A photovoltaic module condition monitoring system based on multi-source data and a method thereof

Abstract

The application provides a photovoltaic module state monitoring system and method based on multi-source data, relates to the technical field of photovoltaic system monitoring, and comprises a photovoltaic module state monitoring system by introducing self data, environmental data and a Kalman filtering algorithm, the self data and the environmental data of the photovoltaic module are collected, the future state of the module is predicted by adopting the Kalman filtering algorithm through a data prediction module, the predicted data of the module are obtained, the environmental influence index introduced through a data correction module can effectively distinguish the cause of the performance decline of the photovoltaic module, the predicted data are taken as a reference standard to generate a self data error index, the self data fluctuation rate is generated by using the self data and the predicted data of the module at the three corrected time points, the state monitoring index can be generated by combining the data analysis model of a monitoring analysis module, and the state monitoring index is compared with a preset threshold value, a warning signal is sent, and the accuracy of photovoltaic system fault detection is improved.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic system monitoring technology, specifically to a photovoltaic module status monitoring system and method based on multi-source data. Background Technology

[0002] With the rapid development of the renewable energy sector, photovoltaic (PV) power generation systems, as a crucial component, have received widespread attention for their efficiency and stability. PV modules, as the core of a PV power generation system, have a decisive impact on the overall system's power generation efficiency and operational safety. However, during long-term operation, PV modules are affected by environmental factors such as temperature and light intensity, as well as internal factors such as material aging and mechanical damage, leading to a gradual decline in their performance. Therefore, how to monitor and evaluate the status of PV modules in real time, and promptly detect and prevent potential faults, has become a key technical issue in ensuring the stable operation of PV systems.

[0003] Traditional methods for monitoring the condition of photovoltaic (PV) modules often rely on simple data collection and manual analysis, which have significant limitations in data processing and fault prediction. First, the environment in which PV modules operate is constantly changing; even small changes can have an impact, such as variations in ambient temperature and light intensity. This often makes it difficult to accurately distinguish whether performance degradation is caused by environmental changes or module malfunctions. Second, traditional methods are inefficient when processing the massive amounts of data generated by large-scale PV systems, making real-time monitoring and fault early warning difficult. Furthermore, manual analysis methods depend on experienced professionals, are susceptible to subjective judgment, and struggle to guarantee the consistency and accuracy of the analysis results.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a photovoltaic module status monitoring system and method based on multi-source data to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A photovoltaic module status monitoring system based on multi-source data includes:

[0008] The data acquisition module is used to collect the photovoltaic module's own data and environmental data at time k and time k-1. The own data includes the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data. The environmental data includes the photovoltaic module's surface sunlight intensity data, the temperature data of the surrounding environment, and the time it has been put into use.

[0009] The data prediction module is used to predict the photovoltaic module's own data at time k+1 based on the photovoltaic module's own data and environmental data at time k-1 and time k-1, and to obtain the module prediction data. The module prediction data includes the photovoltaic module's output current data, voltage data and photovoltaic module surface temperature data at time k+1.

[0010] The data correction module is used to collect the photovoltaic module's own data and environmental data at time k+1. Based on the photovoltaic module's own data and environmental data at time k and time k-1, a first environmental impact index set is generated, and the first environmental impact index set is used to correct the photovoltaic module's own data at time k+1.

[0011] The fluctuation judgment module is used to generate its own data error index based on the corrected self-data of the photovoltaic module at time k+1 and the module prediction data. It uses the first environmental impact dataset to correct the self-data at time k. Based on the self-data at time k-1, the corrected self-data at time k and time k+1, it generates its own data volatility.

[0012] The monitoring and analysis module is used to analyze its own data volatility and data error index, establish a data analysis model, generate a status monitoring index, compare the status monitoring index with a threshold, and select whether to issue an early warning signal based on the comparison result.

[0013] Furthermore, the k+1 time is the current time, and the time length between time k and time k-1 is the same as the time length between time k and time k+1.

[0014] Furthermore, the Kalman filter algorithm is used to predict the photovoltaic module's own data at time k+1. The method used to obtain the module prediction data is as follows:

[0015] Based on the photovoltaic module's own data at time k-1, a state vector of the photovoltaic module at time k-1 is constructed, and a state equation for the state vector transition from time k-1 to time k is constructed in combination with the state transition matrix. This generates self-predicted data at time k based on the photovoltaic module's own data at time k-1 and environmental data, and updates the error covariance matrix of the self-predicted data at time k.

[0016] Based on the environmental data of the photovoltaic module at time k, the observation vector of the photovoltaic module at time k is constructed, and the observation matrix is ​​constructed. The Kalman gain is calculated by combining the updated error covariance matrix and the observation matrix.

[0017] The Kalman gain and the observation vector at time k are used to update the self-predicted data at time k, and the updated self-predicted data at time k is obtained.

[0018] Using the updated self-predicted data at time k, and combining it with the state transition matrix, we construct the state equation from time k to time k+1, and obtain the predicted self-data at time k+1, i.e., component prediction data.

[0019] Furthermore, the state vector of the photovoltaic module at time k-1 is:

[0020]

[0021] Where, x k―1 I represents the state vector of the photovoltaic module at time k-1. k―1 V k―1 and T k―1 These represent the output current, voltage, and surface temperature data of the photovoltaic module at time k-1, respectively.

[0022] The state transition matrix is:

[0023]

[0024] Where F represents the state transition matrix;

[0025] The formula used to generate the self-predicted data at time k based on the photovoltaic module's own data at time k-1 and environmental data is as follows:

[0026] x k|k―1 =x k―1 *F

[0027] Where, x k|k―1 This represents the self-predicted data at time k based on the photovoltaic module's own data at time k-1 and environmental data;

[0028] The formula used to update the error covariance matrix of the self-predicted data at time k is:

[0029] P k|k―1 =F*P k―1|k―1 *F T

[0030] Among them, P k|k―1 F represents the error covariance matrix of the self-predicted data at time k after the update. T P represents the transpose of the state transition matrix. k―1|k―1 Let represent the error covariance matrix of the self-predicted data at time k-1, and:

[0031]

[0032] Where, σ I σ U and σT These represent the tolerances for the current data, voltage data, and surface temperature data of the photovoltaic module, respectively.

[0033] Furthermore, the method used to calculate the Kalman gain is as follows:

[0034] K k =P k|k―1 *H T *(H*P k|k―1 *H T ) ―1

[0035] Among them, K k H represents the Kalman gain, and H represents the observation matrix. T Let represent the transpose of the observation matrix, and:

[0036]

[0037] The observation vector of the photovoltaic module at time k is:

[0038]

[0039] Among them, z k I represents the observation vector of the photovoltaic module at time k. k V k and T k S represents the output current, voltage, and surface temperature data of the photovoltaic module at time k, respectively. k E k and t k These represent the surface solar intensity, ambient temperature, and usage time of the photovoltaic module at time k, respectively.

