prior art and will not be described in detail in the present invention.
[0097] The connection relationships in the mechanical structure of the present invention are all well-known connection methods in the prior art, such as threaded connection, bonding, welding and other connection methods. The technicians can choose according to the specific situation. Simple changes in the connection methods all fall within the protection scope of the present invention.
[0098] The preferred embodiments of the present application are described in detail above in conjunction with the accompanying drawings; however, the present application is not limited to the specific details in the above embodiments. Within the technical concept of the present application, a variety of simple modifications can be made to the technical solution of the present application, and these simple modifications all fall within the protection scope of the present application.
[0099] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the various possible combinations of this application will not be described separately.
[0100] In addition, the various implementation modes of the present application may be arbitrarily combined, and as long as they do not violate the concept of the present application, they should also be regarded as the contents disclosed in the present application.
Claims
1. A method for evaluating the service life of a photovoltaic module, characterized in that: The steps include: S1: Arrange the soundprint sensor on the photovoltaic module; S11: Collect images of the entire photovoltaic module, analyze the flatness of the collected images, set the locations with poor flatness as voiceprint collection modules, and preset multiple voiceprint sensors in each voiceprint collection module to form a sensor network; S12: Setting the initial sampling frequency and the sound wave characteristic information triggering threshold of the voiceprint sensor in each voiceprint collection module; S13: The voiceprint sensor collects the voiceprint information of the photovoltaic module at a regular interval and transmits the collected information; S14: when the voiceprint sensors in one or more voiceprint collection modules detect sound wave feature information, the sound wave feature information in each voiceprint collection module is collected to form a sound wave data set; the sound wave data set in each voiceprint collection module is transmitted to the comparison module; S15: The comparison module compares the sound wave data set in the voiceprint collection module with the previous data set of the corresponding voiceprint collection module; S16: Obtaining the acoustic wave feature deviation of the voiceprint collection module according to the difference between the acoustic wave data set and the previous data set; S17: setting the optimal interval between adjacent voiceprint sensors and the optimal sampling frequency interval of each voiceprint collection module according to the difference between the acoustic wave feature deviation and the acoustic wave feature information trigger threshold; S18: locating the sound wave feature points collected by the voiceprint collection module at the sampling frequency in S17; S19: Input the difference between the acoustic wave feature deviation and the acoustic wave feature information trigger threshold, and the coordinates of the acoustic wave feature points into the neural network model, and calculate and output the optimal placement position of the voiceprint sensor; S2: The voiceprint sensor collects the voiceprints generated by the photovoltaic modules at different aging stages, converts the collected voiceprints into electrical signals, and sends them to the data processing unit; S3: The data processing unit analyzes the received signal and extracts characteristic parameters related to the aging of the photovoltaic components; S4: Establish an aging model by training the voiceprint data generated by a large number of aging photovoltaic modules; S5: The voiceprint data collected in real time is input into the aging model, and the aging model predicts the aging degree and remaining life of the component based on the voiceprint characteristics.
2. A method for evaluating the service life of a photovoltaic module according to claim 1, characterized in that: The method further includes S6, which outputs the aging degree and remaining lifespan predicted in S5 to the user in an intuitive manner.
3. A method for evaluating the service life of a photovoltaic module according to claim 2, characterized in that: It also includes S7. When the aging degree of the photovoltaic module exceeds the set threshold, the system will send out a warning signal to remind the user to take corresponding measures.
4. A method for evaluating the service life of a photovoltaic module according to claim 1, characterized in that: The sound wave characteristic information in S12 includes vibration information, abnormal sound, and sound intensity information, wherein the vibration information includes the frequency and amplitude of the vibration.
5. A method for evaluating the service life of a photovoltaic module according to claim 1, characterized in that: S17 includes: A1: Classify the differences into small difference interval, medium difference interval, medium-large difference interval, and large difference interval; A2: Set the corresponding sampling frequency and the distance between adjacent voiceprint sensors according to the classification of the difference.
6. A method for evaluating the service life of a photovoltaic module according to claim 1, characterized in that: The coordinates of the optimal placement of the voiceprint sensor in the output layer of the neural network model in S19 are G (X, Y): ; Among them, i represents the node of the hidden layer, i = 1-n; f is the activation function, w 1i 、w 2i 、w 3i is the weight of different neurons, b 1i is the bias term.
7. A method for evaluating the service life of a photovoltaic module according to claim 1, characterized in that: The aging stages in S2 include an initial aging stage, a mid-term aging stage, a severe aging stage, and a near-failure stage.
8. A method for evaluating the service life of a photovoltaic module according to claim 1, characterized in that: In S3, the characteristic parameters include sound frequency, sound intensity, sound duration, abnormal sound, sound attenuation, sound resonance, and sound amplitude.
9. A method for evaluating the service life of a photovoltaic module according to claim 1, characterized in that: The voiceprint data generated by the photovoltaic components in S4 include: vibration frequency, vibration amplitude, vibration duration, vibration periodicity, vibration pattern, sound frequency distribution, main frequency characteristics, frequency spectrum characteristics, amplitude change pattern, sound rhythm and rhythm, abnormal sound characteristics, timbre characteristics, and complexity of the voiceprint.
