Intelligent monitoring method and system for photovoltaic module
By calculating the horizontal and vertical difference coefficients and abnormal weights of photovoltaic modules, combined with the angle-based anomaly factor, intelligent monitoring of photovoltaic modules is achieved, solving the problem of small differences caused by difficult to distinguish environmental factors from component abnormalities in the existing technology, and improving Monitoring accuracy and sensitivity.
Patent Information
- Application Number
- CN202411042541.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-07-31
AI Technical Summary
Existing intelligent monitoring technology for photovoltaic modules is difficult to accurately distinguish between environmental factors and small differences caused by component abnormalities, resulting in the failure to detect potential abnormalities in time and affect the life of the component.
By obtaining the current, output power and meteorological data of the photovoltaic module, calculating the horizontal difference coefficient and the longitudinal difference coefficient, combining the abnormal weight and angle-based anomaly factors, intelligent monitoring and abnormal detection of the photovoltaic module are achieved.
It improves the accuracy and sensitivity of the abnormality of photovoltaic modules, can detect small differences in time, and extend the life of the module.
Smart Images

Figure CN119010787A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of photovoltaic module monitoring technology, specifically to a method and system for intelligent monitoring of photovoltaic modules. Background Technology
[0002] With the transformation of the global energy structure, photovoltaic (PV) power generation, as a clean and renewable energy source, is finding increasingly wider applications. As the core component of a PV power generation system, the performance of PV modules directly affects the overall power generation efficiency of the system. Therefore, monitoring PV modules has become a key research focus in the field of PV power generation.
[0003] In existing intelligent monitoring of photovoltaic (PV) modules, two common methods are used: one is to construct a simulation model of the PV system and determine whether there are PV module anomalies based on the difference between the simulated values and the actual values; the other is to determine whether there are PV module anomalies based on the difference between the monitored data and the historical data of the PV module. Both methods are highly dependent on the historical data of the PV module. There will always be differences between the monitored data and the historical data. The reasons for these differences may be due to environmental factors or anomalies in the PV module itself, but they are difficult to distinguish. In current technology, the approach to handling these differences is to determine that there are no anomalies in the PV module when the difference is small. However, if these small differences are caused by anomalies in the PV module itself, if they are not addressed in time and are only dealt with when significant differences appear, the accumulation of these small anomalies can have an irreversible impact on the lifespan of the PV module itself. Summary of the Invention
[0004] To address the aforementioned technical problems, a method and system for intelligent monitoring of photovoltaic modules are provided to resolve existing issues.
[0005] The solution to the technical problem in this application is to provide a method and system for intelligent monitoring of photovoltaic modules, including the following steps:
[0006] In a first aspect, embodiments of this application provide a method for intelligent monitoring of photovoltaic modules, the method comprising the following steps:
[0007] Acquire the current and output power of each photovoltaic module at each time, as well as meteorological environmental data at each time, including light intensity, ambient temperature, wind speed and humidity;
[0008] Based on the differences in current, output power, and seasonal cyclical characteristics of output power among different photovoltaic modules at different times, the horizontal difference coefficient of each photovoltaic module at each time is determined.
[0009] Based on the differences in current and output power of each photovoltaic module at each time point compared to the same time point of each historical day, and the impact of meteorological environmental data on output power, the longitudinal difference coefficient of each photovoltaic module at each time point is determined.
[0010] Based on the horizontal difference coefficient and the vertical difference coefficient, the anomaly weight of each photovoltaic module at each time point is determined;
[0011] Based on the aforementioned anomaly weights, the angle-based anomaly factor for each photovoltaic module at each time point is determined, and anomaly monitoring is performed on the photovoltaic modules.
[0012] Preferably, determining the lateral difference coefficient of each photovoltaic module at each time point includes:
[0013] Analyze the differences in seasonal cyclical characteristics between each photovoltaic module and the rest of the photovoltaic modules, as well as the correlation between current and output power, to determine the first difference between each photovoltaic module and the rest of the photovoltaic modules;
[0014] Analyze the correlation between the current and output power of each photovoltaic module at each time point and the other photovoltaic modules at the same time point to determine the first similarity between each photovoltaic module and the other photovoltaic modules at each time point;
[0015] The ratio of the first difference between each photovoltaic module and the other photovoltaic modules to the first similarity is denoted as the first ratio.
[0016] The first ratio of each photovoltaic module to all other photovoltaic modules at each time point is combined to determine the horizontal difference coefficient of each photovoltaic module at each time point.
