An intelligent heat dissipation control method and control system for a photovoltaic inverter

Abstract

The present invention discloses an intelligent heat dissipation control method and control system for a photovoltaic inverter, and relates to the technical field of heat dissipation control for photovoltaic inverters. The method comprises the following steps: using a temperature sensor to collect component temperatures and the internal ambient temperature of the photovoltaic inverter, and simultaneously collecting internal thermal imaging images, and performing a first screening process; then establishing a thermal imaging temperature correction model; performing temperature correction processing on the thermal imaging images inside the photovoltaic inverter to obtain corrected thermal imaging temperature data; and then dynamically controlling the heat dissipation fan of the photovoltaic inverter. The present invention is used to solve the problem that the existing photovoltaic inverter heat dissipation control technology cannot obtain accurate overall temperature information inside the photovoltaic inverter through thermal imaging images when performing heat dissipation control by measuring the internal temperature of the photovoltaic inverter, and uses the temperature sensor to perform error correction on the thermal imaging images to obtain accurate overall temperature information inside the photovoltaic inverter and then perform heat dissipation control.

Description

Technical Field

[0001] The present invention relates to the technical field of heat dissipation control of photovoltaic inverters, and in particular to an intelligent heat dissipation control method and control system for photovoltaic inverters. Background Art

[0002] Photovoltaic inverter heat dissipation control technology refers to the management and regulation of the heat generation, conduction and dissipation process inside the photovoltaic inverter through various methods and equipment to ensure that the photovoltaic inverter operates stably within an appropriate temperature range, thereby ensuring the working efficiency, reliability and service life of the photovoltaic inverter.

[0003] Existing photovoltaic inverter heat dissipation control technology controls the heat dissipation device according to the temperature changes inside the photovoltaic inverter. When measuring the internal temperature of the photovoltaic inverter, a temperature sensor installed inside the photovoltaic inverter is often used for measurement. However, the temperature sensor can only be installed in a limited number of key positions, such as key components such as power semiconductor devices, transformers and inductors; the temperature conditions of many other components and areas inside the inverter cannot be directly known; for example, the patent application with publication number CN117529070A discloses a heat dissipation control method and system for stable operation of a photovoltaic inverter. This solution is to measure the temperature of several points through a temperature sensor and then perform heat dissipation control; however, the temperature of key components is normal, and other unmonitored areas may also have hot spots; for example, a local temperature rise may occur in a corner of the inverter due to poor ventilation or electromagnetic interference. , hot spots appear. These hot spots are not monitored by temperature sensors. Problems can only be discovered when heat is transferred to key components equipped with temperature sensors or when the fault has caused the performance of the inverter to decline. This makes the fault warning delayed and increases the risk of equipment damage. The overall temperature information inside the photovoltaic inverter can be obtained through thermal imaging equipment, but thermal imaging equipment is easily disturbed by the ambient temperature, and the thermal imager itself has a certain measurement error. When measuring the internal temperature of the inverter, different temperature distributions are measured under different ambient temperatures and component temperatures, resulting in deviations in heat dissipation control. Therefore, the existing photovoltaic inverter heat dissipation control technology cannot use thermal imaging images to perform error correction on the thermal imaging images using temperature sensors when measuring the internal temperature of the photovoltaic inverter for heat dissipation control, and then obtain accurate overall temperature information inside the photovoltaic inverter for heat dissipation control. Summary of the Invention

[0004] The present invention aims to solve, at least to a certain extent, one of the technical problems in the prior art. It uses a temperature sensor to collect the component temperatures and the internal ambient temperature of a photovoltaic inverter, and simultaneously collects internal thermal imaging images, performs a first screening process, and then establishes a thermal imaging temperature correction model; performs temperature correction processing on the thermal imaging images inside the photovoltaic inverter to obtain corrected thermal imaging temperature data; and then dynamically controls the cooling fan of the photovoltaic inverter. This solves the problem that the existing photovoltaic inverter heat dissipation control technology cannot obtain accurate overall temperature information inside the photovoltaic inverter through thermal imaging images when performing heat dissipation control by measuring the internal temperature of the photovoltaic inverter, and uses a temperature sensor to perform error correction on the thermal imaging images to obtain accurate overall temperature information inside the photovoltaic inverter and then perform heat dissipation control.

[0005] To achieve the above objectives, in a first aspect, the present application provides an intelligent heat dissipation control method for a photovoltaic inverter, comprising the following steps:

[0006] Using a temperature sensor to collect component temperatures and internal ambient temperature of the photovoltaic inverter, while also collecting a thermal imaging image of the photovoltaic inverter, and performing a first screening process to obtain sensor temperature data and thermal imaging temperature data;

[0007] Establish a thermal imaging temperature correction model based on sensor temperature data and thermal imaging temperature data;

[0008] Performing temperature correction processing on the thermal imaging image inside the photovoltaic inverter based on the thermal imaging temperature correction model and data from the temperature sensor inside the photovoltaic inverter to obtain corrected thermal imaging temperature data;

[0009] The cooling fan of the photovoltaic inverter is dynamically controlled based on the corrected thermal imaging temperature data and the data of the temperature sensor inside the photovoltaic inverter.

[0010] Furthermore, using a temperature sensor to collect component temperatures and internal ambient temperature of the photovoltaic inverter, and simultaneously collecting a thermal imaging image of the interior of the photovoltaic inverter includes the following sub-steps:

[0011] Using a temperature sensor inside the photovoltaic inverter, continuously collect the component temperature of a corresponding component inside the photovoltaic inverter at a first time interval, and record the collection time and collection location, marking it as first component temperature data; at the same time, continuously collect the ambient temperature inside the photovoltaic inverter at a first time interval, and record the collection time, marking it as first internal temperature data, and the first time interval is t1;

[0012] Using a thermal imager, while maintaining a relative position between the thermal imager and the photovoltaic inverter in space, continuously collecting thermal imaging images of the entire interior of the photovoltaic inverter at a first time interval, marking the images as first thermal imaging data;

[0013] A position inside the photovoltaic inverter where the temperature can be collected by the temperature sensor and obtained by the thermal imaging image is marked as a first reference position.

[0014] Furthermore, performing a first screening process to obtain sensor temperature data and thermal imaging temperature data includes the following sub-steps:

[0015] Based on the first thermal imaging data and the first component temperature data, respectively extracting temperatures corresponding to the first thermal imaging data and temperatures corresponding to the first component temperature data at all acquisition moments and all first reference positions, and marking them in sequence as second thermal imaging data and second component temperature data, respectively;

[0016] Performing data cleaning on the second thermal imaging data, the second component temperature data, and the first internal temperature data, respectively, to obtain thermal imaging temperature data, the third component temperature data, and the second internal temperature data;

[0017] The data cleaning process includes: sorting the corresponding data according to the corresponding first reference position and the acquisition time, recording them as a time data series, calculating the mean value and standard deviation of the time data series, recording them as E0 and E1 respectively in order, and for any data in the time data series, recording them as Ei, where i represents the sequence number in the time data series. If Ei satisfies |Ei-E0|>3*E1, then Ei is marked as abnormal data; if Ei does not satisfy |Ei-E0|>3*E1, then Ei is marked as normal data, and the marking process is repeated for all data in the time data series.