[0040] The formula used to obtain the updated self-prediction data at time k is:

[0041] x k|k =x k|k―1 +K k (z k ―H*x k|k―1 )

[0042] Where, x k|k This represents the updated self-predicted data at time k, and the formula used to obtain the predicted self-data at time k+1, i.e., the component prediction data, is as follows:

[0043]

[0044] Where, x k+1|k Represents component prediction data, Iy k+1 Vyk+1 and Ty k+1 These represent the predicted current, voltage, and surface temperature data of the photovoltaic module at time k+1 in the module prediction data.

[0045] Furthermore, the first environmental impact index set includes the solar intensity impact index, temperature impact index, and time impact index. The logic underlying the generation of the first environmental impact index set is as follows:

[0046]

[0047] Among them, Gq y Wd y and Sj y These represent the influence indices of sunlight intensity, temperature, and time, respectively. k V k and T k S represents the output current, voltage, and surface temperature data of the photovoltaic module at time k, respectively. k E k and t k Represent the surface solar intensity, ambient temperature, and time of use of the photovoltaic module at time k, respectively. k―1 V k―1 and T k―1 S represents the output current, voltage, and surface temperature data of the photovoltaic module at time k-1, respectively. k―1 E k―1 and t k―1 These represent the surface solar intensity, ambient temperature, and time of use of the photovoltaic module at time k-1, respectively.

[0048] The logic for correcting the photovoltaic module's own data at time k+1 using the first environmental impact index set is as follows:

[0049] Ix k+1 =I k+1 +Gq y *(S k+1 —S k―1 )

[0050] Vx k+1 =V k+1 +Wd y *(E k+1 —E k―1 )

[0051] Tx k+1 =T k+1 +Sj y *(t k+1 ―tk―1 )

[0052] Among them, Ix k+1 Vx k+1 and Tx k+1 I represents the corrected output current, voltage, and surface temperature data of the photovoltaic module at time k+1, respectively. k+1 V k+1 and T k+1 These represent the current, voltage, and surface temperature data of the photovoltaic module at time k+1, respectively. k+1 E k+1 and t k+1 These represent the surface sunlight intensity, ambient temperature, and time of use of the photovoltaic module at time k+1, respectively.

[0053] Furthermore, the formula used to correct the data at time k using the first environmental impact dataset is as follows:

[0054] Ix k =I k +Gq y *(S k —S k―1 )

[0055] Vx k =V k +Wd y *(E k —E k―1 )

[0056] Tx k =T k +Sj y *(t k ―t k―1 )

[0057] Among them, Ix k Vx k and Tx k These represent the corrected current, voltage, and surface temperature data of the photovoltaic module at time k, respectively.

[0058] Furthermore, when generating the self-data error index, the errors in the photovoltaic module's output current data, voltage data, and surface temperature data at time k+1, both from the corrected self-data and the module's predicted data, are assigned corresponding weights. These weights are then summed to calculate the self-data error index. The formula for generating the self-data error index is as follows:

[0059]

[0060] Among them, ZW z Ix represents the data error index. k+1 Vx k+1 and Tx k+1 These represent the corrected output current, voltage, and surface temperature data of the photovoltaic module at time k+1, respectively. k+1 Vy k+1 and Ty k+1 Let α, β, and γ represent the current, voltage, and surface temperature data of the photovoltaic module at time k+1 in the module prediction data, respectively. Let α, β, and γ be the current anomaly weights, voltage anomaly weights, and temperature anomaly weights, respectively. Let α > β > γ > 0, and α + β + γ = 1.

[0061] Furthermore, based on the self-data at time k-1, the corrected self-data at time k, and time k+1, when generating the self-data volatility, the variances of the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data at the three times are calculated respectively. The self-data volatility is obtained by calculating the average of the three sets of variances, specifically:

[0062]

[0063] Among them, Zs b For its own data volatility, Ix k+1 Vx k+1 and Tx k+1 Represent the corrected output current, voltage, and surface temperature data of the photovoltaic module at time k+1, respectively. Ix k Vx k and Tx k Represent the corrected output current, voltage, and surface temperature data of the photovoltaic module at time k, respectively. k―1 V k―1 and T k―1 These represent the output current, voltage, and surface temperature data of the photovoltaic module at time k-1, respectively. and Let represent the average current, average voltage, and average surface temperature, respectively, and:

[0064]

[0065] The formula used to generate the status monitoring index is:

[0066]

[0067] Among them, ZJ z This represents the status monitoring index. When comparing the status monitoring index with a threshold, if ZJ...z ≥ZJ yz If ZJ z <ZJ yz If not, no warning signal will be issued.

[0068] The present invention also provides a photovoltaic module status monitoring method based on multi-source data. The monitoring method is executed by the aforementioned photovoltaic module status monitoring system, and the specific steps include:

[0069] Step 1: Collect the photovoltaic module's own data and environmental data at time k and time k-1. The own data includes the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data. The environmental data includes the photovoltaic module surface sunlight intensity data, the ambient temperature data, and the time of use.

[0070] Step 2: Based on the photovoltaic module's own data and environmental data at time k and k-1, the Kalman filter algorithm is used to predict the photovoltaic module's own data at time k+1 to obtain the module prediction data. The module prediction data includes the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data at time k+1.

[0071] Step 3: Collect the photovoltaic module's own data and environmental data at time k+1. Based on the photovoltaic module's own data and environmental data at time k and time k-1, generate a first environmental impact index set. Use the first environmental impact index set to correct the photovoltaic module's own data at time k+1.

[0072] Step 4: Generate the self-data error index based on the corrected self-data of the photovoltaic module at time k+1 and the module prediction data. Use the first environmental impact dataset to correct the self-data at time k. Generate the self-data volatility based on the self-data at time k-1, the corrected self-data at time k, and the self-data at time k+1.

[0073] Step 5: Analyze the volatility and error index of the data itself, establish a data analysis model, generate a status monitoring index, compare the status monitoring index with the threshold, and choose whether to issue a warning signal based on the comparison results.

[0074] Compared with the prior art, the beneficial effects of the present invention are:

[0075] This invention constructs a photovoltaic module status monitoring system by introducing its own data, environmental data, and Kalman filtering algorithm. By collecting the photovoltaic module's own data and environmental data, and using the Kalman filtering algorithm through the data prediction module to predict the future state of the module, the system obtains the module's predicted data by predicting the photovoltaic module's own data at the next moment. Through the environmental impact index introduced by the data correction module, the system can effectively distinguish the causes of photovoltaic module performance degradation, and promptly distinguish whether the performance degradation is caused by external environmental factors or problems within the module itself, thereby providing a more accurate decision-making basis for the operation and maintenance monitoring of photovoltaic modules.

