A photovoltaic support load deformation monitoring system

Abstract

The present invention relates to the field of photovoltaic power generation technology, and specifically to a photovoltaic support load deformation monitoring system, the system comprising: a data acquisition module, used to collect wind data, photovoltaic support center column inclination data and wind load data; a data transmission module, used to transmit the collected data to a monitoring center; a monitoring center, used to receive the data transmitted by the data transmission module, and analyze, store and display it; an alarm module, used to send an alarm signal when the photovoltaic support load exceeds a threshold after the monitoring center analyzes the data. A photovoltaic support load deformation monitoring system proposed by the present invention improves monitoring accuracy and real-time performance by real-time monitoring of the deformation of the photovoltaic support under load, thereby ensuring the safe operation of the photovoltaic power station.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power generation, and in particular to a photovoltaic support load deformation monitoring system. Background Art

[0002] With the widespread application of solar photovoltaic power generation technology, photovoltaic brackets, as an important part of photovoltaic power stations, are directly related to the safety and stability of the entire power station. Since photovoltaic brackets are exposed to the outdoor environment for a long time and are affected by natural factors, they are prone to deformation and damage. Therefore, real-time monitoring of the load deformation of photovoltaic brackets is of great practical significance. Most of the existing photovoltaic bracket monitoring systems have problems such as low monitoring accuracy, poor real-time performance and lack of predictability, which cannot meet the needs of engineering applications. Summary of the invention

[0003] In view of the above-mentioned problems, the present invention provides a photovoltaic support load deformation monitoring system.

[0004] The present invention is achieved through the following technical solutions:

[0005] A photovoltaic support load deformation monitoring system is provided, the system comprising:

[0006] Data acquisition module: used to collect wind data, photovoltaic bracket center column inclination data and wind load data;

[0007] Data transmission module: used to transmit the collected data to the monitoring center;

[0008] Monitoring center: used to receive data transmitted by the data transmission module, and analyze, store and display it;

[0009] Alarm module: used to set the wind load threshold of the photovoltaic support and compare it with the data transmitted by the monitoring center. When the data exceeds the safety threshold, an alarm signal is issued;

[0010] Among them, the monitoring center uses artificial intelligence algorithms to predict the wind speed of the photovoltaic bracket based on the wind data collected by the data acquisition module, including wind speed and wind direction, and calculates the wind load of the photovoltaic bracket based on the predicted wind speed, and compares it with the actual wind load collected by the data acquisition module to verify the accuracy of the prediction;

[0011] Among them, when the wind load size of the photovoltaic support predicted by the monitoring center exceeds the photovoltaic support wind load threshold set in the alarm module, an alarm is issued through the alarm module.

[0012] Preferably, the data acquisition module further includes:

[0013] Wind monitoring unit: including a wind outlet meteorological instrument installed on the photovoltaic bracket, which is used to collect wind data in real time and monitor changes in wind speed and direction. A high-precision wind outlet meteorological instrument is used to ensure the accuracy and reliability of the measurement data. It can work normally even in extreme weather conditions to ensure real-time monitoring of wind power;

[0014] Tilt monitoring unit: including wiring-free high-precision tilt sensors installed at mountain passes and slopes, used to collect real-time tilt data of the center column of the photovoltaic support, monitor and observe the tilt state of the center column of the photovoltaic support in real time. The use of high-precision tilt sensors can accurately capture the slight tilt changes of the center column of the photovoltaic support, calculate the bearing threshold of the bearing capacity of the axle pin at the bottom of the photovoltaic support, and effectively prevent safety accidents and property losses caused by excessive tilt of the photovoltaic support;

[0015] Wind load monitoring unit: includes a pressure sensor installed on the photovoltaic bracket, which is used to collect the actual wind load data of the photovoltaic bracket to help evaluate the stability and safety of the photovoltaic bracket under different wind conditions.

[0016] Preferably, the data transmission module comprises:

[0017] Wired transmission unit: used to transmit the collected wind data, the inclination data of the center column of the photovoltaic support and the wind load data to the monitoring center through wired transmission. The wired transmission unit consists of cables, interface devices and wired communication protocols. The cables are responsible for physical connection and realize the input and output of the interface devices. The wired communication protocol is used to ensure the integrity and security of the data during the transmission process. Specifically, the wired transmission unit transmits the data collected by the sensor directly to the monitoring center through the cable. During the data transmission process, the interface device encodes and decodes the data to ensure the accuracy of the data during long-distance transmission. Among them, the wired communication protocols used include Ethernet, RS-232, RS-485, etc.