[0017] Preferably, determining the first difference between each photovoltaic module and the remaining photovoltaic modules includes:
[0018] The seasonality intensity of each photovoltaic module is calculated by using a sequence decomposition algorithm for the output power of each photovoltaic module at all times.
[0019] The difference in seasonal intensity between each photovoltaic module and the rest of the photovoltaic modules is taken as the first difference between each photovoltaic module and the rest of the photovoltaic modules.
[0020] Preferably, determining the first similarity between each photovoltaic module and the other photovoltaic modules at each time point includes:
[0021] The current and output power of each photovoltaic module at each time moment are used to form the photovoltaic output vector of each photovoltaic module at each time moment;
[0022] The similarity between each photovoltaic module and the other photovoltaic modules at the same time is calculated, and the result of the calculation of the exponential function with the natural constant as the base and the similarity as the exponent is used as the first similarity between each photovoltaic module and the other photovoltaic modules at each time.
[0023] Preferably, determining the longitudinal difference coefficient of each photovoltaic module at each time point includes:
[0024] The ratio between the sum of light intensity and wind speed at each moment and the sum of ambient temperature and humidity is used as the environmental impact coefficient at each moment.
[0025] Analyze the correlation between the output power of each photovoltaic module and the environmental impact coefficient at all previous times to determine the second similarity of each photovoltaic module at each time.
[0026] Divide all moments into moments of each day according to their respective dates, and calculate the mean of the similarity between each moment and the same moment of all historical days for each photovoltaic module's photovoltaic output vector;
[0027] The product of the mean and the second similarity is denoted as the first product; the reciprocal of the result of the exponential function with the natural constant as the base and the first product as the exponent is used as the longitudinal difference coefficient of each photovoltaic module at each time point.
[0028] Preferably, determining the second similarity of each photovoltaic module at each time point includes:
[0029] The output power sequence and environmental impact sequence for each photovoltaic module at each time point are respectively composed of the output power of each photovoltaic module at all times before each time point and the environmental impact coefficient;
[0030] The second similarity of each photovoltaic module at each time point is determined based on the degree of correlation between the output power sequence and the environmental impact sequence.
[0031] Preferably, the expression for determining the anomaly weight of each photovoltaic module at each time point is: Among them, W i,t Let A be the outlier weight of the i-th photovoltaic module at time t. i,t Let C be the lateral difference coefficient of the i-th photovoltaic module at time t. i,t denoted as the longitudinal difference coefficient of the i-th photovoltaic module at time t, w0 is the preset initial weight, and α is the preset control coefficient.
[0032] Preferably, determining the angle-based anomaly factor for each photovoltaic module at each time point includes:
[0033] The current, output power, light intensity, ambient temperature, wind speed and humidity of each photovoltaic module at each time moment are used to form the anomaly monitoring vector of each photovoltaic module at each time moment.
[0034] The abnormal weights of each photovoltaic module at all times are normalized. The abnormal monitoring vectors of each photovoltaic module at each time and all previous times, along with the normalized abnormal weights, are input into the ABOD anomaly detection algorithm to obtain the angle-based anomaly factor of each photovoltaic module at each time.
[0035] Preferably, the abnormal monitoring of photovoltaic modules includes:
[0036] A threshold segmentation algorithm is applied to the angle-based anomaly factor of each photovoltaic module at all times to obtain the optimal threshold;
[0037] If the angle-based anomaly factor is less than or equal to the optimal threshold, the photovoltaic module has an anomaly; otherwise, the photovoltaic module does not have an anomaly.
[0038] Secondly, embodiments of this application also provide a photovoltaic module intelligent monitoring system, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of any of the above-described photovoltaic module intelligent monitoring methods.
[0039] This application has at least the following beneficial effects:
[0040] This application analyzes the differences in current, output power, and seasonal cyclical characteristics of output power among different photovoltaic modules at various times to determine the horizontal difference coefficient of each photovoltaic module at each time. Its advantage lies in considering the differences between different photovoltaic modules within the same external environment at the same time, thereby eliminating errors caused by environmental factors. Based on the differences in current and output power of each photovoltaic module at each time and the same time of each historical day, as well as the influence of meteorological environment on output power, the application determines the vertical difference coefficient of each photovoltaic module at each time. Its advantage lies in considering the differences in current and output power changes of the same photovoltaic module at different times, which helps improve the judgment of abnormal conditions in photovoltaic modules. Based on the horizontal and vertical difference coefficients, the application determines the anomaly weight of each photovoltaic module at each time. Based on the anomaly weight, the application determines the angle-based anomaly factor of each photovoltaic module at each time for anomaly monitoring. Its advantage lies in measuring the degree of anomaly in photovoltaic modules, considering both horizontal and vertical differences, enhancing the ability to capture various differences in photovoltaic module monitoring, detecting subtle differences in monitoring data, and improving the sensitivity and accuracy of photovoltaic module anomaly investigation. Attached Figure Description
[0041] The following description, in conjunction with the accompanying drawings, provides a further detailed explanation of an intelligent monitoring method for photovoltaic modules according to this application.