[0018] The third component temperature data and the second internal temperature data are combined and stored according to the acquisition time, and marked as sensor temperature data.

[0019] Furthermore, establishing a thermal imaging temperature correction model based on the sensor temperature data and the thermal imaging temperature data includes the following sub-steps:

[0020] Merging the thermal imaging temperature data and the sensor temperature data according to the acquisition time and the first reference position of the acquisition, and recording the result as first sample data;

[0021] For the first sample data, the first sample data at the same acquisition time and the same first reference position are combined into a sample vector, denoted as M = [m1, m2, m3], where m1 represents the temperature of the first reference position measured by the temperature sensor, m2 represents the temperature of the first reference position corresponding to m1 displayed by the thermal imaging image, and m3 represents the ambient temperature inside the photovoltaic inverter. If there are abnormal data in m1, m2, and m3 in the sample vector M, they are removed. After completion, the second sample data is obtained.

[0022] Normalizing the second sample data according to the component types of the sample vectors, scaling the sizes of all components of all sample vectors in the second sample data to [0, 1], and obtaining the third sample data after completion;

[0023] The third sample data is divided into a training data set and a test data set in a ratio of 9:1.

[0024] Furthermore, establishing a thermal imaging temperature correction model based on the sensor temperature data and the thermal imaging temperature data further includes the following sub-steps:

[0025] A first temperature correction model is constructed based on a multilayer perceptron. The first temperature correction model includes an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is set to a1, the number of hidden layers is set to a2, the number of neurons in each hidden layer is set to a3, and the number of neurons in the output layer is set to a4.

[0026] The first temperature correction model is trained using the training data set, and a thermal imaging temperature correction model is obtained after the training is completed;

[0027] The thermal imaging temperature correction model is tested using the test data set, and the average error AE of the thermal imaging temperature correction model is calculated. The average error calculation formula is as follows: , where Yj is the temperature of the first reference position measured by the temperature sensor, Xj is the temperature of the first reference position output by the thermal imaging temperature correction model, and n is the number of data input to the thermal imaging temperature correction model.

[0028] Furthermore, performing temperature correction processing on the thermal imaging image inside the photovoltaic inverter based on the thermal imaging temperature correction model and data from the temperature sensor inside the photovoltaic inverter to obtain corrected thermal imaging temperature data includes the following sub-steps:

[0029] Collecting thermal imaging images of the entire interior of the photovoltaic inverter at a first time interval, and simultaneously collecting component temperatures obtained by a temperature sensor inside the photovoltaic inverter and the internal ambient temperature;

[0030] The temperature values corresponding to all pixels in the collected thermal imaging image are extracted and recorded as the original thermal imaging temperature information. The original thermal imaging temperature information is normalized with the corresponding ambient temperature inside the photovoltaic inverter and input into the thermal imaging temperature correction model to obtain the first corrected temperature information.

[0031] Furthermore, performing temperature correction processing on the thermal imaging image inside the photovoltaic inverter based on the thermal imaging temperature correction model and the data of the temperature sensor inside the photovoltaic inverter to obtain the corrected thermal imaging temperature data further includes the following sub-steps:

[0032] Based on the first corrected temperature information and the component temperature obtained by the temperature sensor at the same time, a first error between the first corrected temperature information at all first reference positions and the temperature obtained by the temperature sensor is calculated using a first error calculation formula. The first error calculation formula is as follows: , where AF represents the first error, U represents the temperature of the first reference position measured by the temperature sensor, and V is the temperature of the first reference position corresponding to the first corrected temperature information; and all the first errors are averaged to obtain the second error AG;

[0033] Then calculate the temperature correction error using the correction error calculation formula. The correction error calculation formula is as follows: , where AH is the temperature correction error, q1 and q2 are weight coefficients, q1+q2=1, and the value range of q1 and q2 is [0, 1];

[0034] Based on the temperature correction error, the temperature correction formula is used to correct the error of the first correction temperature information. The temperature correction formula is as follows: , where TY is the temperature value in the first corrected temperature information, and TX is the temperature value after error correction; after completion, the second corrected temperature information is obtained, which is marked as the corrected thermal imaging temperature data.

[0035] Furthermore, dynamically controlling the cooling fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and the data of the temperature sensor inside the photovoltaic inverter includes the following sub-steps:

[0036] Merge the corrected thermal imaging temperature data at the same moment with the data from the temperature sensor inside the photovoltaic inverter, and replace the temperature value corresponding to the first reference position in the corrected thermal imaging temperature data with the temperature value collected by the corresponding temperature sensor, to obtain the internal temperature distribution data.

[0037] Set the safe temperature range, recorded as [B1, C1], and set the optimal temperature range according to the safe temperature range, recorded as [B2, C2], where B1<B2,C2> C1.

[0038] Furthermore, dynamically controlling the cooling fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and the data of the temperature sensor inside the photovoltaic inverter further includes the following sub-steps:

[0039] Extract the maximum value of the internal temperature distribution data, record it as the internal maximum temperature TM, obtain the internal maximum temperature TM at a first time interval, record it as the internal maximum temperature sequence, perform linear fitting on the internal maximum temperature sequence, obtain the internal maximum temperature sequence of the first time length in the future, record it as the predicted maximum temperature sequence, record any temperature value in the predicted maximum temperature sequence as Tk, where k represents the sequence number in the predicted maximum temperature sequence, and the first time length is t2;

[0040] If the internal maximum temperature TM at the current moment satisfies TM ∈ [B2, C2] and any Tk ∈ [B2, C2], then the rotation speed of the cooling fan is not adjusted;

[0041] If the internal maximum temperature TM at the current moment satisfies TM ∈ [B2, C2] and there exists Tk > C2, then increase the rotation speed of the cooling fan;

[0042] If the internal maximum temperature TM at the current moment satisfies TM ∈ [B2, C2] and there exists Tk < B2, then decrease the rotation speed of the cooling fan;

[0043] If the internal maximum temperature TM at the current moment satisfies TM ∈ (C2, C1] and any Tk < C1, then increase the rotation speed of the cooling fan;

[0044] If the internal maximum temperature TM at the current moment satisfies TM ∈ (C2, C1] and any Tk > C1 or TM > C1, then increase the rotation speed of the cooling fan to the maximum rotation speed and issue a warning signal;

[0045] If the internal maximum temperature TM at the current moment satisfies TM ∈ [B1, B2) and any Tk > B1, then decrease the rotation speed of the cooling fan;

[0046] If the internal maximum temperature TM at the current moment satisfies TM ∈ [B1, B2) and any Tk < C1 or TM < B1, then decrease the rotation speed of the cooling fan to the minimum rotation speed and issue a warning signal.