[0018] Preferably, the monitoring center includes:

[0019] Data analysis unit: used to predict the wind speed of the photovoltaic bracket using artificial intelligence algorithm according to the wind data collected by the data acquisition module, including wind speed and wind direction, calculate the wind load of the photovoltaic bracket according to the predicted wind speed, and compare it with the actual wind load collected by the data acquisition module to verify the accuracy of the prediction;

[0020] Data storage unit: used to store and manage the data collected by the data acquisition module and the wind load data calculated by the data analysis unit;

[0021] Interface display unit: Uses a graphical interface to display in real time the data collected by the data acquisition module and the wind load data calculated by the data analysis unit.

[0022] The data analysis unit uses a long short-term memory network model to predict the magnitude and direction of the wind load suffered by the photovoltaic bracket, and the specific steps include:

[0023] Data collection and preprocessing: Real-time wind speed, wind direction and wind load data are obtained from the wind outlet meteorological instrument in the wind monitoring unit and the pressure sensor in the wind load monitoring unit as input features of the long short-term memory network model. Data cleaning is performed, and Z-score standardization is performed after outliers and missing values ​​are removed to convert the data into a distribution with a mean of 0 and a standard deviation of 1.

[0024] Construct a long short-term memory network model: determine the model structure, including the input layer, hidden layer, and output layer, and set the model hyperparameters, including the learning rate, batch size, and number of iterations; the input layer inputs the wind data collected by the wind monitoring unit after data preprocessing, including the size and direction of the wind load. The hidden layer contains one or more LSTM units. The number of LSTM units can be determined according to the length and complexity of the data sequence. The output layer outputs the predicted wind data, including the predicted size and direction of the wind load. The learning rate can determine the speed of weight update. Too high a rate may cause the model to diverge, while too low a rate may lead to slow convergence. The batch size refers to the number of samples used for each gradient update. A larger batch can reduce noise but takes up more memory. The number of iterations refers to a complete traversal of the training set as one iteration. The number of iterations determines the number of rounds of model training;

[0025] Model training: Divide the data set into training set, validation set and test set. The training set is used to train the model, the validation set is used to adjust the model's hyperparameters, and the test set is used to evaluate the model's prediction performance. During the training process, the weights and biases of the model are optimized through the back-propagation algorithm to minimize the prediction error.

[0026] Model evaluation: Use the test set to evaluate the performance of the model, using mean square error and root mean square error as error indicators; analyze the evaluation results, and when overfitting and underfitting occur, adjust the model structure and model hyperparameters, and retrain the model;

[0027] Model deployment: Deploy the trained LSTM network model to the data analysis unit of the monitoring center to predict the size and direction of the wind load on the photovoltaic bracket;

[0028] Improvement and optimization: Continue to collect wind speed, wind direction and wind load data and add them to the training set to obtain a new training set. Use the new training set to retrain the long short-term memory network model and update the model version regularly.

[0029] The steps of performing Z-score standardization after removing outliers and missing values ​​in the data collection and preprocessing include:

[0030] Calculate the mean and standard deviation: The wind data collected by the data acquisition module is removed from outliers and missing values, and then the mean and standard deviation of the wind data are calculated;

[0031] Standardization transformation: After calculating the mean and standard deviation of the wind data, the wind data is standardized, as shown in formula (1):

[0032]

[0033] Among them, x' is the standardized wind data, x is the original wind data, μ is the calculated mean, and σ is the calculated standard deviation.

[0034] In addition, in the long short-term memory network model, the input of the input layer is the wind data after preprocessing and Z-score standardization; the hidden layer includes 3 LSTM layers, each LSTM layer includes 3 LSTM units, which are used to capture the long-term dependencies in the wind data; the output of the output layer is the predicted value of the wind load on the photovoltaic bracket.