[0042] Figure 1 A flowchart illustrating the steps of an intelligent monitoring method for photovoltaic modules provided in this application embodiment;
[0043] Figure 2 A flowchart illustrating the steps of the method for obtaining the lateral difference coefficient of each photovoltaic module at various times, as provided in the embodiments of this application;
[0044] Figure 3 A flowchart illustrating the steps of the method for obtaining the longitudinal difference coefficient of each photovoltaic module at various times, as provided in the embodiments of this application. Detailed Implementation
[0045] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of a photovoltaic module intelligent monitoring method and system proposed in this application, in conjunction with the accompanying drawings and implementation examples, is provided. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit the scope of this application.
[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.
[0047] Please see Figure 1 The diagram illustrates a flowchart of a photovoltaic module intelligent monitoring method according to an embodiment of this application. The method includes the following steps:
[0048] Step 1: Obtain the current and output power of each photovoltaic module at each time, as well as the meteorological environment data at each time, including light intensity, ambient temperature, wind speed and humidity.
[0049] By installing current transformers on the output lines of each photovoltaic (PV) module and power meters at the output terminals of each module, the DC current and output power of each PV module are collected in real time. Then, environmental data during PV operation, including irradiance, ambient temperature, wind speed, and humidity, are collected in real time using radiometers, temperature sensors, anemometers, and humidity sensors at intervals of t. All collected data is uploaded to the PV module intelligent monitoring system for storage. The system retrieves the current, output power, irradiance, ambient temperature, wind speed, and humidity of each PV module at various times.
[0050] Preferably, in this embodiment, the data collection time interval is 10 minutes. As for other implementation methods, the implementer can set it according to the actual situation.
[0051] At this point, the current and output power of each photovoltaic module at each moment, as well as the light intensity, ambient temperature, wind speed, and humidity at each moment, are obtained.
[0052] Step 2: Determine the horizontal difference coefficient of each photovoltaic module at each time point based on the differences in current, output power, and seasonal cyclical characteristics of output power among different photovoltaic modules at each time point.
[0053] In intelligent monitoring of photovoltaic (PV) modules, errors caused by minor anomalies within the PV modules themselves can be determined by observing the differences between different PV modules at the same time. However, due to the inherent conditions of different PV modules, such as shading or dirt, the current and output power of different PV modules are affected. There are many reasons for changes in the current and output power of PV modules, and it is impossible to completely determine the anomaly of PV modules solely based on the differences in current and output power between different PV modules. A multi-dimensional analysis is required. Therefore, it is necessary to conduct horizontal comparisons based on the differences between different PV modules, and then combine this with a comprehensive analysis of the vertical differences between each PV module and its historical time, in order to determine whether an anomaly exists in the PV modules.
[0054] Under normal circumstances, the operating status of all photovoltaic (PV) modules in a photovoltaic (PV) system is related to changes in the environment, ensuring that the current and output power of each PV module should vary relatively consistently. Furthermore, since the power generation of PV modules is affected by sunlight intensity, which exhibits seasonal cyclical characteristics, the output power of PV modules also varies periodically. However, when the operating status of PV modules is not entirely uniform, they may be affected by variable shading (such as tree shadows or building shadows), constant shading (such as leaves or bird droppings), or global anomalies (such as dust), resulting in differences between different PV modules. Variable shading is mostly caused by shadows and thus exhibits periodic changes, while constant shading is more random and can disrupt this periodicity to some extent.
[0055] Based on the above analysis, the differences in current and output power among different photovoltaic modules are analyzed, and a horizontal difference coefficient is constructed, specifically as follows:
[0056] The seasonality intensity of each photovoltaic module is calculated by using a sequence decomposition algorithm for the output power of each photovoltaic module at all times.
[0057] Preferably, in this embodiment, the STL time series decomposition algorithm is used, wherein the seasonal cycle length parameter is 144. The STL time series decomposition algorithm and the calculation of seasonal intensity are well-known techniques and will not be described in detail here. As other implementation methods, implementers may use other methods of the prior art, such as the X-11 sequence decomposition algorithm, etc. This embodiment does not impose any special restrictions on this.