[0047] In a second aspect, the present application provides an intelligent cooling control system for a photovoltaic inverter, including a temperature acquisition module, a model establishment module, a temperature correction module, and a cooling control module;

[0048] The temperature acquisition module includes a data acquisition unit and a data screening unit. The data acquisition unit uses temperature sensors to acquire the component temperature and the internal environmental temperature inside the photovoltaic inverter, and simultaneously acquires the thermal imaging image inside the photovoltaic inverter. The data screening unit is used for the first screening process to obtain sensor temperature data and thermal imaging temperature data;

[0049] The model establishment module establishes a thermal imaging temperature correction model based on the sensor temperature data and the thermal imaging temperature data;

[0050] The temperature correction module performs temperature correction processing on the thermal imaging image inside the photovoltaic inverter based on the thermal imaging temperature correction model and the data of the temperature sensors inside the photovoltaic inverter to obtain corrected thermal imaging temperature data;

[0051] The heat dissipation control module dynamically controls the heat dissipation fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and the data of the temperature sensor inside the photovoltaic inverter.

[0052] Beneficial effects of the present invention: The present invention uses a temperature sensor to collect the component temperatures and the internal ambient temperature of the photovoltaic inverter, and simultaneously collects a thermal imaging image of the photovoltaic inverter, and performs a first screening process to obtain sensor temperature data and thermal imaging temperature data; a thermal imaging temperature correction model is established based on the sensor temperature data and the thermal imaging temperature data; a temperature correction process is performed on the thermal imaging image of the photovoltaic inverter based on the thermal imaging temperature correction model and the data of the temperature sensor inside the photovoltaic inverter to obtain corrected thermal imaging temperature data; a cooling fan of the photovoltaic inverter is dynamically controlled based on the corrected thermal imaging temperature data and the data of the temperature sensor inside the photovoltaic inverter; accurate overall temperature information inside the photovoltaic inverter can be obtained through the thermal imaging image and error correction of the thermal imaging image using the temperature sensor, and then heat dissipation control is performed to ensure that no local hot spots appear inside the photovoltaic inverter;

[0053] The present invention establishes a thermal imaging temperature correction model and performs error correction in combination with the internal ambient temperature. The advantage is that the internal ambient temperature will change the deviation between the temperature measured by the thermal imaging image and the actual temperature. By taking the internal ambient temperature as input, the model can learn these complex relationships, thereby more accurately correcting the temperature measurement error of the thermal imager, improving the accuracy of the final temperature data, and improving the versatility and adaptability of the model under different ambient temperature conditions. By performing two error corrections on the thermal imaging image, corrected thermal imaging temperature data is obtained. The advantage is that the first time the error of the thermal imaging image caused by environmental interference is reduced, making the temperature data closer to the true value, and the second time the temperature data is calibrated again, further eliminating possible deviations, making the final temperature data more accurate, and providing a more reliable basis for subsequent heat dissipation control. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a functional block diagram of the system of the present invention;

[0055] Figure 2 is a flow chart of the steps of the method of the present invention;

[0056] Figure 3 This is a flow chart of the temperature correction process of the present invention;

[0057] Figure 4 This is a flow chart of the dynamic control of the cooling fan of the present invention;

[0058] Figure 5 Schematic diagram of the structure of the electronic device of the present invention. DETAILED DESCRIPTION

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0060] Example 1, please refer to Figure 1 As shown, the present application provides an intelligent heat dissipation control system for a photovoltaic inverter, including a temperature acquisition module, a model building module, a temperature correction module and a heat dissipation control module;

[0061] The temperature acquisition module includes a data acquisition unit and a data screening unit. The data acquisition unit uses a temperature sensor to collect the temperature of components inside the photovoltaic inverter and the internal ambient temperature, and simultaneously collects a thermal imaging image inside the photovoltaic inverter. The data screening unit is used to perform a first screening process to obtain sensor temperature data and thermal imaging temperature data.

[0062] The data acquisition unit is configured with a data acquisition strategy, which includes: using the temperature sensor inside the photovoltaic inverter to continuously acquire the component temperature of the corresponding component inside the photovoltaic inverter at a first time interval, and recording the acquisition time and the acquisition position, which are marked as the first component temperature data; at the same time, continuously acquiring the ambient temperature inside the photovoltaic inverter at a first time interval, and recording the acquisition time, which are marked as the first internal temperature data, the first time interval is t1, and in this embodiment, the first time interval t1 is 3 seconds; that is, the temperature is acquired once every 3 seconds.

[0063] Using a thermal imager, while maintaining a relative position between the thermal imager and the photovoltaic inverter in space, continuously collecting thermal imaging images of the entire interior of the photovoltaic inverter at a first time interval, marking the images as first thermal imaging data;

[0064] Mark the locations inside the photovoltaic inverter where the temperature can be measured by both the temperature sensor and the thermal imager as the first reference location. For example, some temperature sensors inside the photovoltaic inverter are located on the surface of the circuit board, and the temperature at these locations can also be measured by the thermal imager, while some temperature sensors are located inside the components and cannot be measured by the imager.

[0065] The data screening unit is configured with a data screening strategy, which includes: based on the first thermal imaging data and the first component temperature data, extracting the temperature corresponding to the first thermal imaging data and the temperature corresponding to the first component temperature data at all acquisition times and all first reference positions, and marking them in sequence as second thermal imaging data and second component temperature data, respectively; that is, only the temperature sensor data, the thermal imaging image, and the internal environment temperature at the first reference position at the same acquisition time are required;

[0066] Performing data cleaning on the second thermal imaging data, the second component temperature data, and the first internal temperature data, respectively, to obtain thermal imaging temperature data, the third component temperature data, and the second internal temperature data;

[0067] The data cleaning process includes: sorting the corresponding data according to the corresponding first reference position and the acquisition time, recording it as a time data series, calculating the mean value and standard deviation of the time data series, recording them as E0 and E1 respectively in order, and for any data in the time data series, recording it as Ei, where i represents the sequence number in the time data series, if Ei satisfies |Ei-E0|>3*E1, then mark Ei as abnormal data, if Ei does not satisfy |Ei-E0|>3*E1, then mark Ei as normal data, and repeat the marking of all data in the time data series; for example, in the time data series of a temperature sensor at a first reference position, E2=20.1℃, and in the time data series, E0=35.0℃ and E1=2.2℃, then |20.1-35.0|>3*2.2, then mark E2 as abnormal data

[0068] Merging and storing the third component temperature data and the second internal temperature data according to the acquisition time, and marking them as sensor temperature data;

[0069] During the specific implementation process, the ambient temperature inside the photovoltaic inverter will affect the measurement results of the thermal imaging image. For example, in a high temperature environment, thermal noise will increase, thereby interfering with the detector's accurate capture of the target object's thermal radiation, resulting in increased noise and reduced resolution in the thermal imaging image. This will in turn affect the accuracy of the thermal imager's temperature measurement. Therefore, collecting the internal ambient temperature for subsequent error correction can more accurately correct the errors in the thermal imaging image and improve the accuracy of the final temperature data.