[0035] Preferably, the alarm module comprises:

[0036] Threshold setting unit: used to set the wind load threshold of the photovoltaic support;

[0037] Deformation judgment unit: used for the monitoring center to predict the wind speed of the photovoltaic bracket using artificial intelligence algorithm according to the wind data collected by the data acquisition module, including wind speed and wind direction, and calculate the wind load of the photovoltaic bracket according to the predicted wind speed. When the wind load of the photovoltaic bracket predicted by the monitoring center exceeds the wind load threshold of the photovoltaic bracket set in the alarm module, the alarm module will sound an alarm;

[0038] Deformation judgment unit: used to send deformation alarm instructions to the sound and light alarm unit and the SMS alarm unit according to the data of the data acquisition module and the monitoring center, when the actual wind load collected by the pressure sensor in the data acquisition module is greater than the predicted wind load threshold of the monitoring center; when the inclination data of the photovoltaic support column collected by the inclination sensor in the inclination monitoring unit exceeds the safety threshold, the deformation alarm instruction is sent to the sound and light alarm unit and the SMS alarm unit;

[0039] Sound and light alarm unit: installed on the photovoltaic support, used to send out sound and light alarm signals at the photovoltaic support site after receiving the deformation alarm command sent by the deformation judgment unit, usually including alarm sound and flashing light, to attract the attention of on-site staff;

[0040] SMS alarm unit: It is used to receive the deformation alarm command sent by the deformation judgment unit and send the alarm information to the staff via SMS. The SMS content usually includes the specific alarm location, alarm type (such as excessive wind load or abnormal inclination) and other relevant information, so that the staff can quickly take corresponding measures.

[0041] The advantages and effects of the present invention are:

[0042] The method of the present invention realizes real-time and accurate monitoring of the status of the photovoltaic bracket, improves the accuracy and real-time performance of the traditional photovoltaic bracket status monitoring, and ensures the safe operation of the photovoltaic power station. In addition, the wind speed, wind direction and real-time wind load of the photovoltaic bracket can be monitored, and the wind load threshold borne by the photovoltaic bracket can be calculated by predicting the size and direction of the wind load that the photovoltaic bracket may suffer based on the monitored wind speed and wind direction. When the photovoltaic bracket has a risk of deformation, an alarm can be issued in time, so that the staff can make a response plan in time to ensure the safety of the personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 This is a structural diagram of a photovoltaic support load deformation monitoring system of the present invention.

[0045] Figure 2 This is a flow chart of the long short-term memory network model training of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0047] The present invention provides a photovoltaic support load deformation monitoring system, such as Figure 1 As shown, including:

[0048] Data acquisition module: used to collect wind data, photovoltaic bracket center column inclination data and wind load data;

[0049] Data transmission module: used to transmit the collected data to the monitoring center;

[0050] Monitoring center: used to receive data transmitted by the data transmission module, and analyze, store and display it;

[0051] Alarm module: used to set the wind load threshold of the photovoltaic support and compare it with the data transmitted by the monitoring center. When the data exceeds the safety threshold, an alarm signal will be issued.

[0052] Specifically, the data acquisition module includes:

[0053] Wind monitoring unit: including a wind outlet meteorological instrument installed on the photovoltaic bracket, which is used to collect wind data in real time and monitor changes in wind speed and direction. A high-precision wind outlet meteorological instrument is used to ensure the accuracy and reliability of the measurement data. It can work normally even in extreme weather conditions to ensure real-time monitoring of wind power;

[0054] Tilt monitoring unit: including wiring-free high-precision tilt sensors installed at mountain passes and slopes, used to collect real-time tilt data of the center column of the photovoltaic support, monitor and observe the tilt state of the center column of the photovoltaic support in real time. The use of high-precision tilt sensors can accurately capture the slight tilt changes of the center column of the photovoltaic support, calculate the bearing threshold of the bearing capacity of the axle pin at the bottom of the photovoltaic support, and effectively prevent safety accidents and property losses caused by excessive tilt of the photovoltaic support;

[0055] Wind load monitoring unit: includes a pressure sensor installed on the photovoltaic bracket, which is used to collect the actual wind load data of the photovoltaic bracket to help evaluate the stability and safety of the photovoltaic bracket under different wind conditions.

[0056] Specifically, the data transmission module includes:

[0057] Wired transmission unit: used to transmit the collected wind data, the inclination data of the center column of the photovoltaic support and the wind load data to the monitoring center through wired transmission. The wired transmission unit consists of cables, interface devices and wired communication protocols. The cables are responsible for physical connection and realize the input and output of the interface devices. The wired communication protocol is used to ensure the integrity and security of the data during the transmission process. Specifically, the wired transmission unit transmits the data collected by the sensor directly to the monitoring center through the cable. During the data transmission process, the interface device encodes and decodes the data to ensure the accuracy of the data during long-distance transmission. Among them, the wired communication protocols used include Ethernet, RS-232, RS-485, etc.