[0058] It should be noted that since the output power of photovoltaic modules is related to light intensity, and light intensity has a diurnal periodicity (meaning that the light intensity and its changes are similar or identical at most times of the day), data can be collected at 144 times per day. Therefore, the seasonal period length parameter in the STL time series decomposition algorithm is set to 144. Secondly, the seasonal intensity of each photovoltaic module is calculated, including: applying a sequence decomposition algorithm to the output power of each photovoltaic module at all times to obtain the seasonal series and residual series. The formula for calculating the seasonal intensity is: Among them, F T For trend strength, Var(R) t ) represents the variance of the residual sequence, Var(R) S +T t ) represents the variance of the seasonal sequence and the residual sequence, and max() is used to find the maximum value.
[0059] The difference in seasonal intensity between each photovoltaic module and the rest of the photovoltaic modules is taken as the first difference between each photovoltaic module and the rest of the photovoltaic modules;
[0060] The current and output power of each photovoltaic module at each time moment are used to form the photovoltaic output vector of each photovoltaic module at each time moment;
[0061] Calculate the similarity between the photovoltaic output vector of each photovoltaic module at each time step and the other photovoltaic modules at the same time step, and use the result of the calculation of the exponential function with the natural constant as the base and the similarity as the exponent as the exponent, as the first similarity between each photovoltaic module and the other photovoltaic modules at each time step.
[0062] The ratio of the first difference to the first similarity between each photovoltaic module and the other photovoltaic modules is denoted as the first ratio.
[0063] The first ratio of each photovoltaic module to all other photovoltaic modules at each time point is combined to determine the horizontal difference coefficient of each photovoltaic module at each time point;
[0064] It is understood that the specific relationship of fusion can be an additive relationship, a multiplicative relationship, etc., and the specific relationship is determined according to the actual scenario. This embodiment does not impose any special restrictions on this.
[0065] Preferably, in this embodiment, the first similarity is obtained by calculating the cosine similarity between the photovoltaic output vector of each photovoltaic module at each time and the other photovoltaic modules at the same time. The calculation method of cosine similarity is a well-known technique and will not be described in detail here.
[0066] In this embodiment, the method for calculating the lateral difference coefficient of each photovoltaic module at each time point is as follows: Among them, A i,t S is the lateral difference coefficient of the i-th photovoltaic module at time t. i,t Let S be the seasonal intensity of the i-th photovoltaic module at time t. j,t Let E be the seasonal intensity of the j-th photovoltaic module at time t. i,t Let E be the photovoltaic output vector of the i-th photovoltaic module at time t. j,t Let be the photovoltaic output vector of the j-th photovoltaic module at time t, N be the total number of photovoltaic modules, R() be the similarity calculation function, and exp() be the exponential function with the natural constant as the base.
[0067] It is understood that this embodiment measures the similarity between two vectors using cosine similarity. As other implementation methods, implementers may use other methods of the prior art, such as Pearson correlation coefficient, DTW distance, etc. This embodiment does not impose any special restrictions on this.
[0068] It should be noted that when photovoltaic modules are in normal operating condition, the first similarity is relatively high when the operating state of each photovoltaic module is basically the same as that of the others. The output power of a normal photovoltaic module is consistent with the light intensity. Since the light intensity is periodic, the output power of the photovoltaic module at different times is also periodic. The periodic difference in output power between different photovoltaic modules is small, that is, the first difference is small. The horizontal difference coefficient of the photovoltaic module is small, indicating that the operating state of the photovoltaic module is relatively consistent with that of the others. At this time, the possibility of the photovoltaic module malfunctioning is smaller.
[0069] Furthermore, the flowchart of the method for obtaining the lateral difference coefficient of each photovoltaic module at each time point provided in the embodiments of this application is as follows: Figure 2 As shown.
[0070] Thus, the lateral difference coefficient of each photovoltaic module at each time point is obtained.
[0071] Step 3: Based on the differences in current and output power of each photovoltaic module at each time point compared to the same time point of each historical day, and the impact of meteorological environmental data on output power, determine the longitudinal difference coefficient of each photovoltaic module at each time point.
[0072] Based on the above analysis, the horizontal difference coefficient can only compare the differences between different photovoltaic modules at the same time. However, if there is an anomaly in the entire photovoltaic system, such as the photovoltaic module surface being covered with dust, thus affecting power generation, this global anomaly cannot be detected by horizontal difference alone. Therefore, it is necessary to analyze and compare it with all the historical data of the photovoltaic module to further determine whether there is an anomaly in the photovoltaic module.