[0070] The model building module builds a thermal imaging temperature correction model based on the sensor temperature data and the thermal imaging temperature data;

[0071] The model building module is configured with a model building strategy, which includes: merging the thermal imaging temperature data and the sensor temperature data according to the acquisition time and the first reference position of the acquisition, and recording them as first sample data;

[0072] For the first sample data, the first sample data at the same acquisition time and the same first reference position are combined into a sample vector, denoted as M = [m1, m2, m3], where m1 represents the temperature of the first reference position measured by the temperature sensor, m2 represents the temperature of the first reference position corresponding to m1 displayed by the thermal imaging image, and m3 represents the ambient temperature inside the photovoltaic inverter; if there are abnormal data in m1, m2, and m3 in the sample vector M, they are removed; after completion, the second sample data is obtained; abnormal data will affect the results of subsequent model training, so the sample vector containing abnormal data needs to be removed;

[0073] Normalizing the second sample data according to the component types of the sample vectors, i.e., normalizing the sample vectors m1, m2, and m3 respectively; scaling the sizes of all components of all sample vectors in the second sample data to [0, 1], and obtaining the third sample data after completion;

[0074] The third sample data is divided into a training data set and a test data set in a ratio of 9:1;

[0075] A first temperature correction model is constructed based on a multi-layer perceptron. The first temperature correction model includes an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is set to a1, the number of hidden layers is set to a2, the number of neurons in each hidden layer is set to a3, and the number of neurons in the output layer is set to a4. In this embodiment, a1=2, which is used to input the temperature of the first reference position displayed by the thermal imaging image and the ambient temperature inside the photovoltaic inverter, that is, sample vectors m2 and m3; a2=2; a3=24; and a4=1, which is used to output the temperature after error correction of m2, with the temperature of the first reference position measured by the temperature sensor as a reference, that is, m1.

[0076] The first temperature correction model is trained using the training data set, and a thermal imaging temperature correction model is obtained after the training is completed;

[0077] The thermal imaging temperature correction model is tested using the test data set, and the average error AE of the thermal imaging temperature correction model is calculated. The average error calculation formula is as follows: , where Yj is the temperature of the first reference position measured by the temperature sensor, Xj is the temperature of the first reference position output by the thermal imaging temperature correction model, and n is the number of data input to the thermal imaging temperature correction model. If the average error AE of the thermal imaging temperature correction model input is too large, for example, AE>0.2, secondary training can be performed using the training data set until the average error of the thermal imaging temperature correction model reaches a reasonable range.

[0078] During the specific implementation process, the number of hidden layers a2 and the number of neurons in each hidden layer a3 can be increased or decreased according to the actual application scenario. Generally, you can try it from 1 to 3 layers, starting with a smaller number of neurons and gradually increasing the number of neurons; but it should not be too large. Too many hidden layers and a large number of neurons will significantly increase the amount of calculation and the number of parameters of the model, resulting in large consumption of computing resources and low computing efficiency. It should not be too small either. If it is too small, it will not be able to fully learn the complex nonlinear relationship between input data and output data; in the thermal imaging error correction model, it may not be possible to accurately correct the error of the thermal imaging image, resulting in the corrected temperature data still having a large deviation from the actual temperature, and unable to provide a reliable basis for heat dissipation control.

[0079] The temperature correction module performs temperature correction processing on the thermal imaging image inside the photovoltaic inverter based on the thermal imaging temperature correction model and the data of the temperature sensor inside the photovoltaic inverter to obtain corrected thermal imaging temperature data;

[0080] The temperature correction module is configured with a temperature correction strategy, which includes: collecting a thermal imaging image of the entire interior of the photovoltaic inverter at a first time interval, and simultaneously collecting component temperatures obtained by a temperature sensor inside the photovoltaic inverter and an internal ambient temperature;

[0081] See also Figure 3 As shown, the temperature corresponding to all pixels in the collected thermal imaging image is extracted and recorded as the original thermal imaging temperature information. The original thermal imaging temperature information is normalized with the corresponding ambient temperature inside the photovoltaic inverter and input into the thermal imaging temperature correction model to obtain the first corrected temperature information; that is, the ambient temperature and the temperature corresponding to the pixel at the same time are input into the thermal imaging temperature correction model to obtain the error-corrected temperature corresponding to the pixel; the first corrected temperature is obtained by using the ambient temperature and thermal imaging image data based on the thermal imaging error correction model, which preliminarily reduces the error caused by the ambient temperature interference of the thermal imager and makes the temperature data closer to the true value;

[0082] Based on the first corrected temperature information and the component temperature obtained by the temperature sensor at the same time, a first error between the first corrected temperature information at all first reference positions and the temperature obtained by the temperature sensor is calculated using a first error calculation formula. The first error calculation formula is as follows: , where AF represents the first error, U represents the temperature of the first reference position measured by the temperature sensor, and V is the temperature of the first reference position corresponding to the first corrected temperature information. For example, if the temperature measured by the temperature sensor at a first reference position is 35.2°C, while the temperature in the first corrected temperature information is 34.8°C, then the first error of the first reference position is 35.2-34.8=0.4. The second error AG is obtained by averaging all the first errors, i.e., averaging the first errors of all the first reference positions.

[0083] Then calculate the temperature correction error using the correction error calculation formula. The correction error calculation formula is as follows: , where AH is the temperature correction error, q1 and q2 are weight coefficients, q1+q2=1, and the value range of q1 and q2 is [0, 1]. In this embodiment, q1=q2=0.5. q1 and q2 can be set according to the actual application scenario; for example, AG=-0.5, AE=-0.4, then AH=0.5*(-0.5)+0.5*(-0.4)=-0.45;

[0084] Based on the temperature correction error, the temperature correction formula is used to correct the error of the first correction temperature information. The temperature correction formula is as follows: , where TY is the temperature value in the first corrected temperature information, and TX is the temperature value after error correction; after completion, the second corrected temperature information is obtained, marked as the corrected thermal imaging temperature data; for example, if AH=-0.45, then all temperature values in the first corrected temperature information are added with -0.45; the second time, the difference between the internal temperature sensor measurement value and the first corrected temperature is calculated, and the weighted average is combined with the model average error to obtain the average corrected error, and then the first corrected temperature information is corrected to obtain the second corrected temperature information; this process recalibrates the temperature data, further eliminating possible deviations, making the final temperature data more accurate, and more accurately reflecting the actual temperature conditions inside the photovoltaic inverter, providing a more reliable basis for subsequent heat dissipation control;

[0085] During the specific implementation process, the first correction mainly focuses on the relationship between the thermal imager's measurement error and the ambient temperature, and adjusts the error from a macro perspective; the second correction takes into account the difference between the temperature sensor's measurement value and the first correction temperature; through this two-correction method, the impact of many complex factors on temperature measurement, such as the thermal imager's own characteristics, environmental factors, and differences between different measuring equipment, is comprehensively considered. This can more comprehensively grasp the source of error in the inverter's internal temperature measurement, thereby achieving more accurate error correction and ensuring that the corrected temperature data can truly reflect the complex thermal state inside the inverter.