[0076] This invention uses predicted data as a reference standard to generate its own data error index, which is used to assess whether the actual state of the photovoltaic module is normal. It uses the corrected self-data at three time points and the module's predicted data to generate its own data volatility, which reflects the performance changes of the photovoltaic module at adjacent time points. By using the self-data error index and self-data volatility for analysis, combined with the data analysis model of the monitoring and analysis module, a status monitoring index can be generated and compared with a preset threshold to issue an early warning signal. This improves the accuracy of photovoltaic system fault detection and reduces the workload of manual analysis and the uncertainty of subjective judgment. Attached Figure Description

[0077] Figure 1 This is a schematic diagram of the overall system structure of the present invention;

[0078] Figure 2 This is a schematic diagram of the overall method flow of the present invention. Detailed Implementation

[0079] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0080] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0081] Example:

[0082] Please see Figure 1 The present invention provides a technical solution:

[0083] A photovoltaic module status monitoring system based on multi-source data includes a data acquisition module, a data prediction module, a data correction module, a fluctuation judgment module, and a monitoring and analysis module, wherein:

[0084] The data acquisition module collects the photovoltaic module's own data and environmental data at time k and time k-1. The own data includes the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data. The environmental data includes the photovoltaic module's surface sunlight intensity data, ambient temperature data, and the time it has been in use. The data acquisition module then sends the collected data to [the relevant authority / organization].

[0085] In this embodiment, a photovoltaic module refers to a photovoltaic panel or solar panel, which is a device that uses solar energy to convert light energy into electrical energy. It is composed of multiple solar cell units, which are the basic components for making a photovoltaic module. Each photovoltaic module has a junction box containing circuit connection and protection devices, which can extract the current generated by the solar panel. In this embodiment, a photovoltaic module refers to a single solar panel module that can output electrical energy. Photovoltaic modules can be used independently or installed in an array with other components to increase power output.

[0086] The current and voltage data output by the photovoltaic module are the values ​​of current and voltage output by the power output terminal of the photovoltaic module at a specific moment. The unit of the current data is amperes, and the unit of the voltage data is volts. The surface temperature data of the photovoltaic module is the surface temperature data of the photovoltaic module measured by a temperature sensor placed on the surface of the photovoltaic module at a specific moment. In order to avoid errors, the surface temperature of the photovoltaic module can be collected at multiple locations simultaneously, and then the average value is calculated and calibrated as the surface temperature data of the photovoltaic module. The unit of temperature is degrees Celsius.

[0087] Solar intensity data on the photovoltaic (PV) module surface refers to the intensity of sunlight hitting the PV module surface, measured in watts per square meter. This data reflects the amount of solar energy received by the PV module and directly affects its power generation efficiency. Solar intensity data is measured using a solar radiometer. To avoid errors, the solar intensity at multiple locations on the PV module surface can be collected simultaneously, and the average value is calculated and calibrated as the solar intensity data for the PV module surface. Ambient temperature data refers to the temperature of the environment surrounding the PV module, measured in degrees Celsius. Ambient temperature affects the performance of PV modules because the efficiency of PV cells decreases as temperature increases. Therefore, monitoring ambient temperature helps assess the performance of PV modules under different climatic conditions. Similarly, averaging multiple ambient temperature values ​​can reduce errors. The operational time refers to the cumulative time since the PV module was installed and began generating electricity, measured in hours. Over time, the performance of PV modules may degrade due to aging, pollution, damage, etc.

[0088] The data prediction module uses the Kalman filter algorithm to predict the photovoltaic module's own data at time k+1 based on the photovoltaic module's own data and environmental data at time k-1, and obtains the module prediction data. The module prediction data includes the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data at time k+1.

[0089] The Kalman filter algorithm is a recursive algorithm that can effectively combine historical data and current measurement data to make optimal estimates of future states. In photovoltaic systems, it can predict key parameters such as the output current, voltage, and surface temperature of photovoltaic modules.

[0090] In this embodiment, the Kalman filter algorithm is used to predict the photovoltaic module's own data at time k+1. The method used to obtain the module prediction data is as follows:

[0091] Based on the photovoltaic module's own data at time k-1, a state vector of the photovoltaic module at time k-1 is constructed. Combined with the state transition matrix, a state equation is constructed to transition the state vector from time k-1 to time k. Self-predicted data at time k is generated based on the photovoltaic module's own data at time k-1 and environmental data. The error covariance matrix of the self-predicted data at time k is then updated.

[0092] Constructing the state vector is the first step in Kalman filtering, representing the photovoltaic module's own data as a vector. The state transition matrix describes the change of the system state over time, generating current-moment prediction data based on historical self-data and environmental data, and updating the error covariance matrix to reflect the uncertainty of the prediction, providing a foundation for subsequent predictions and updates.

[0093] Based on the environmental data of the photovoltaic module at time k, the observation vector of the photovoltaic module at time k is constructed, and the observation matrix is ​​also constructed. The Kalman gain is then calculated by combining the updated error covariance matrix and the observation matrix.

[0094] The observation vector contains the actual measurements obtained from the environment, and the observation matrix links these measurements with the data of the photovoltaic module itself. Calculating the Kalman gain is the core of Kalman filtering. It determines the weight between prediction and observation, uses actual measurement data to correct the prediction, and ensures the accuracy of the prediction.

[0095] The Kalman gain and the observation vector at time k are used to update the self-predicted data at time k, and the updated self-predicted data at time k is obtained.

[0096] The predicted data is updated using Kalman gain and observation vectors. This step is the "correction" phase of Kalman filtering, which combines predictions and actual observations to obtain a more accurate state estimate. By fusing predictions and actual observations, the accuracy of the prediction is improved.

[0097] Using the updated self-predicted data at time k, and combining it with the state transition matrix, we construct the state equation from time k to time k+1, and obtain the predicted self-data at time k+1, i.e., component prediction data.

[0098] By using the updated current-time prediction data and combining it with the state transition matrix, a state equation from the current time to the future time is constructed, thereby obtaining the prediction data for the future time. Based on the validated system model and the updated state estimate, a reliable foundation is provided for future predictions.

[0099] Furthermore, the state vector of the photovoltaic module at time k-1 is:

[0100]

[0101] Where, x k―1 I represents the state vector of the photovoltaic module at time k-1. k―1 V k―1 and T k―1 These represent the output current, voltage, and surface temperature data of the photovoltaic module at time k-1, respectively.

[0102] Current, voltage, and temperature are key parameters for the operating state of photovoltaic (PV) modules. The output current of a PV module is a direct indicator of its power generation capacity. Current is significantly affected by light intensity and temperature, changing over time, and is one of the core variables for state estimation. Output voltage is also an important parameter for evaluating PV module performance; voltage variations are related to load, environmental conditions, and the module's internal impedance. The surface temperature of the PV module directly affects its performance. By combining these parameters into a state vector, the operating state of the PV module at a given moment can be comprehensively and accurately described. These state variables are not only interrelated but also jointly influence the performance of the PV module. Constructing the state vector is to effectively track and predict these key parameters using the Kalman filter algorithm during the state estimation process.

[0103] Since the state vector contains key performance indicators, it can comprehensively reflect the operating status of photovoltaic modules. By integrating multiple relevant variables into a single vector, the mathematical processing of state estimation is simplified, and it can provide reliable input for the prediction and correction steps of Kalman filtering.