[0073] For a photovoltaic module in normal operation, its power generation is related to the intensity of sunlight, which has a diurnal periodicity, thus ensuring that the daily power generation of the photovoltaic module should be basically consistent. However, due to varying environmental conditions, the power generation will fluctuate to some extent, but this fluctuation is correlated with changes in environmental data.
[0074] By calculating the correlation coefficients between light intensity, wind speed, ambient temperature, and humidity and output power at all times, it was found that light intensity and wind speed have a positive correlation with the output power of photovoltaic modules, while ambient temperature and humidity have a negative correlation. Theoretically, higher ambient temperature corresponds to stronger light intensity, and thus higher output power of photovoltaic modules. However, in reality, higher ambient temperature results in lower output power and reduced power generation. Therefore, the correlation coefficient only represents the correlation between ambient temperature and photovoltaic power generation and cannot reflect the influence of multiple couplings between other meteorological factors on the correlation. Based on the correlation coefficients, the impact of meteorological environment on the output power of photovoltaic modules is analyzed, specifically:
[0075] The sum of light intensity and wind speed at each moment is recorded as the first sum; the sum of ambient temperature and humidity at each moment is recorded as the second sum.
[0076] The ratio of the first sum to the second sum is used as the environmental impact coefficient at each time point;
[0077] Furthermore, since output power is related to light intensity, and light intensity has a diurnal periodicity (i.e., it first increases and then decreases each day), a longitudinal difference coefficient for each photovoltaic module at each time point is constructed by comparing and analyzing the differences in output power between each moment and the same historical moment. Specifically:
[0078] The output power sequence and environmental impact sequence for each photovoltaic module at each time point are respectively composed of the output power of each photovoltaic module at all times before each time point and the environmental impact coefficient;
[0079] The second similarity of each photovoltaic module at each time point is determined based on the degree of correlation between the output power sequence and the environmental impact sequence.
[0080] Divide all moments into moments of each day according to their respective dates, and calculate the mean of the similarity between each moment and the same moment of all historical days for each photovoltaic module's photovoltaic output vector;
[0081] The product of the mean and the second similarity is denoted as the first product; the reciprocal of the result of the exponential function with the natural constant as the base and the first product as the exponent is used as the longitudinal difference coefficient of each photovoltaic module at each time point.
[0082] Preferably, in this embodiment, the Pearson correlation coefficient between the output power sequence and the environmental impact sequence is calculated as the second similarity of each photovoltaic module at each time point; the mean of the cosine similarity of the photovoltaic output vector of each photovoltaic module between each time point and the same time point of each day in all its historical data is also calculated.
[0083] In this embodiment, the method for calculating the longitudinal difference coefficient of each photovoltaic module at each time point is as follows: Among them, C i,t R is the longitudinal variation coefficient of the i-th photovoltaic module at time t. i ′ ,t The second similarity of the i-th photovoltaic module at time t is given. Let be the mean of the similarity between the photovoltaic output vector of the i-th photovoltaic module at time t and at the same time of all its historical days.
[0084] It is understood that this embodiment uses Pearson correlation coefficient and cosine similarity to measure the degree of similarity. As other implementation methods, implementers can use other methods of existing technology to measure similarity, such as DTW distance, Euclidean distance, etc. This embodiment does not impose any special restrictions on this.
[0085] It should be noted that when the photovoltaic modules are operating normally, the output power of the photovoltaic modules at the corresponding moment is basically consistent with the changes in output power at the same time on all historical days. The first difference is relatively large; and there is a positive correlation between the output power of photovoltaic modules and the environmental impact coefficient, meaning the second similarity is relatively large, so the longitudinal difference coefficient C is relatively small. Conversely, when there are anomalies in photovoltaic modules, the output power of the photovoltaic modules will be affected, resulting in a larger difference from historical data; and due to the existence of such anomalies, such as shading itself, the power generation of photovoltaic modules will decrease, which will affect the correlation between output power and environmental impact coefficient, thus weakening the positive correlation, resulting in a larger longitudinal difference coefficient.
[0086] Furthermore, the flowchart of the method for obtaining the longitudinal difference coefficient of each photovoltaic module at each time point provided in the embodiments of this application is as follows: Figure 3 As shown.
[0087] Thus, the longitudinal difference coefficient of each photovoltaic module at each time point is obtained.
[0088] Step 4: Based on the horizontal difference coefficient and the vertical difference coefficient, determine the abnormal weight of each photovoltaic module at each time point.