[0058] Specifically, the monitoring center includes:

[0059] Data analysis unit: used to predict the wind speed of the photovoltaic bracket using artificial intelligence algorithm according to the wind data collected by the data acquisition module, including wind speed and wind direction, calculate the wind load of the photovoltaic bracket according to the predicted wind speed, and compare it with the actual wind load collected by the data acquisition module to verify the accuracy of the prediction;

[0060] Data storage unit: used to store and manage the data collected by the data acquisition module and the wind load data predicted by the data analysis unit. The data collected by the data acquisition module mainly includes wind data, inclination data and the wind load data actually borne by the photovoltaic bracket. In addition, the stored data is classified, indexed and backed up to facilitate subsequent search and generation of wind data reports.

[0061] Interface display unit: A graphical interface is used to display in real time the data collected by the data acquisition module and the wind load data calculated by the data analysis unit, so that the staff can clearly understand the current status of the photovoltaic support.

[0062] The data analysis unit uses a long short-term memory network model to predict the magnitude and direction of wind loads on photovoltaic brackets. The specific steps are as follows: Figure 2 As shown, including:

[0063] Data collection and preprocessing: Real-time wind speed, wind direction and wind load data are obtained from the wind outlet meteorological instrument in the wind monitoring unit and the pressure sensor in the wind load monitoring unit as input features of the long short-term memory network model. Data cleaning is performed, and Z-score standardization is performed after outliers and missing values ​​are removed to convert the data into a distribution with a mean of 0 and a standard deviation of 1.

[0064] The steps for Z-score standardization after removing outliers and missing values ​​in data collection and preprocessing include:

[0065] Calculate the mean and standard deviation: The wind data collected by the data acquisition module is removed from outliers and missing values, and then the mean and standard deviation of the wind data are calculated;

[0066] Standardization transformation: After calculating the mean and standard deviation of the wind data, the wind data is standardized, as shown in formula (1):

[0067]

[0068] Among them, x' is the standardized wind data, x is the original wind data, μ is the calculated mean, and σ is the calculated standard deviation;

[0069] In the data collection and preprocessing step, the 3σ principle in statistical methods is used to deal with outliers. For a given data set, if a data point falls outside the mean value μ plus or minus three standard deviations σ, then this data point can be regarded as an outlier. The calculation formula of the mean value μ is shown in formula (2):

[0070]

[0071] Among them, x i represents the i-th data point, and n is the total number of data points in the dataset.

[0072] After obtaining the mean value μ, calculate the standard deviation σ. The calculation formula is shown in formula (3):

[0073]

[0074] The meaning of each parameter is the same as that of formula (2). After calculating the mean value μ and standard deviation σ, the range of outliers can be determined. The definition of the outlier range is shown in formula (4):

[0075] (μ-3σ,μ+3σ)(4)

[0076] That is, any data point lower than μ-3σ or higher than μ+3σ can be considered as an outlier.

[0077] Specifically, assuming that the wind monitoring unit collects a set of wind speed data, first calculate the mean and standard deviation of the set of data. Then, according to the 3σ principle, the range of outliers can be determined. If a certain wind speed data deviates far from this range, then this value may be an outlier caused by measurement error or other reasons, and the outlier is cleaned;

[0078] Construct a long short-term memory network model: determine the model structure, including the input layer, hidden layer and output layer, and set the model hyperparameters, including the learning rate, batch size and number of iterations; the input layer inputs the wind data collected by the wind monitoring unit after data preprocessing, including the size and direction of the wind load. The hidden layer contains one or more LSTM units. The number of LSTM units can be determined according to the length and complexity of the data sequence. Each LSTM layer includes one or more LSTM units to capture the long-term dependencies in the wind data; the output layer outputs the predicted wind data, including the predicted size and direction of the wind load. The learning rate can determine the speed of weight update. Too high a learning rate may cause the model to diverge, while too low a learning rate may lead to slow convergence. The batch size refers to the number of samples used for each gradient update. A larger batch can reduce noise but takes up more memory. The number of iterations refers to a complete traversal of the training set as one iteration. The number of iterations determines the number of rounds of model training.