[0104] The state transition matrix is:

[0105]

[0106] Wherein, F represents the state transition matrix, which is an identity matrix. It indicates that each state variable (current, voltage, and temperature) is independent in the time series and is assumed to be unaffected by other state variables within a time step. When the identity matrix is ​​used as the state transition matrix, it means that the predicted state value at the next moment is equal to the state value at the current moment. That is, without external intervention or further dynamic modeling, each state variable is assumed to be a temporal self-replication. Using the identity matrix simplifies the complexity of state estimation. When there is insufficient system dynamic information or insufficient data to support complex models, using the identity matrix can avoid introducing erroneous dynamic factors. The state transition matrix in this embodiment is used in the initial stage or in the verification stage, or when the system dynamics are relatively stable and do not change much.

[0107] Under normal conditions, the output voltage, current, and surface temperature of photovoltaic modules exhibit relatively stable characteristics. Within a short period, the light intensity remains relatively stable without sudden changes. This results in minimal variation in the amount of solar radiation received by the photovoltaic modules, leading to relatively stable output current and voltage levels. In specific seasons and geographical locations, the ambient temperature remains relatively constant, further minimizing surface temperature variations in the photovoltaic modules and enabling them to maintain stable output performance. Moreover, current photovoltaic systems are typically equipped with MPPT controllers, which continuously adjust the system operating point to match the maximum power point. This means that even with slight changes in environmental conditions, near-optimal operating conditions can be maintained, ensuring relatively stable output current and voltage. Therefore, the state transition matrix used in this embodiment is highly suitable for scenarios where the state of the photovoltaic modules does not change significantly.

[0108] The formula used to generate the self-predicted data at time k based on the photovoltaic module's own data at time k-1 and environmental data is as follows:

[0109] x k|k―1 =x k―1 *F

[0110] Where, x k|k―1 This represents the self-predicted data at time k based on the photovoltaic module's own data at time k-1 and environmental data.

[0111] As described above, photovoltaic modules have a certain degree of inherent stability. Therefore, the changes in current and voltage states are linear. That is, from one moment to the next, the state change can be described by a linear matrix, thus enabling rapid state prediction.

[0112] The formula used to update the error covariance matrix of the self-predicted data at time k is:

[0113] P k|k―1 =F*P k―1|k―1 *F T

[0114] Among them, P k|k―1 F represents the error covariance matrix of the self-predicted data at time k after the update. T P represents the transpose of the state transition matrix. k―1|k―1 Let represent the error covariance matrix of the self-predicted data at time k-1, and:

[0115]

[0116] Where, σ I σ U and σ TThese represent the tolerances for the current data, voltage data, and surface temperature data of the photovoltaic module, respectively.

[0117] In the Kalman filter prediction algorithm, updating the error covariance matrix is ​​based on the assumptions of a linear system and Gaussian noise. The update process is to predict the state uncertainty of the next time step and to pass this uncertainty into the model. The error covariance matrix represents the uncertainty or error covariance of the state estimate at time k-1. It captures the correlation of errors between state variables and the magnitude of their respective uncertainties. In photovoltaic module monitoring, the uncertainty of each state variable is quantified by the error covariance matrix. The diagonal elements in the matrix represent the variance of each state variable, i.e., the uncertainty of measurement or prediction, while the off-diagonal elements represent the error correlation between state variables. This implementation can effectively pass the state uncertainty to the next time step by updating the error covariance matrix of its own prediction data at time k, providing a mathematical basis for subsequent state estimation and updates.

[0118] The tolerances for the current and voltage data output by the photovoltaic module are obtained by consulting the module's instruction manual. The surface temperature data of the photovoltaic module is obtained by consulting the instruction manual of the sensor that collects the temperature values. When processing the output data of the photovoltaic module, σ... I 2 σ V 2 and σ T 2 The error covariance matrix placed on the diagonal of the self-predicted data at time k-1 intuitively represents the measurement uncertainty of current, voltage, and temperature at time k-1, i.e., the fluctuation or error of the measured values. The error covariance matrix can take into account the uncertainty in historical information and new observation data, thereby making a more accurate state estimate. During the update, the error covariance matrix is ​​used to combine with the new measurement data to reduce the uncertainty of the estimate.

[0119] Furthermore, the method used to calculate the Kalman gain is as follows:

[0120] K k =P k|k―1 *H T *(H*P k|k―1 *H T ) ―1

[0121] Among them, K k H represents the Kalman gain, and H represents the observation matrix. T Let represent the transpose of the observation matrix, and:

[0122]

[0123] The observation vector of the photovoltaic module at time k is:

[0124]

[0125] Among them, z k I represents the observation vector of the photovoltaic module at time k. k V k and T k S represents the output current, voltage, and surface temperature data of the photovoltaic module at time k, respectively. k E k and t k These represent the surface sunlight intensity, ambient temperature, and usage time of the photovoltaic module at time k, respectively.

[0126] The Kalman gain reflects how the estimation of the system state should be adjusted when given the observation vector of the photovoltaic module at time k. The error covariance matrix of the updated self-predicted data at time k represents the uncertainty of the current state estimate after considering the previous state and process model. The observation matrix defines how to map from the state vector to the observation vector, and its transpose is used for dimension matching and matrix operations. This embodiment dynamically adjusts the gain by considering the previous prediction uncertainty and the information of the current observation, thereby more accurately fusing the prediction and observation data.

[0127] In this embodiment, the observation matrix is ​​set to identify and select specific parts of the state vector corresponding to current, voltage, and temperature to ensure that the measurements of current, voltage, and temperature in the state vector can be directly mapped to the observation vector. The observation vector includes the current, voltage, and surface temperature data of the photovoltaic module at time k, as well as the sunlight intensity, ambient temperature, and usage time. This allows the Kalman filter algorithm to not only update the state estimate based on its own measured data, but also to integrate changes in environmental conditions.

[0128] The formula used to obtain the updated self-prediction data at time k is:

[0129] x k|k =x k|k―1 +K k (z k ―H*x k|k―1 )

[0130] Where, x k|k Let x represent the updated self-predicted data at time k. k|k The observation information z at time k was taken into consideration. k The latest estimate of the system state is then made. This update process involves two key steps: prediction and correction. In the prediction phase, x... k|k―1It is a prediction based on the state estimate of the previous time step and the dynamic model of the system. It represents the best estimate of the system state at time k before considering the latest observations. The correction stage involves changing the predicted state x. k|k―1 Compared with actual observation z k Compare the values ​​to adjust the state estimates. Difference z k ―H*x k|k―1 This represents the error between the observed value and the expected observed value based on the predicted state.

[0131] Kalman gain balances the degree of confidence in the predicted state with the degree of confidence in new observation data. If the Kalman gain is large, it indicates a high degree of confidence in the observation data, so the correction of the state estimate depends more on the observation error. Conversely, if the Kalman gain is small, it indicates a higher degree of confidence in the prediction model, so the state update depends less on the observation error.