[0089] Based on the above analysis, four different scenarios can occur using the horizontal and vertical difference coefficients: First, if both the horizontal and vertical difference coefficients are large, the photovoltaic module shows a significant difference from other photovoltaic modules and a large difference from historical data. This indicates that the anomaly in the photovoltaic module may have appeared suddenly, such as persistent shading from bird droppings or leaves. Such anomalies require timely handling, otherwise, they will affect the lifespan of the photovoltaic module. Second, if the vertical difference coefficient is large, but the horizontal difference coefficient is small, the photovoltaic module shows a small difference from other photovoltaic modules, but a large difference from historical data. This indicates that the photovoltaic module or the entire photovoltaic system has experienced a global anomaly, such as dust. Such anomalies require timely cleaning, otherwise, they will affect the power generation efficiency of the photovoltaic module and even the entire photovoltaic system. The safety of the photovoltaic module is as follows: Third, if the longitudinal difference coefficient is small, but the lateral difference coefficient is large, then the difference between the photovoltaic module and other photovoltaic modules is large, while the difference with historical data is small. This indicates that the photovoltaic module has a long-term, continuous anomaly. This anomaly causes differences with other photovoltaic modules, such as variable shading anomaly. This anomaly is due to shading, which causes the photovoltaic module's power generation to differ from other modules. However, this type of shading only affects the light intensity of the photovoltaic module to a certain extent, and it is determined by the geographical location of the photovoltaic module installation. It does not damage the photovoltaic module itself and does not require treatment. Fourth, if the longitudinal difference coefficient is small and the lateral difference coefficient is small, then the difference between the photovoltaic module and other photovoltaic modules is small, and the difference with historical data is also small. This indicates that the photovoltaic module is in normal operating condition.
[0090] Based on the above analysis, anomaly weights are constructed to reflect the anomalies of each photovoltaic module, specifically as follows:
[0091] The anomaly weight for each photovoltaic module at each time point is: Among them, W i,t Let A be the outlier weight of the i-th photovoltaic module at time t. i,t Let C be the lateral difference coefficient of the i-th photovoltaic module at time t. i,t α is the longitudinal difference coefficient of the i-th photovoltaic module at time t, w0 is the preset initial weight, which is 1 in this embodiment; α is the preset control coefficient, which is 2 in this embodiment.
[0092] It should be noted that the preset control coefficient should be set to a value greater than 1. The specific value can be determined based on the range of the longitudinal difference coefficient. The range of the longitudinal difference coefficient for photovoltaic modules in normal operation is as follows: but because The maximum value of C is 1, so C i,t Minimum That is, when C i,t The range of values is At this time, the corresponding photovoltaic module is in normal operation. Therefore, when the preset control coefficient α is 2, the range of the vertical difference coefficient of the photovoltaic module is... At this time, for normal photovoltaic modules, when the preset control coefficient α is 3, the range of the longitudinal difference coefficient of the photovoltaic module is as follows: At that time, the photovoltaic modules are considered normal, and the restrictions on abnormal conditions of photovoltaic modules are more stringent; therefore, as an alternative implementation method, the implementer can set it according to the actual situation.
[0093] It should be noted that when photovoltaic modules experience constant shading, both the lateral and longitudinal difference coefficients are relatively large, and A value greater than 0 indicates a higher anomaly weight, and the more severe the anomaly in the corresponding photovoltaic module, the greater its anomaly weight. When a global anomaly exists in a photovoltaic module, the vertical difference coefficient is larger, while the horizontal difference coefficient is smaller, meaning C is larger and A is smaller. A value greater than 0 indicates an anomaly weight greater than w0, meaning its anomaly weight is larger compared to a normal photovoltaic module. When photovoltaic modules experience variable shading, the vertical difference is smaller, while the horizontal difference is larger; that is, A is larger, and C is smaller. A value less than 0 results in an abnormal weight of 1, which is relatively small. When the photovoltaic module is functioning normally, both A and C are small, and If the value is less than 0, the abnormal weight is relatively small, so the value is 1.
[0094] Thus, the anomaly weights of each photovoltaic module at each time point are obtained.
[0095] Step 5: Based on the aforementioned anomaly weights, determine the angle-based anomaly factor for each photovoltaic module at each time point, and perform anomaly monitoring on the photovoltaic modules.
[0096] Furthermore, based on the aforementioned anomaly weights, anomalies in each photovoltaic module are monitored, specifically as follows:
[0097] The current, output power, light intensity, ambient temperature, wind speed and humidity of each photovoltaic module at each time moment are used to form the anomaly monitoring vector of each photovoltaic module at each time moment.