[0079] Model training: The data set is divided into a training set (about 70% to 80%), a validation set (about 10% to 15%), and a test set (about 10% to 15%). The training set is used to train the model, and the weights and biases of the model are continuously adjusted through the back-propagation algorithm to minimize the prediction error; the validation set is used to adjust the hyperparameters of the model to ensure that the model is neither overfitting nor underfitting; the test set is used to evaluate the prediction performance of the model to ensure that the model has good generalization ability;

[0080] Model evaluation: Use the test set to evaluate the performance of the model, using mean square error and root mean square error as error indicators; analyze the evaluation results, and when overfitting and underfitting occur, adjust the model structure and model hyperparameters, and retrain the model;

[0081] Model deployment: Deploy the trained LSTM network model to the data analysis unit of the monitoring center to predict the size and direction of the wind load on the photovoltaic bracket;

[0082] Improvement and optimization: Continue to collect wind speed, wind direction and wind load data and add them to the training set, create a new training set, including old data and the new data set, use the new training set to retrain the long short-term memory network model to adapt to environmental changes or improve prediction accuracy, and regularly update the model version to ensure that the model is always kept up to date and provides the most accurate predictions.

[0083] Specifically, the alarm module includes:

[0084] Threshold setting unit: used to set the wind load threshold of the photovoltaic support;

[0085] Deformation judgment unit: used for the monitoring center to predict the wind speed of the photovoltaic bracket using artificial intelligence algorithm according to the wind data collected by the data acquisition module, including wind speed and wind direction, and calculate the wind load of the photovoltaic bracket according to the predicted wind speed. When the wind load of the photovoltaic bracket predicted by the monitoring center exceeds the wind load threshold of the photovoltaic bracket set in the alarm module, the alarm module will sound an alarm;

[0086] Sound and light alarm unit: installed on the photovoltaic support, used to send out sound and light alarm signals at the photovoltaic support site after receiving the deformation alarm command sent by the deformation judgment unit, usually including alarm sound and flashing light, to attract the attention of on-site staff;

[0087] SMS alarm unit: It is used to receive the deformation alarm command sent by the deformation judgment unit and send the alarm information to the staff via SMS. The SMS content usually includes the specific alarm location, alarm type (such as excessive wind load or abnormal inclination) and other relevant information, so that the staff can quickly take corresponding measures.

[0088] The photovoltaic support load deformation monitoring system provided in the present application adopts the photovoltaic support load deformation monitoring system in the above embodiment, which can solve the technical problem of low monitoring accuracy of the traditional photovoltaic support load monitoring system. Compared with the prior art, the beneficial effects of the photovoltaic support load deformation monitoring system provided in the present application are the same as the beneficial effects of the photovoltaic support load deformation monitoring system provided in the above embodiment, and the other technical features in the photovoltaic support load deformation monitoring system are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0089] The various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0090] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A photovoltaic support load deformation monitoring system, characterized in that: The system comprises: Data acquisition module: used to collect wind data, photovoltaic bracket center column inclination data and wind load data; Data transmission module: used to transmit the collected data to the monitoring center; Monitoring center: used to receive data transmitted by the data transmission module, and analyze, store and display it; Alarm module: used to set the wind load threshold of the photovoltaic support and compare it with the data transmitted by the monitoring center. When the data exceeds the safety threshold, an alarm signal is issued; Among them, the monitoring center analyzes the received wind data by combining artificial intelligence algorithms.

2. A photovoltaic support load deformation monitoring system according to claim 1, characterized in that: The data acquisition module also includes: Wind monitoring unit: including a wind outlet meteorological instrument installed on the photovoltaic bracket, used to collect wind data and monitor changes in wind speed and direction; Tilt monitoring unit: includes a wiring-free tilt sensor installed at mountain passes and slopes, which is used to collect the tilt data of the center column of the photovoltaic support and monitor and observe the tilt state of the center column of the photovoltaic support in real time; Wind load monitoring unit: includes a pressure sensor installed on the photovoltaic bracket, which is used to collect the actual wind load data of the photovoltaic bracket.

3. A photovoltaic support load deformation monitoring system according to claim 1, characterized in that: The data transmission module comprises: Wired transmission unit: used to transmit the collected wind data, photovoltaic bracket center column inclination data and wind load data to the monitoring center through wired transmission.

4. A photovoltaic support load deformation monitoring system according to claim 1, characterized in that: The monitoring center predicts the wind speed of the photovoltaic bracket using an artificial intelligence algorithm based on the wind data, including wind speed and wind direction, collected by the data acquisition module, calculates the wind load of the photovoltaic bracket based on the predicted wind speed, and compares it with the actual wind load collected by the data acquisition module to verify the accuracy of the prediction.