[0132] Obtaining the updated self-prediction data at time k reflects an optimization process aimed at minimizing the uncertainty of the system state estimate. Therefore, the Kalman filter algorithm can dynamically adjust its correction of the state estimate at each time step.

[0133] The formula used to obtain the predicted data for time k+1, i.e., the component prediction data, is as follows:

[0134]

[0135] Where, x k+1|k Represents component prediction data, Iy k+1 Vy k+1 and Ty k+1 These represent the predicted current, voltage, and surface temperature data of the photovoltaic module at time k+1 in the module prediction data.

[0136] x k+1|k The best estimate of x based on the current time, i.e., time k. k|k To predict the state at the next time step, i.e., time k+1, the Kalman filter algorithm can not only provide an accurate estimate of the current state but also predict the future state of the system. k+1|k This represents the prediction of the state at time k+1, given the conditions at time k.

[0137] The data correction module collects the photovoltaic module's own data and environmental data at time k+1. Based on the photovoltaic module's own data and environmental data at time k and time k-1, it generates a first environmental impact index set and uses the first environmental impact index set to correct the photovoltaic module's own data at time k+1.

[0138] By considering the photovoltaic module's own data and environmental data at time k and k-1, it is possible to more accurately predict the photovoltaic module's own data at time k+1. Under different environmental conditions, the performance of the photovoltaic module will change to some extent, although the change is small. Therefore, the generation and application of the environmental impact index set enables the own data to be corrected according to the impact of environmental changes on the output of the photovoltaic module.

[0139] In this embodiment, the first environmental impact index set includes the sunlight intensity impact index, the temperature impact index, and the time impact index. The logic for generating the first environmental impact index set is as follows:

[0140]

[0141] Among them, Gq y Wd y and Sj y These represent the influence indices of sunlight intensity, temperature, and time, respectively. k V k and T k S represents the output current, voltage, and surface temperature data of the photovoltaic module at time k, respectively. k E k and t k Represent the surface solar intensity, ambient temperature, and time of use of the photovoltaic module at time k, respectively. k―1 V k―1 and T k―1 S represents the output current, voltage, and surface temperature data of the photovoltaic module at time k-1, respectively. k―1 E k―1 and t k―1 These represent the surface sunlight intensity, ambient temperature, and time of use of the photovoltaic module at time k-1, respectively.

[0142] In this embodiment, surface sunlight intensity data is used to correct the current data, ambient temperature data is used to correct the voltage data, and the usage time is used to correct the surface temperature data of the photovoltaic module. This is because the current output of the photovoltaic module mainly depends on the sunlight intensity. Increased light intensity usually leads to more photons hitting the solar cells, exciting more electrons, and thus generating a larger current. The voltage output of the photovoltaic module is closely related to the ambient temperature. Increased ambient temperature leads to a decrease in the band gap of the semiconductor material, thereby reducing the voltage output. Therefore, using ambient temperature to correct the voltage data can effectively calibrate voltage fluctuations caused by temperature changes. Long-term use of the photovoltaic module will affect its thermal characteristics because material aging may affect its thermal conductivity and heat capacity, thus affecting the surface temperature of the module. The length of time the module has been in use can be used as an indicator to measure the degree of aging of the module. Therefore, using it to correct the surface temperature data can help assess and adjust changes in thermal characteristics caused by usage time. The correction strategy adopted in this embodiment allows the specific effects of each environmental and operational factor to be specifically considered and calibrated.

[0143] In this embodiment, the influence index assesses the degree of impact of environmental factors on the performance of photovoltaic modules by calculating the rate of change between two consecutive time points. The solar intensity influence index is used to assess the impact of changes in solar intensity on the output current of photovoltaic modules by calculating the change in current I. k —I k―1 With the change in sunlight intensity S k —S k―1 The ratio of the solar intensity to the solar intensity index is used to obtain the change in current when the solar intensity changes by one unit. The solar intensity influence index reflects the sensitivity of the photovoltaic module to changes in sunlight.

[0144] The temperature effect index is used to assess the impact of changes in ambient temperature on the output voltage of photovoltaic modules. It is calculated by measuring the voltage change V. k ―V k―1 With temperature change E k —E k―1 The ratio of the voltage change to the surface temperature (T0) is used to determine the change in voltage per unit temperature change, reflecting the sensitivity of the photovoltaic module to temperature variations. The time-effect index is used to assess the impact of photovoltaic module usage time on surface temperature, calculated by measuring the surface temperature change T0. k ―T k―1 Change in usage time t k ―t k―1 The ratio of the two values ​​gives the change in surface temperature for each additional unit of usage time. This index reflects the change in the thermal properties of photovoltaic modules as usage time increases.

[0145] By calculating the rate of change, the dynamic impact of environmental factors on the performance of photovoltaic modules can be captured more accurately. This takes into account not only the direct impact of environmental factors but also the changing trends of these factors over time, thus providing a more precise basis for data correction.

[0146] Furthermore, the logic for correcting the photovoltaic module's own data at time k+1 using the first environmental impact index set is as follows:

[0147] Ix k+1 =I k+1 +Gq y *(S k+1 —S k―1 )

[0148] Vx k+1 =V k+1 +Wd y *(E k+1 —E k―1 )

[0149] Tx k+1 =T k+1 +Sj y *(t k+1 ―t k―1 )

[0150] Among them, Ix k+1 Vx k+1 and Tx k+1 I represents the corrected output current, voltage, and surface temperature data of the photovoltaic module at time k+1, respectively. k+1 V k+1 and T k+1 S represents the current, voltage, and surface temperature data of the photovoltaic module at time k+1, respectively. k+1 E k+1 and t k+1 These represent the surface sunlight intensity, ambient temperature, and time of use of the photovoltaic module at time k+1, respectively.

[0151] In this embodiment, known influence indices are used to adjust the photovoltaic module data at a future time, thereby obtaining data that more accurately reflects the actual working state. The current data at time k+1 is adjusted by using the solar intensity influence index and the difference in solar intensity at two times. The voltage data at time k+1 is adjusted by using the temperature influence index and the difference in ambient temperature at two times, taking into account the impact of ambient temperature changes on voltage output. The surface temperature data at time k+1 is adjusted by using the time influence index and the difference in usage time at two times, taking into account the potential impact of long-term use on surface temperature.

[0152] The current output of a photovoltaic (PV) module is primarily determined by the process of photons exciting electrons to generate current. As sunlight intensity increases, more photons are absorbed, producing more electron-hole pairs, thus increasing the current output. This relationship can be considered linear because the mechanism of photon absorption and current generation is stable under normal operating conditions. However, as ambient temperature rises, the band gap of the semiconductor material decreases, leading to a drop in voltage output. This temperature effect on voltage is typically described by a negative temperature coefficient, meaning that for every degree the temperature increases, the voltage decreases by a fixed percentage. Therefore, this relationship can also be considered linear within a certain temperature range.