[0098] The abnormal weights of each photovoltaic module at all times are normalized. The abnormal monitoring vectors of each photovoltaic module at each time and all previous times, along with the normalized abnormal weights, are input into the ABOD anomaly detection algorithm to obtain the angle-based anomaly factor of each photovoltaic module at each time.
[0099] Preferably, in this embodiment, the maximum and minimum value normalization method is used to normalize the abnormal weights of each photovoltaic module at all times. The maximum and minimum value normalization method is a well-known technology and will not be described in detail here. As other implementation methods, implementers may use other methods of the prior art, such as the mean normalization method, etc. This embodiment does not impose any special restrictions on this.
[0100] It should be noted that the ABOD anomaly detection algorithm is a well-known technology, and its specific process will not be elaborated further. Calculating the angle basis anomaly factor for each photovoltaic module at each time point using the ABOD anomaly detection algorithm requires certain adjustments. When calculating the angle of each photovoltaic module at each time point, the anomaly weight for that time point needs to be considered. By assigning different anomaly weights to different times, the degree of lateral and longitudinal differences in the scene at different times is taken into account. A higher weight is given to the corresponding time point of photovoltaic modules with more severe anomalies, which helps improve the accuracy of anomaly detection. The following angle calculation formula is used: θ i,t =W i,t ×ω i,t θ i,t Let W be the angle of the i-th photovoltaic module at time t. i,t Let ω be the outlier weight of the i-th photovoltaic module at time t. i,t It is the original angle of the i-th photovoltaic module at the t-th moment, obtained by using the angle calculation method in the ABOD anomaly detection algorithm.
[0101] A threshold segmentation algorithm is applied to the angle-based anomaly factor of each photovoltaic module at all times to obtain the optimal threshold;
[0102] If the angle base anomaly factor of each photovoltaic module at any given time is less than or equal to the optimal threshold, the corresponding photovoltaic module is abnormal and needs to be repaired. If the angle base anomaly factor of each photovoltaic module at any given time is greater than the optimal threshold, the corresponding photovoltaic module is not abnormal and the photovoltaic module is monitored.
[0103] Preferably, in this embodiment, the Otsu threshold segmentation algorithm is used to obtain the optimal threshold. As for other implementation methods, the implementer may use other methods of the prior art, such as histogram construction methods, etc. This embodiment does not impose any special restrictions on this.
[0104] It should be noted that the larger the angle base anomaly factor, the smaller the degree of anomaly of the corresponding photovoltaic module and the greater the possibility that it is in normal operation. Conversely, the smaller the angle base anomaly factor, the greater the possibility that the corresponding photovoltaic module is experiencing anomalies.
[0105] Based on the same inventive concept as the above method, this application embodiment also provides a photovoltaic module intelligent monitoring system, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described photovoltaic module intelligent monitoring methods.
[0106] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.
[0107] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0108] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this application, without departing from the content of the technical solution of this application, shall fall within the protection scope of the technical solution of this application.
Claims
1. A photovoltaic module intelligent monitoring method, characterized in that: The method comprises the following steps: Obtain the current and output power of each photovoltaic module at each moment, as well as the meteorological environment data at each moment, including light intensity, ambient temperature, wind speed and humidity; According to the differences in current, output power and seasonal cycle characteristics of output power between different photovoltaic modules at each moment, the lateral difference coefficient of each photovoltaic module at each moment is determined; According to the difference in current and output power of each photovoltaic module between each moment and the same moment of each day in history, as well as the impact of meteorological environment data on output power, the longitudinal difference coefficient of each photovoltaic module at each moment is determined; Determining the abnormal weight of each photovoltaic module at each moment based on the lateral difference coefficient and the longitudinal difference coefficient; Based on the abnormal weight, the angle-based abnormal factor of each photovoltaic component at each moment is determined, and the photovoltaic components are monitored for abnormalities.
2. A photovoltaic module intelligent monitoring method as claimed in claim 1, characterized in that: The determining of the lateral difference coefficient of each photovoltaic module at each moment includes: Analyze the difference between the seasonal cycle characteristics of each photovoltaic module and the other photovoltaic modules, as well as the correlation between the current and the output power, and determine the first difference between each photovoltaic module and the other photovoltaic modules; Analyze the correlation between the current and the output power of each photovoltaic module at each moment and the other photovoltaic modules at the same moment, and determine the first similarity between each photovoltaic module and the other photovoltaic modules at each moment; Recording the ratio of the first difference between each photovoltaic component and the other photovoltaic components to the first similarity as a first ratio; The first ratio of each photovoltaic module at each moment to all other photovoltaic modules is integrated to determine the lateral difference coefficient of each photovoltaic module at each moment.