[0086] The heat dissipation control module dynamically controls the cooling fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and the data of the internal temperature sensors of the photovoltaic inverter;

[0087] The heat dissipation control module is configured with a heat dissipation control strategy, which includes: merging the corrected thermal imaging temperature data and the data of the internal temperature sensors of the photovoltaic inverter at the same moment, and replacing the temperature value corresponding to the first reference position in the corrected thermal imaging temperature data with the temperature value collected by the corresponding temperature sensor. After completion, the internal temperature distribution data is obtained; because the temperature measured by the temperature sensor is more accurate, directly replacing a part can improve the accuracy of subsequent heat dissipation control;

[0088] Set a safety temperature range, denoted as [B1, C1]. According to the safety temperature range, set an optimal temperature range, denoted as [B2, C2], where B1 < B2 and C2 > C1; in this embodiment, the safety temperature range is [-20°C, 90°C], and the optimal temperature range is [0°C, 50°C]. The optimal temperature range refers to the optimal operating temperature range of the photovoltaic inverter; extract the maximum value of the internal temperature distribution data, denoted as the internal highest temperature TM. Obtain the internal highest temperature TM at a first time interval, denoted as the internal highest temperature sequence. Perform a linear fit on the internal highest temperature sequence, that is, establish a relationship between the highest temperature TM and time; obtain the internal highest temperature sequence for the first future time length, denoted as the predicted highest temperature sequence. Denote any temperature value in the predicted highest temperature sequence as Tk, where k represents the serial number in the predicted highest temperature sequence. The first time length is t2. In this embodiment, the first time length t2 is 1 minute, that is, obtain the internal highest temperature for the next 1 minute. The first time interval is 3 seconds, that is, obtain 20 internal highest temperatures within the next 1 minute;

[0089] Please refer to Figure 4 As shown, if the internal highest temperature TM at the current moment satisfies TM ∈ [B2, C2] and any Tk ∈ [B2, C2], then the rotation speed of the cooling fan is not adjusted; that is, the internal highest temperature at the current moment and the internal highest temperature for a period of time in the future are both within the optimal temperature range;

[0090] If the internal highest temperature TM at the current moment satisfies TM ∈ [B2, C2] and there exists Tk > C2, then increase the rotation speed of the cooling fan; that is, the internal highest temperature at the current moment is within the optimal temperature range, but the internal highest temperature for a period of time in the future is greater than the optimal temperature range;

[0091] If the internal highest temperature TM at the current moment satisfies TM ∈ [B2, C2] and there exists Tk < B2, then decrease the rotation speed of the cooling fan; that is, the internal highest temperature at the current moment is within the optimal temperature range, but the internal highest temperature for a period of time in the future is less than the optimal temperature range;

[0092] If the internal maximum temperature TM at the current moment satisfies TM ∈ (C2, C1] and any Tk < C1, then increase the rotational speed of the cooling fan; that is, the internal maximum temperature at the current moment is greater than the optimal temperature range, but within the safe temperature range, and the internal maximum temperature within a certain period in the future is within the safe temperature range;

[0093] If the internal maximum temperature TM at the current moment satisfies TM ∈ (C2, C1] and any Tk > C1 or TM > C1, then increase the rotational speed of the cooling fan to the maximum rotational speed and issue a warning signal; that is, the internal maximum temperature at the current moment is greater than the optimal temperature range, but within the safe temperature range, and the internal maximum temperature within a certain period in the future is greater than the safe temperature range, or the internal maximum temperature at the current moment is greater than the safe temperature range

[0094] If the internal maximum temperature TM at the current moment satisfies TM ∈ [B1, B2) and any Tk > B1, then reduce the rotational speed of the cooling fan; that is, the internal maximum temperature at the current moment is less than the optimal temperature range, but within the safe temperature range, and the internal maximum temperature within a certain period in the future is within the safe temperature range;

[0095] If the internal maximum temperature TM at the current moment satisfies TM ∈ [B1, B2) and any Tk < C1 or TM < B1, then reduce the rotational speed of the cooling fan to the minimum rotational speed and issue a warning signal; that is, the internal maximum temperature at the current moment is less than the optimal temperature range, but within the safe temperature range, and the internal maximum temperature within a certain period in the future is less than the safe temperature range, or the internal maximum temperature at the current moment is less than the safe temperature range;

[0096] In the specific implementation process, use linear fitting to predict the future change trend of the maximum temperature, and adjust the rotational speed of the cooling fan in advance according to the prediction result and the current temperature; this forward-looking control strategy can effectively avoid the occurrence of over-temperature situations, respond to temperature changes in a timely manner, and ensure the stable operation of the photovoltaic inverter.

[0097] Embodiment 2, please refer to Figure 2 As shown, the present application provides an intelligent heat dissipation control method for a photovoltaic inverter, including the following steps:

[0098] Step S1, use a temperature sensor to collect the component temperature and the internal ambient temperature inside the photovoltaic inverter, and at the same time collect the thermal imaging image inside the photovoltaic inverter, and perform a first screening process to obtain sensor temperature data and thermal imaging temperature data; Step S1 includes the following sub-steps:

[0099] Step S101: using a temperature sensor inside the photovoltaic inverter to continuously collect component temperatures of corresponding components inside the photovoltaic inverter at first time intervals, and recording the collection time and collection location, marking them as first component temperature data;

[0100] Step S102: continuously collecting the ambient temperature inside the photovoltaic inverter at a first time interval, and recording the collection time as first internal temperature data, the first time interval being t1;

[0101] Step S103: using a thermal imager, while maintaining a relative position between the thermal imager and the photovoltaic inverter in space, continuously capturing thermal imaging images of the entire interior of the photovoltaic inverter at a first time interval, and marking the images as first thermal imaging data;

[0102] Step S104, marking a position inside the photovoltaic inverter where the temperature can be collected by the temperature sensor and obtained by the thermal imaging image as a first reference position;