[0153] The surface temperature of photovoltaic modules changes with the increase of usage time. Over a certain period of time (e.g., several months to several years), the change in surface temperature is proportional to the change in usage time, which can be approximated as a linear relationship. Therefore, by using various influence indices from previous periods, data for future moments can be corrected to obtain performance that is closer to the actual situation.

[0154] The fluctuation judgment module generates its own data error index based on the corrected self-data of the photovoltaic module at time k+1 and the module prediction data. It uses the first environmental impact dataset to correct its own data at time k. Based on its own data at time k-1, the corrected self-data at time k, and its own data at time k+1, it generates its own data volatility.

[0155] Furthermore, the formula used to correct the data at time k using the first environmental impact dataset is as follows:

[0156] Ix k =I k +Gq y *(S k —S k―1 )

[0157] Vx k =V k +Wd y *(E k —E k―1 )

[0158] Tx k =T k +Sj y *(t k ―t k―1 )

[0159] Among them, Ix k Vx k and Tx k These represent the corrected current, voltage, and surface temperature data of the photovoltaic module at time k, respectively.

[0160] The correction of the photovoltaic module's output current, voltage, and surface temperature data at time k is also achieved by using the changes in environmental data from the previous time step to correct the current photovoltaic module data, thereby obtaining more accurate current, voltage, and surface temperature to compensate for the instantaneous impact of environmental factors on the photovoltaic module's performance.

[0161] A photovoltaic module status monitoring system is constructed by introducing its own data, environmental data, and Kalman filtering algorithm. By collecting the photovoltaic module's own data and environmental data, and using the Kalman filtering algorithm through the data prediction module to predict the future state of the module, the system obtains the module's predicted data by predicting the self-data of the photovoltaic module at the next moment. The environmental impact index introduced by the data correction module can effectively distinguish the causes of photovoltaic module performance degradation, and promptly distinguish whether the performance degradation is caused by external environmental factors or problems within the module itself, thus providing a more accurate basis for decision-making in the operation and maintenance monitoring of photovoltaic modules.

[0162] In this embodiment, when generating the self-data error index, the errors of the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data in the corrected self-data and module prediction data at time k+1 are assigned corresponding weights, and the self-data error index is calculated by summing them. The formula for generating the self-data error index is as follows:

[0163]

[0164] Among them, ZW z Ix represents the data error index. k+1 Vx k+1 and Tx k+1 These represent the corrected output current, voltage, and surface temperature data of the photovoltaic module at time k+1, respectively. k+1 Vy k+1 and Ty k+1 Let α, β, and γ represent the current, voltage, and surface temperature data of the photovoltaic module at time k+1 in the module prediction data, respectively. Let α, β, and γ be the current anomaly weights, voltage anomaly weights, and temperature anomaly weights, respectively. Let α > β > γ > 0, and α + β + γ = 1.

[0165] By calculating the corrected data Ix k+1 Vx k+1 and Tx k+1 With the predicted data Iy k+1 Vy k+1 and Ty k+1The absolute difference between the actual performance and the prediction is used to measure the deviation between the actual performance and the prediction. The absolute difference of each item is divided by the minimum value of these items to convert it into a relative error, so as to compare the accuracy of each item more fairly.

[0166] By assigning different weights to the relative errors of current, voltage, and surface temperature—weights that reflect the importance of different types of data in assessing the accuracy of photovoltaic module performance—the weighted relative errors of current, voltage, and surface temperature are summed to generate a data error index.

[0167] The self-data error index uses predicted data as a reference standard to assess whether the actual state of photovoltaic modules is normal. The larger the index, the more abnormal the state of the photovoltaic modules. The predicted data is calculated based on historical self-data and environmental conditions, and represents the expected performance of photovoltaic modules at a specific time. Therefore, the predicted data can serve as a reasonable benchmark for comparing with the actual observed data.

[0168] The self-corrected data error index aims to promptly identify and assess performance anomalies in photovoltaic modules by comparing actual and predicted data. If the corrected actual data (Ix) is not corrected... k+1 Vx k+1 and Tx k+1 With measurement data Iy k+1 Vy k+1 and Ty k+1 Consistency indicates that the performance of the module meets expectations and is in normal condition. If there is a significant deviation between the corrected actual data and the predicted data, it may indicate that the performance of the module has been affected by some abnormal factor and that an abnormal condition may have occurred. Therefore, the larger the data error index, the more abnormal the failure of the photovoltaic module may be.

[0169] Relative error is used instead of absolute error because relative error better reflects the percentage change in data, making it fairer for comparisons of data of different magnitudes. For example, the absolute values ​​of current, voltage, and surface temperature can vary greatly, and directly comparing absolute errors may not accurately reflect actual performance deviations. Different performance indicators have varying degrees of impact on the overall performance of photovoltaic modules. Current is a direct factor in the energy output of a photovoltaic system. The magnitude of the current directly affects the electrical energy generated by the photovoltaic panel; therefore, the accuracy of the current reading is extremely important for system performance. Voltage changes have a smaller direct impact on energy output than current. Voltage is a key factor affecting system stability and safety; voltage stability is crucial for maintaining normal system operation and preventing electrical accidents such as overloads or short circuits. Although surface temperature has a smaller impact on immediate power output, high temperatures reduce the efficiency of photovoltaic modules and accelerate aging, affecting the long-term stability and lifespan of the modules.

[0170] Therefore, current is the most important because it is directly related to energy output, followed by voltage. Although surface channel temperature is more indirect, it is equally critical for long-term operation and is therefore given the least weight. Thus, α>β>γ. This weighting helps to more accurately assess and monitor the performance status of photovoltaic systems.

[0171] The monitoring and analysis module analyzes its own data volatility and data error index, establishes a data analysis model, generates a status monitoring index, compares the status monitoring index with a threshold, and selects whether to issue an early warning signal based on the comparison result.

[0172] In this embodiment, when generating the self-data volatility based on the self-data at time k-1, the corrected self-data at time k, and the self-data at time k+1, the variances of the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data at the three times are calculated respectively. The self-data volatility is obtained by calculating the average of the three sets of variances. Specifically:

[0173]

[0174] Among them, Zs b For its own data volatility, Ix k+1 Vx k+1 and Tx k+1 Represent the corrected output current, voltage, and surface temperature data of the photovoltaic module at time k+1, respectively. Ix k Vx k and Tx k Represent the corrected output current, voltage, and surface temperature data of the photovoltaic module at time k, respectively. k―1 V k―1 and T k―1 These represent the output current, voltage, and surface temperature data of the photovoltaic module at time k-1, respectively. and Let represent the average current, average voltage, and average surface temperature, respectively, and:

[0175]

[0176] In this embodiment, the data at time k-1, k, and k+1 are used to evaluate the performance stability of the photovoltaic module. By analyzing the data fluctuations of current, voltage, and surface temperature, information about the performance changes of the module within a specific time period can be obtained. By selecting data at three times, the short-term fluctuation characteristics of the photovoltaic module performance in the time series can be captured. The square of the difference between the current, voltage, and temperature data at each time point and its corresponding average value is calculated, and then summed and averaged. This is actually a variation of calculating variance, used to measure the dispersion or volatility of the data.