5. A photovoltaic support load deformation monitoring system according to claim 1, characterized in that: The monitoring center includes: Data analysis unit: used to predict the wind speed of the photovoltaic bracket using artificial intelligence algorithm according to the wind data collected by the data acquisition module, including wind speed and wind direction, calculate the wind load of the photovoltaic bracket according to the predicted wind speed, and compare it with the actual wind load collected by the data acquisition module to verify the accuracy of the prediction; Data storage unit: used to store and manage the data collected by the data acquisition module and the wind load data calculated by the data analysis unit; Interface display unit: Uses a graphical interface to display in real time the data collected by the data acquisition module and the wind load data calculated by the data analysis unit.

6. A photovoltaic support load deformation monitoring system according to claim 4, characterized in that: The data analysis unit uses a long short-term memory network model to predict the magnitude and direction of the wind load suffered by the photovoltaic bracket, and the specific steps include: Data collection and preprocessing: Real-time wind speed, wind direction and wind load data are obtained from the wind outlet meteorological instrument in the wind monitoring unit and the pressure sensor in the wind load monitoring unit. Data cleaning is performed to remove outliers and missing values, and then Z-score standardization is performed to convert the data into a distribution with a mean of 0 and a standard deviation of 1 as the input features of the long short-term memory network model. Build a long short-term memory network model: determine the model structure, including the input layer, hidden layer, and output layer, and set the model hyperparameters, including learning rate, batch size, and number of iterations; Model training: The real-time wind speed, wind direction and wind load data obtained by the wind outlet meteorological instrument in the wind monitoring unit and the pressure sensor in the wind load monitoring unit are used as data sets and divided into training set, validation set and test set. The training set is used to train the model, the validation set is used to adjust the hyperparameters of the model, and the test set is used to evaluate the prediction performance of the model. During the training process, the weight and bias of the model are optimized through the back propagation algorithm to minimize the prediction error. Model evaluation: Use the test set to evaluate the performance of the model, using mean square error and root mean square error as error indicators; analyze the evaluation results, and when overfitting and underfitting occur, adjust the model structure and model hyperparameters, and retrain the model; Model deployment: Deploy the trained LSTM network model to the data analysis unit of the monitoring center to predict the size and direction of the wind load on the photovoltaic bracket; Improvement and optimization: Continue to collect wind speed, wind direction and wind load data and add them to the training set to obtain a new training set. Use the new training set to retrain the long short-term memory network model and update the model version regularly.

7. A photovoltaic support load deformation monitoring system according to claim 6, characterized in that: The steps of performing Z-score standardization after removing outliers and missing values ​​in the data collection and preprocessing include: Calculate the mean and standard deviation: The wind data collected by the data acquisition module is removed from outliers and missing values, and then the mean and standard deviation of the wind data are calculated; Standardization transformation: After calculating the mean and standard deviation of the wind data, the wind data is standardized, as shown in formula (1): Among them, x' is the standardized wind data, x is the original wind data, μ is the calculated mean, and σ is the calculated standard deviation.

8. A photovoltaic support load deformation monitoring system according to claim 6, characterized in that: In the long short-term memory network model, the input of the input layer is the wind data after preprocessing and Z-score standardization; the hidden layer includes 3 LSTM layers, each LSTM layer includes 3 LSTM units, which are used to capture the long-term dependencies in the wind data; the output of the output layer is the predicted value of the wind load on the photovoltaic bracket.

9. A photovoltaic support load deformation monitoring system according to claim 1, characterized in that: In the alarm module, an alarm is triggered when the wind load on the photovoltaic support predicted by the monitoring center exceeds the wind load threshold of the photovoltaic support set in the alarm module.

10. A photovoltaic support load deformation monitoring system according to claim 1, characterized in that: The alarm module comprises: Threshold setting unit: used to set the wind load threshold of the photovoltaic support; Deformation judgment unit: used for the monitoring center to predict the wind speed of the photovoltaic bracket using artificial intelligence algorithm according to the wind data collected by the data acquisition module, including wind speed and wind direction, and calculate the wind load of the photovoltaic bracket according to the predicted wind speed. When the wind load of the photovoltaic bracket predicted by the monitoring center exceeds the wind load threshold of the photovoltaic bracket set in the alarm module, the alarm module will sound an alarm; Sound and light alarm unit: installed on the photovoltaic support, used to send out sound and light alarm signals on the photovoltaic support site; SMS alarm unit: used to send alarm information to staff via SMS.

Images