[0103] Step S105 , based on the first thermal imaging data and the first component temperature data, extracting temperatures corresponding to the first thermal imaging data and temperatures corresponding to the first component temperature data at all acquisition moments and all first reference positions, and marking them in sequence as second thermal imaging data and second component temperature data, respectively;

[0104] Step S106, performing data cleaning processing on the second thermal imaging data, the second component temperature data, and the first internal temperature data, to obtain thermal imaging temperature data, the third component temperature data, and the second internal temperature data;

[0105] Step S107: data cleaning. Step S107 includes the following sub-steps:

[0106] Step S1071, sorting the corresponding data according to the corresponding first reference position and collection time, recording them as a time data series, calculating the average value and standard deviation of the time data series, recording them as E0 and E1 respectively;

[0107] Step S1072: For any data in the time data sequence, denoted as Ei, where i represents the sequence number in the time data sequence, if Ei satisfies |Ei-E0|>3*E1, then Ei is marked as abnormal data;

[0108] Step S1073: If Ei does not satisfy |Ei-E0|>3*E1, then mark Ei as normal data, and repeat the marking process for all data in the time data series;

[0109] Step S108 : The third component temperature data and the second internal temperature data are combined and stored according to the collection time, and marked as sensor temperature data.

[0110] Step S2: establishing a thermal imaging temperature correction model based on the sensor temperature data and the thermal imaging temperature data. Step S2 includes the following sub-steps:

[0111] Step S201 , merging the thermal imaging temperature data and the sensor temperature data according to the acquisition time and the first reference position of the acquisition, and recording the data as first sample data;

[0112] Step S202: For the first sample data, the first sample data at the same collection time and the same first reference position are combined into a sample vector, denoted as M = [m1, m2, m3], where m1 represents the temperature of the first reference position measured by the temperature sensor, m2 represents the temperature of the first reference position corresponding to m1 displayed on the thermal imaging image, and m3 represents the ambient temperature inside the photovoltaic inverter;

[0113] Step S203: If there are abnormal data in m1, m2, and m3 in the sample vector M, remove them; after completion, obtain the second sample data;

[0114] Step S204: normalize the second sample data according to the component types of the sample vectors, and scale the sizes of all components of all sample vectors in the second sample data to [0, 1]. After completion, third sample data is obtained.

[0115] Step S205, dividing the third sample data into a training data set and a test data set in a ratio of 9:1;

[0116] Step S206: construct a first temperature correction model based on a multilayer perceptron. The first temperature correction model includes an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is set to a1, the number of hidden layers is set to a2, the number of neurons in each hidden layer is set to a3, and the number of neurons in the output layer is set to a4.

[0117] Step S207, using the training data set to perform model training on the first temperature correction model, and upon completion, a thermal imaging temperature correction model is obtained;

[0118] Step S208: Use the test data set to perform model testing on the thermal imaging temperature correction model, and calculate the average error AE of the thermal imaging temperature correction model. The average error calculation formula is as follows: , where Yj is the temperature of the first reference position measured by the temperature sensor, Xj is the temperature of the first reference position output by the thermal imaging temperature correction model, and n is the number of data input to the thermal imaging temperature correction model.

[0119] Step S3, performing temperature correction processing on the thermal imaging image inside the photovoltaic inverter based on the thermal imaging temperature correction model and data from the temperature sensor inside the photovoltaic inverter to obtain corrected thermal imaging temperature data; Step S3 includes the following sub-steps:

[0120] Step S301: collecting thermal imaging images of the entire interior of the photovoltaic inverter at first time intervals, and simultaneously collecting component temperatures obtained by temperature sensors inside the photovoltaic inverter and the internal ambient temperature;

[0121] Step S302: extracting the temperature values corresponding to all pixels in the collected thermal imaging image, recording the values as raw thermal imaging temperature information, normalizing the raw thermal imaging temperature information with the corresponding ambient temperature inside the photovoltaic inverter, and inputting the normalized values into a thermal imaging temperature correction model to obtain first corrected temperature information.

[0122] Step S303: Based on the first corrected temperature information and the component temperature obtained by the temperature sensor at the same time, a first error between the first corrected temperature information at all first reference positions and the temperature obtained by the temperature sensor is calculated using a first error calculation formula. The first error calculation formula is as follows: , where AF represents the first error, U represents the temperature of the first reference position measured by the temperature sensor, and V is the temperature of the first reference position corresponding to the first corrected temperature information; and all the first errors are averaged to obtain the second error AG;

[0123] In step S304, the temperature correction error is calculated using the correction error calculation formula. The correction error calculation formula is as follows: , where AH is the temperature correction error, q1 and q2 are weight coefficients, q1+q2=1, and the value range of q1 and q2 is [0, 1];

[0124] Step S305: Based on the temperature correction error, the first corrected temperature information is corrected using a temperature correction formula. The temperature correction formula is as follows: , where TY is the temperature value in the first corrected temperature information, and TX is the temperature value after error correction; after completion, the second corrected temperature information is obtained, which is marked as the corrected thermal imaging temperature data.

[0125] Step S4, dynamically controlling the cooling fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and the data from the temperature sensor inside the photovoltaic inverter; Step S4 includes the following sub-steps:

[0126] Step S401: Combine the corrected thermal imaging temperature data and the data of the internal temperature sensors of the photovoltaic inverter at the same moment, and replace the temperature value corresponding to the first reference position in the corrected thermal imaging temperature data with the temperature value collected by the corresponding temperature sensor. After completion, obtain the internal temperature distribution data;

[0127] Step S402: Set the safe temperature range, denoted as [B1, C1]. According to the safe temperature range, set the optimal temperature range, denoted as [B2, C2], where B1 < B2 and C2 > C1;

[0128] Step S403: Extract the maximum value of the internal temperature distribution data, denoted as the internal maximum temperature TM. Obtain the internal maximum temperature TM at the first time interval, denoted as the internal maximum temperature sequence. Perform linear fitting on the internal maximum temperature sequence to obtain the internal maximum temperature sequence for the future first time length, denoted as the predicted maximum temperature sequence. Denote any temperature value in the predicted maximum temperature sequence as Tk, where k represents the serial number in the predicted maximum temperature sequence, and the first time length is t2;

[0129] Step S404: If the internal maximum temperature TM at the current moment satisfies TM ∈ [B2, C2] and any Tk ∈ [B2, C2], then do not adjust the rotation speed of the cooling fan;

[0130] Step S405: If the internal maximum temperature TM at the current moment satisfies TM ∈ [B2, C2] and there exists Tk > C2, then increase the rotation speed of the cooling fan;

[0131] Step S406: If the internal maximum temperature TM at the current moment satisfies TM ∈ [B2, C2] and there exists Tk < B2, then decrease the rotation speed of the cooling fan;

[0132] Step S407: If the internal maximum temperature TM at the current moment satisfies TM ∈ (C2, C1] and any Tk < C1, then increase the rotation speed of the cooling fan;

[0133] Step S408: If the internal maximum temperature TM at the current moment satisfies TM ∈ (C2, C1] and any Tk > C1 or TM > C1, then increase the rotation speed of the cooling fan to the maximum rotation speed and issue a warning signal;

[0134] Step S409: If the internal maximum temperature TM at the current moment satisfies TM ∈ [B1, B2) and any Tk > B1, then decrease the rotation speed of the cooling fan;

[0135] Step S410: If the internal maximum temperature TM at the current moment satisfies TM ∈ [B1, B2) and any Tk < C1 or TM < B1, then decrease the rotation speed of the cooling fan to the minimum rotation speed and issue a warning signal.