3. A photovoltaic module intelligent monitoring method as claimed in claim 2, characterized in that: The determining of a first difference between each photovoltaic component and the remaining photovoltaic components comprises: The seasonal intensity of each PV module is calculated by using a sequence decomposition algorithm for the output power of each PV module at all times; The difference in seasonal intensity between each photovoltaic module and the other photovoltaic modules is taken as the first difference between each photovoltaic module and the other photovoltaic modules.
4. A photovoltaic module intelligent monitoring method as claimed in claim 2, characterized in that: The determining of the first similarity between each photovoltaic component and the other photovoltaic components at each moment includes: The current and output power of each photovoltaic module at each moment are combined to form the photovoltaic output vector of each photovoltaic module at each moment; The similarity of the photovoltaic output vector between each photovoltaic component at each moment and the remaining photovoltaic components at the same moment is calculated, and the calculation result of an exponential function with a natural constant as the base and the similarity as the exponent is used as the first similarity between each photovoltaic component at each moment and the remaining photovoltaic components.
5. A photovoltaic module intelligent monitoring method as claimed in claim 4, characterized in that: The determining of the longitudinal difference coefficient of each photovoltaic module at each moment includes: The ratio between the sum of the light intensity and wind speed at each moment and the sum of the ambient temperature and humidity is taken as the environmental impact coefficient at each moment; Analyze the correlation between the output power of each photovoltaic module at all moments before each moment and the environmental impact coefficient, and determine the second similarity of each photovoltaic module at each moment; Divide all moments into moments of each day according to the date to which they belong, and calculate the average of the similarity between each moment and the same moment of all historical days of the photovoltaic output vector of each photovoltaic module; The product of the mean and the second similarity is recorded as the first product; the reciprocal of the result of the exponential function with the natural constant as the base and the first product as the exponent is taken as the longitudinal difference coefficient of each photovoltaic component at each moment.
6. A photovoltaic module intelligent monitoring method as claimed in claim 5, characterized in that: The determining of the second similarity of each photovoltaic module at each moment includes: The output power of each photovoltaic module at all moments before each moment and the environmental impact coefficient are respectively combined to form an output power sequence and an environmental impact sequence at each moment; The second similarity of each photovoltaic component at each moment is determined according to the correlation degree between the output power sequence and the environmental impact sequence.
7. A photovoltaic module intelligent monitoring method as claimed in claim 1, characterized in that: The abnormal weight of each photovoltaic module at each moment is determined by the expression: Among them, W i,t is the abnormal weight of the ith PV module at the tth moment, A i,t is the lateral difference coefficient of the ith photovoltaic module at the tth moment, C i,t is the longitudinal difference coefficient of the i-th PV module at the t-th moment, w0 is the preset initial weight; α is the preset control coefficient.
8. A photovoltaic module intelligent monitoring method as claimed in claim 1, characterized in that: The step of determining the angle-based anomaly factor of each photovoltaic module at each moment includes: The current, output power, light intensity, ambient temperature, wind speed and humidity of each photovoltaic module at each moment are combined to form an abnormal monitoring vector of each photovoltaic module at each moment; The abnormal weight of each photovoltaic component at all times is normalized, and the abnormal monitoring vector of each photovoltaic component at each time and all previous times and the normalized abnormal weight are input into the ABOD abnormality detection algorithm to obtain the angle-based abnormal factor of each photovoltaic component at each time.
9. A photovoltaic module intelligent monitoring method according to claim 1, characterized in that: The abnormal monitoring of photovoltaic components includes: A threshold segmentation algorithm is used for the angle-based anomaly factor of each photovoltaic module at all times to obtain the optimal threshold; If the angle-based abnormality factor is less than or equal to the optimal threshold, the photovoltaic component is abnormal; otherwise, the photovoltaic component is not abnormal.
10. A photovoltaic assembly intelligent monitoring system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the photovoltaic component intelligent monitoring method as described in any one of claims 1 to 9 are implemented.
Citation Information
Patent Citations
Method and device for judging fault type of photovoltaic module
CN109271736A
Photovoltaic panel fault detection method based on difference characteristic analysis technology
CN112260650A
Method and device for determining running state of photovoltaic power station
CN115833746A
Method and device for determining dust coverage of photovoltaic power station based on neural network
CN116562355A
Photovoltaic new energy intelligent monitoring platform
CN117118353A