[0177] By considering data from three consecutive moments, short-term fluctuations in the output of photovoltaic modules can be captured more sensitively. Since the volatility of the data is generated by using the data from the past k-1 moment, the corrected data from the k moment, and the data from the k+1 moment, the influence of environmental factors has been eliminated. This fluctuation is determined to be caused by a fault in the photovoltaic module itself.

[0178] The greater the volatility of the photovoltaic module's own data, the greater the performance variation of the photovoltaic module at adjacent time points, indicating a worse condition. It may indicate that there is a fault inside the photovoltaic module, resulting in unstable performance. For example, a greater volatility of the photovoltaic module's own data may indicate that the fluctuation of current or voltage exceeds the normal range, which may indicate circuit problems, module damage or connection problems, resulting in a worse condition of the photovoltaic module.

[0179] The formula used to generate the status monitoring index is:

[0180]

[0181] Among them, ZJ z This represents the status monitoring index. When comparing the status monitoring index with a threshold, if ZJ... z ≥ZJ yz If ZJ z <ZJ yz No warning signal will be issued.

[0182] The condition monitoring index assesses the overall condition of photovoltaic modules by using its own data error index and its own data volatility. The larger the condition monitoring index, the more likely there are problems or anomalies in the operation of the photovoltaic modules. As the error index increases, the condition monitoring index will also increase, reflecting the deterioration of module performance. As the volatility increases, the condition monitoring index will also increase, indicating that the output of the module is more unstable, which may point to unstable module performance or potential faults.

[0183] The self-data error index is fitted using a logarithmic function because, while larger error data are important, their growth does not linearly affect the overall index. This avoids the overall assessment result being dominated by a single or a few large error values. Moreover, the logarithmic function is particularly suitable for amplifying small values ​​close to zero, allowing the condition monitoring index to react sensitively to small self-data error indices. Therefore, a logarithmic function is used to reflect the impact of the self-data error index on the condition monitoring index. The self-data volatility is fitted using an exponential function because higher volatility usually means higher risk or instability. Through exponential transformation, the impact of larger volatility is significantly amplified, helping to give it sufficient attention in the assessment. Small fluctuations in voltage, current, etc., will not have a significant impact on photovoltaic modules, while large fluctuations may cause serious problems. The growth rate of the exponential function is much faster than linear, which allows the condition monitoring index to respond quickly and provide rapid warnings in the face of extreme fluctuations.

[0184] The predicted data is used as a reference standard to generate its own data error index, which is used to assess whether the actual state of the photovoltaic module is normal. The self-data volatility is generated using the corrected self-data at three time points and the module prediction data, which reflects the performance changes of the photovoltaic module at adjacent time points. The self-data error index and self-data volatility are used for analysis.

[0185] By combining the data analysis model of the monitoring and analysis module, a status monitoring index can be generated and compared with a preset threshold to issue an early warning signal, which improves the accuracy of photovoltaic system fault detection and reduces the workload of manual analysis and the uncertainty of subjective judgment.

[0186] In this embodiment, time k+1 is the current time, the time length between time k and time k-1 is the same as the time length between time k and time k+1, and the time interval between time k-1, time k and time k+1 is the same, which means that the sampling frequency is consistent, such as every hour, every day, every second, etc.

[0187] Please see Figure 2 The present invention also provides a photovoltaic module status monitoring method based on multi-source data. The monitoring method is executed by the aforementioned photovoltaic module status monitoring system, and the specific steps include:

[0188] Step 1: Collect the photovoltaic module's own data and environmental data at time k and time k-1. The own data includes the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data. The environmental data includes the photovoltaic module surface sunlight intensity data, the ambient temperature data, and the time of use.

[0189] Step 2: Based on the photovoltaic module's own data and environmental data at time k and k-1, the Kalman filter algorithm is used to predict the photovoltaic module's own data at time k+1 to obtain the module prediction data. The module prediction data includes the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data at time k+1.

[0190] Step 3: Collect the photovoltaic module's own data and environmental data at time k+1. Based on the photovoltaic module's own data and environmental data at time k and time k-1, generate a first environmental impact index set. Use the first environmental impact index set to correct the photovoltaic module's own data at time k+1.

[0191] Step 4: Generate the self-data error index based on the corrected self-data of the photovoltaic module at time k+1 and the module prediction data. Use the first environmental impact dataset to correct the self-data at time k. Generate the self-data volatility based on the self-data at time k-1, the corrected self-data at time k, and the self-data at time k+1.

[0192] Step 5: Analyze the volatility and error index of the data itself, establish a data analysis model, generate a status monitoring index, compare the status monitoring index with the threshold, and choose whether to issue a warning signal based on the comparison results.