[0136] Example 3, please refer to Figure 5 As shown, Figure 5 The present invention provides a schematic structural diagram of an electronic device, which may include a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus. The memory stores computer-readable instructions, and the processor can call the instructions in the memory. When the computer-readable instructions are executed by the processor, the steps of a method for intelligent heat dissipation control of a photovoltaic inverter are executed to implement the following functions: using a temperature sensor to collect component temperatures and internal ambient temperature of the photovoltaic inverter, simultaneously collecting a thermal imaging image of the photovoltaic inverter, and performing a first screening process to obtain sensor temperature data and thermal imaging temperature data; establishing a thermal imaging temperature correction model based on the sensor temperature data and the thermal imaging temperature data; performing temperature correction processing on the thermal imaging image of the photovoltaic inverter based on the thermal imaging temperature correction model and data from the temperature sensor inside the photovoltaic inverter to obtain corrected thermal imaging temperature data; and dynamically controlling the cooling fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and data from the temperature sensor inside the photovoltaic inverter.

[0137] In addition, the logical instructions in the above-mentioned memory can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application, or the part that contributes to the existing technology, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0138] Example 4. The present application also provides a computer-readable storage medium. The present application provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned intelligent heat dissipation control method of a photovoltaic inverter are executed to achieve the following functions: using a temperature sensor to collect the component temperature inside the photovoltaic inverter and the internal ambient temperature, and at the same time collecting the thermal imaging image inside the photovoltaic inverter, and performing a first screening process to obtain sensor temperature data and thermal imaging temperature data; establishing a thermal imaging temperature correction model based on the sensor temperature data and the thermal imaging temperature data; performing temperature correction processing on the thermal imaging image inside the photovoltaic inverter based on the thermal imaging temperature correction model and the data of the temperature sensor inside the photovoltaic inverter to obtain corrected thermal imaging temperature data; and dynamically controlling the cooling fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and the data of the temperature sensor inside the photovoltaic inverter.

[0139] Through the description of the above embodiments, the embodiments of the present invention can be provided as methods, systems, or computer program products. Based on this understanding, the essence of the above technical solutions or the portion that contributes to the prior art can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for enabling a computer device (such as a personal computer, server, or network device) to execute the methods described in various embodiments or certain portions of the embodiments.

[0140] In the embodiments provided in this application, it should be understood that the disclosed system or method can be implemented in other ways. The embodiments described above are merely illustrative. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of systems, modules and units can be electrical, mechanical or other forms.

[0141] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An intelligent heat dissipation control method for a photovoltaic inverter, characterized in that: The steps include: Using a temperature sensor to collect component temperatures and the internal ambient temperature of the photovoltaic inverter, and simultaneously collecting a thermal imaging image of the interior of the photovoltaic inverter, and performing a first screening process to obtain sensor temperature data and thermal imaging temperature data; marking a location within the photovoltaic inverter where the temperature can be collected by the temperature sensor and obtained by the thermal imaging image as a first reference location; Establish a thermal imaging temperature correction model based on sensor temperature data and thermal imaging temperature data; Calculate the average error AE of the thermal imaging temperature correction model; the average error calculation formula is as follows: , where Yj is the temperature of the first reference position measured by the temperature sensor, Xj is the temperature of the first reference position output by the thermal imaging temperature correction model, and n is the number of data input to the thermal imaging temperature correction model; Performing temperature correction processing on the thermal imaging image inside the photovoltaic inverter based on the thermal imaging temperature correction model and data from the temperature sensor inside the photovoltaic inverter to obtain corrected thermal imaging temperature data; Dynamically control the cooling fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and the data of the temperature sensor inside the photovoltaic inverter; Performing temperature correction processing on the thermal imaging image inside the photovoltaic inverter based on the thermal imaging temperature correction model and data from the temperature sensor inside the photovoltaic inverter to obtain corrected thermal imaging temperature data includes the following sub-steps: Collecting thermal imaging images of the entire interior of the photovoltaic inverter at a first time interval, and simultaneously collecting component temperatures obtained by a temperature sensor inside the photovoltaic inverter and the internal ambient temperature; Extracting the temperature corresponding to all pixels in the collected thermal imaging image and recording it as raw thermal imaging temperature information, normalizing the raw thermal imaging temperature information with the corresponding ambient temperature inside the photovoltaic inverter, and inputting it into the thermal imaging temperature correction model to obtain first corrected temperature information; Based on the first corrected temperature information and the component temperature obtained by the temperature sensor at the same time, a first error between the first corrected temperature information and the temperature obtained by the temperature sensor at all first reference positions is calculated using a first error calculation formula. The first error calculation formula is as follows: AF=UV, where AF represents the first error, U represents the temperature of the first reference position measured by the temperature sensor, and V is the temperature of the first reference position corresponding to the first corrected temperature information; and all first errors are averaged to obtain a second error AG. Then calculate the temperature correction error using the correction error calculation formula. The correction error calculation formula is as follows: AH=q1*AE+q2*AG, where AH is the temperature correction error, q1 and q2 are weight coefficients, q1+q2=1, and the value range of q1 and q2 is [0, 1]. Based on the temperature correction error, the first corrected temperature information is corrected using the temperature correction formula. The temperature correction formula is as follows: TX=TY+AH, where TY is the temperature value in the first corrected temperature information and TX is the temperature value after error correction. After completion, the second corrected temperature information is obtained and marked as corrected thermal imaging temperature data.

2. The intelligent heat dissipation control method for a photovoltaic inverter according to claim 1, characterized in that: Using temperature sensors to collect component temperatures and ambient temperature inside the photovoltaic inverter, while also collecting thermal imaging images inside the photovoltaic inverter, includes the following sub-steps: Using a temperature sensor inside the photovoltaic inverter, continuously collect the component temperature of a corresponding component inside the photovoltaic inverter at a first time interval, and record the collection time and collection location, marking it as first component temperature data; at the same time, continuously collect the ambient temperature inside the photovoltaic inverter at a first time interval, and record the collection time, marking it as first internal temperature data, and the first time interval is t1; Using a thermal imager, while keeping the relative positions of the thermal imager and the photovoltaic inverter in space unchanged, thermal imaging images of the entire interior of the photovoltaic inverter are continuously collected at a first time interval, and marked as first thermal imaging data.