[0193] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0194] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0195] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0196] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A photovoltaic module status monitoring system based on multi-source data, characterized in that, include: The data acquisition module is used to collect data from photovoltaic modules. Time and The photovoltaic module's own data and environmental data at any given time. The own data includes the current data, voltage data, and surface temperature data of the photovoltaic module. The environmental data includes the surface sunlight intensity data of the photovoltaic module, the temperature data of the surrounding environment, and the time since it was put into use. The data prediction module is used to predict data based on photovoltaic modules. Time and The photovoltaic module's own data and environmental data at any given time are used to apply the Kalman filter algorithm to the photovoltaic module. The component prediction data is obtained by predicting the component's own data at any given time, and the component prediction data includes... The current data, voltage data, and surface temperature data of the photovoltaic module at any given time; The data correction module is used to collect data from photovoltaic modules. Real-time data from the photovoltaic module itself and environmental data. Time and Based on real-time self-data and environmental data, a first environmental impact index set is generated. This first environmental impact index set is then used to evaluate photovoltaic modules. The data at any given moment is corrected. The fluctuation judgment module is used to determine the fluctuation based on the corrected photovoltaic module. The self-data and component prediction data at each moment generate their own data error index, and the first environmental impact dataset is used for... Correcting the data based on the time frame and the past. The data of the moment itself, the corrected Time and The data at any given moment generates its own volatility. The monitoring and analysis module is used to analyze its own data volatility and data error index, establish a data analysis model, generate a status monitoring index, compare the status monitoring index with a threshold, and select whether to issue an early warning signal based on the comparison result. The The time is the current time. Time and The length of time between moments and Time and The time intervals between moments are equal. The first environmental impact index set includes the solar intensity impact index, temperature impact index, and time impact index. The logic for generating the first environmental impact index set is as follows: in, , and These represent the influence indices of sunlight intensity, temperature, and time, respectively. , and Representing photovoltaic modules The real-time output current data, voltage data, and photovoltaic module surface temperature data of the photovoltaic module. , and These respectively represent the photovoltaic modules in Data on surface sunlight intensity, ambient temperature, and time since commissioning. , and Representing photovoltaic modules The real-time output current data, voltage data, and photovoltaic module surface temperature data of the photovoltaic module. , and These respectively represent the photovoltaic modules in Data on surface sunlight intensity, ambient temperature, and time of use at any given moment; Using the first environmental impact index set for photovoltaic modules The logic for correcting the data at any given time step is as follows: in, , and They represent the corrected versions respectively. The real-time output current data, voltage data, and photovoltaic module surface temperature data of the photovoltaic module. , and These represent the collected photovoltaic modules. The real-time output current data, voltage data, and photovoltaic module surface temperature data of the photovoltaic module. , and These respectively represent the photovoltaic modules in Data on surface sunlight intensity, ambient temperature, and time of use at any given moment; Using the first environmental impact dataset for The formula used to correct the data at any given time is: in, , and They represent the corrected versions respectively. The current data, voltage data, and surface temperature data of the photovoltaic module at any given time; When generating its own data error index, the corrected photovoltaic modules are processed separately. In the self-data and component prediction data at any given time, the errors in the photovoltaic module's output current data, voltage data, and photovoltaic module surface temperature data are assigned corresponding weights, summed, and then the self-data error index is calculated. The formula for generating the self-data error index is as follows: in, This indicates the data error index. , and They represent the corrected versions respectively. The real-time output current data, voltage data, and photovoltaic module surface temperature data of the photovoltaic module. , and These represent the predicted values ​​in the component prediction data. The real-time output current data, voltage data, and photovoltaic module surface temperature data of the photovoltaic module. , and These are the weights for current anomalies, voltage anomalies, and temperature anomalies, respectively. ,and ; Based on the past The data of the moment itself, the corrected Time and To generate the self-data volatility, the variances of the photovoltaic module's output current, voltage, and surface temperature data at three different times are calculated. The self-data volatility is obtained by averaging the three variances. Specifically: in, For its own data volatility, , and They represent the corrected versions respectively. The real-time output current data, voltage data, and photovoltaic module surface temperature data of the photovoltaic module. , and They represent the corrected versions respectively. The real-time output current data, voltage data, and photovoltaic module surface temperature data of the photovoltaic module. , and Representing photovoltaic modules The real-time output current data, voltage data, and photovoltaic module surface temperature data of the photovoltaic module. , and Let represent the average current, average voltage, and average surface temperature, respectively, and: The formula used to generate the status monitoring index is: in, This represents the status monitoring index. When comparing the status monitoring index with a threshold, if... If so, a warning signal is issued indicating an abnormality in the current photovoltaic module. If so, no warning signal will be issued; Using the Kalman filter algorithm for photovoltaic modules The method used to obtain component prediction data by using the data of the time itself is as follows: Based on photovoltaic modules Photovoltaic modules are constructed using real-time data. The state vector at time step 1, combined with the state transition matrix, constructs the state vector from... Time's up The state equations for time transitions are generated based on photovoltaic modules. Real-time self-data and environmental data The self-predicted data at each moment is updated. The error covariance matrix of the self-predicted data at time step; According to photovoltaic modules Real-time environmental data to construct photovoltaic modules The observation vector at time t is used to construct the observation matrix, and the Kalman gain is calculated by combining the updated error covariance matrix and the observation matrix. Using Kalman gain, The observation vector at time t is for Update the self-predicted data at any given time and obtain the updated data. The self-predictive data at any given moment; Use the updated version The self-predicted data at each time step, combined with the state transition matrix, constructs a system from... Time's up The state equations for time transitions are used to obtain predictions. The data at any given moment, i.e., the component prediction data.

2. The photovoltaic module status monitoring system based on multi-source data according to claim 1, characterized in that: photovoltaic modules The state vector at time t is: in, Indicates photovoltaic modules The state vector at time t, , and Representing photovoltaic modules The current data, voltage data, and surface temperature data of the photovoltaic module at any given time; The state transition matrix is: in, Represents the state transition matrix; Generate photovoltaic module-based Real-time self-data and environmental data The formula used to predict the self-prediction data at any given time is: in, Indicates based on photovoltaic modules Real-time self-data and environmental data The self-predictive data at any given moment; renew The formula used to determine the error covariance matrix of the self-predicted data at time step is: in, Indicates the updated The error covariance matrix of the self-predicted data at time step, This represents the transpose of the state transition matrix. express The error covariance matrix of the self-predicted data at time t, and: in, , and These represent the tolerances for the current data, voltage data, and surface temperature data of the photovoltaic module, respectively.

3. The photovoltaic module status monitoring system based on multi-source data according to claim 2, characterized in that: The method used to calculate the Kalman gain is as follows: in, Indicates Kalman gain, Represents the observation matrix. Let represent the transpose of the observation matrix, and: The photovoltaic module The observation vector at time t is: in, Indicates photovoltaic modules The observation vector at time t, , and Representing photovoltaic modules The real-time output current data, voltage data, and photovoltaic module surface temperature data of the photovoltaic module. , and These respectively represent the photovoltaic modules in Data on surface sunlight intensity, ambient temperature, and time of use at any given moment; Get the updated The formula used to predict the self-prediction data at any given time is: in, Indicates the updated The prediction is obtained from the self-predictive data at any given time. The formula used for the component's own data at any given time, i.e., the component's prediction data, is as follows: in, This represents the component's predicted data. , and These represent the predicted values ​​in the component prediction data. The data includes the current output, voltage output, and surface temperature of the photovoltaic module at any given time.

4. A photovoltaic module status monitoring method based on multi-source data, characterized in that: The monitoring method is performed by the photovoltaic module status monitoring system according to any one of claims 1-3, and the specific steps include: Step 1: Collect photovoltaic module data Time and The photovoltaic module's own data and environmental data at any given time. The own data includes the current data, voltage data, and surface temperature data of the photovoltaic module. The environmental data includes the sunlight intensity data on the surface of the photovoltaic module, the temperature data of the surrounding environment, and the time since it was put into use. Step 2: Based on photovoltaic modules Time and The photovoltaic module's own data and environmental data at any given time are used to apply the Kalman filter algorithm to the photovoltaic module. The component prediction data is obtained by predicting the component's own data at any given time, and the component prediction data includes... The current data, voltage data, and surface temperature data of the photovoltaic module at any given time; Step 3: Collect photovoltaic modules Real-time data of the photovoltaic module itself and environmental data. Time and Based on real-time self-data and environmental data, a first environmental impact index set is generated. This first environmental impact index set is then used to evaluate photovoltaic modules. The data at any given moment is corrected. Step 4: Based on the revised photovoltaic modules The self-data and component prediction data at each moment generate their own data error index, and the first environmental impact dataset is used for... The data at any given moment is corrected based on past data. The data of the moment itself, and the corrected data Time and The data at any given moment generates its own volatility. Step 5: Analyze the volatility and error index of the data itself, establish a data analysis model, generate a status monitoring index, compare the status monitoring index with the threshold, and choose whether to issue a warning signal based on the comparison results.

Images