3. The intelligent heat dissipation control method for a photovoltaic inverter according to claim 2, characterized in that: Performing the first screening process to obtain sensor temperature data and thermal imaging temperature data includes the following sub-steps: Based on the first thermal imaging data and the first component temperature data, respectively extracting temperatures corresponding to the first thermal imaging data and temperatures corresponding to the first component temperature data at all acquisition moments and all first reference positions, and marking them in sequence as second thermal imaging data and second component temperature data, respectively; Performing data cleaning on the second thermal imaging data, the second component temperature data, and the first internal temperature data, respectively, to obtain thermal imaging temperature data, the third component temperature data, and the second internal temperature data; The data cleaning process includes: sorting the corresponding data according to the corresponding first reference position and the acquisition time, recording them as a time data series, calculating the mean value and standard deviation of the time data series, recording them as E0 and E1 respectively in order, and for any data in the time data series, recording them as Ei, where i represents the sequence number in the time data series. If Ei satisfies |Ei-E0|>3*E1, then Ei is marked as abnormal data; if Ei does not satisfy |Ei-E0|>3*E1, then Ei is marked as normal data, and the marking process is repeated for all data in the time data series. The third component temperature data and the second internal temperature data are combined and stored according to the acquisition time, and marked as sensor temperature data.

4. The intelligent heat dissipation control method for a photovoltaic inverter according to claim 3, characterized in that: Establishing a thermal imaging temperature correction model based on sensor temperature data and thermal imaging temperature data includes the following sub-steps: Merging the thermal imaging temperature data and the sensor temperature data according to the acquisition time and the first reference position of the acquisition, and recording the result as first sample data; For the first sample data, the first sample data at the same acquisition time and the same first reference position are combined into a sample vector, denoted as M = [m1, m2, m3], where m1 represents the temperature of the first reference position measured by the temperature sensor, m2 represents the temperature of the first reference position corresponding to m1 displayed by the thermal imaging image, and m3 represents the ambient temperature inside the PV inverter; If there are abnormal data in m1, m2 and m3 in the sample vector M, remove them; after completion, obtain the second sample data; Normalizing the second sample data according to the component types of the sample vectors, scaling the sizes of all components of all sample vectors in the second sample data to [0, 1], and obtaining the third sample data after completion; The third sample data is divided into a training data set and a test data set in a ratio of 9:

1.

5. The intelligent heat dissipation control method for a photovoltaic inverter according to claim 4, characterized in that: Establishing a thermal imaging temperature correction model based on sensor temperature data and thermal imaging temperature data further includes the following sub-steps: Construct a first temperature correction model based on a multi-layer perceptron. The first temperature correction model includes an input layer, a hidden layer, and an output layer. Set the number of neurons in the input layer as a1, the number of layers in the hidden layer as a2, the number of neurons in each hidden layer as a3, and the number of neurons in the output layer as a4; Use the training data set to train the first temperature correction model, and after completion, obtain the thermal imaging temperature correction model; Use the test data set to test the thermal imaging temperature correction model.

6. The intelligent heat dissipation control method for a photovoltaic inverter according to claim 5, characterized in that: Dynamically controlling the cooling fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and the data of the internal temperature sensor of the photovoltaic inverter includes the following sub-steps: Merge the corrected thermal imaging temperature data and the data of the internal temperature sensor of the photovoltaic inverter at the same moment, and replace the temperature value corresponding to the first reference position in the corrected thermal imaging temperature data with the temperature value collected by the corresponding temperature sensor. After completion, obtain the internal temperature distribution data; Set the safe temperature range, denoted as [B1, C1], and according to the safe temperature range, set the optimal temperature range, denoted as [B2, C2], where B1 < B2 and C1 > C2.

7. The intelligent heat dissipation control method for a photovoltaic inverter according to claim 6, characterized in that: Dynamically controlling the cooling fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and the data of the internal temperature sensor of the photovoltaic inverter further includes the following sub-steps: Extract the maximum value of the internal temperature distribution data, denoted as the internal highest temperature TM. Obtain the internal highest temperature TM at the first time interval, denoted as the internal highest temperature sequence. Perform linear fitting on the internal highest temperature sequence to obtain the internal highest temperature sequence of the first future time length, denoted as the predicted highest temperature sequence. Denote any temperature value in the predicted highest temperature sequence as Tk, where k represents the serial number in the predicted highest temperature sequence, and the first time length is t2; If the internal highest temperature TM at the current moment satisfies TM ∈ [B2, C2] and any Tk ∈ [B2, C2], then do not adjust the rotation speed of the cooling fan; If the internal highest temperature TM at the current moment satisfies TM ∈ [B2, C2] and there exists Tk > C2, then increase the rotation speed of the cooling fan; If the internal highest temperature TM at the current moment satisfies TM ∈ [B2, C2] and there exists Tk < B 2, then reduce the rotation speed of the cooling fan; If the internal highest temperature TM at the current moment satisfies TM ∈ (C2, C1] and any Tk < C1, then increase the rotation speed of the cooling fan; If the internal highest temperature TM at the current moment satisfies TM ∈ (C2, C1] and any Tk > C1 or TM > C1, then increase the rotation speed of the cooling fan to the maximum rotation speed and issue a warning signal; If the internal highest temperature TM at the current moment satisfies TM ∈ [B1, B2) and any Tk > B1, then reduce the rotation speed of the cooling fan; If the internal highest temperature TM at the current moment satisfies TM ∈ [B1, B2) and any Tk < B1 or TM < B1, then reduce the rotation speed of the cooling fan to the lowest rotation speed and issue a warning signal.

8. An intelligent heat dissipation control system for a photovoltaic inverter, used to implement the intelligent heat dissipation control method for a photovoltaic inverter according to any one of claims 1 to 7, characterized in that: It includes temperature acquisition module, model building module, temperature correction module and heat dissipation control module; The temperature acquisition module includes a data acquisition unit and a data screening unit. The data acquisition unit uses a temperature sensor to collect the temperature of components inside the photovoltaic inverter and the internal ambient temperature, and simultaneously collects a thermal imaging image inside the photovoltaic inverter. The data screening unit is used to perform a first screening process to obtain sensor temperature data and thermal imaging temperature data. The model building module builds a thermal imaging temperature correction model based on the sensor temperature data and the thermal imaging temperature data; The temperature correction module performs temperature correction processing on the thermal imaging image inside the photovoltaic inverter based on the thermal imaging temperature correction model and the data of the temperature sensor inside the photovoltaic inverter to obtain corrected thermal imaging temperature data; The heat dissipation control module dynamically controls the heat dissipation fan of the photovoltaic inverter based on the corrected thermal imaging temperature data and the data of the temperature sensor inside the photovoltaic inverter.

Images