Statistical feature-based photovoltaic module anomaly identification method

Abstract

The application provides a photovoltaic module anomaly recognition method based on statistical characteristics, which comprises the following steps: collecting historical data; establishing current state indicators and voltage state indicators; respectively performing statistical modeling on current state indicator distribution and voltage state indicator distribution; respectively performing feature extraction on the current state indicators and the voltage state indicators; respectively calculating current abnormal operating state thresholds and voltage abnormal operating state thresholds according to the current state feature distribution and the voltage state feature distribution of different photovoltaic modules in the same area; and recognizing photovoltaic module anomalies based on the abnormal operating state thresholds. The method considers the unstable output characteristics of the photovoltaic array, establishes the current and voltage state indicators, sets the confidence to obtain the confidence interval, determines the demarcation point, and determines the threshold range of the real-time operating data, and finally determines the real-time state through the fault diagnosis process, so that the photovoltaic module anomaly recognition can be accurately performed.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of photovoltaic power generation, and particularly relates to a photovoltaic module abnormality identification method based on statistical characteristics. BACKGROUND

[0002] As an important form of renewable energy utilization, photovoltaic power generation has developed rapidly in recent years. According to the report of the International Energy Agency, the world added more than 126 GW of photovoltaic installed capacity in 2020, an increase of 21.8% over 2019, which is unprecedented momentum in energy transformation. Photovoltaic power generation systems work outdoors and need to withstand environmental tests for a long time, which poses great challenges to operation and maintenance. The components, as the core power generation equipment in the system, are more sensitive to environmental changes: partial shadowing and open circuit, short circuit and aging caused by external factors such as dust and fallen leaves can cause photovoltaic panels to be damaged, causing efficiency problems of the entire photovoltaic array and even the entire power plant, and long-term hidden dangers may cause fires. Therefore, how to timely and effectively find and eliminate faults is a problem that the photovoltaic industry needs to consider at all times. However, the current fault diagnosis method often stays at the photovoltaic array or branch level, and these methods are not accurate enough, and the damage of individual components may not be reflected in the branch data, which cannot detect small faults affecting individual modules. Even if the fault is detected, it is not clear which components are faulty. Therefore, it is of great significance to study the component-level abnormal state identification method. SUMMARY

[0003] In view of the above problems in the prior art, the photovoltaic module abnormality identification method based on statistical characteristics can realize the abnormality identification of the voltage and current of the photovoltaic module.

[0004] In order to achieve the above-mentioned application purposes, the technical scheme adopted by the present application is:

[0005] The application provides a photovoltaic module abnormality identification method based on statistical characteristics, which comprises the following steps:

[0006] S1, collecting and cleaning historical operation data and historical environment data of the photovoltaic module;

[0007] S2, establishing current state indicators and voltage state indicators based on the current and voltage data of the photovoltaic module, respectively;

[0008] S3, statistically modeling the current state indicator distribution and the voltage state indicator distribution, respectively, to obtain statistical modeling results;

[0009] S4, extracting features from the current state indicators and the voltage state indicators according to the statistical modeling results to obtain current state features and voltage state features;

[0010] S5, calculating the current abnormal operating state threshold and the voltage abnormal operating state threshold according to the current state characteristic distribution and the voltage state characteristic distribution of different photovoltaic modules in the same area respectively;

[0011] S6, identifying the photovoltaic module abnormality based on the abnormal operating state threshold.

[0012] Further, the specific method of step S1 is:

[0013] Collecting the current and voltage in the photovoltaic module historical operating data and the environmental temperature and irradiance in the historical environmental data, deleting the current, voltage and irradiance data corresponding to the night and the daytime with irradiance lower than 100 W / m 2 ; wherein the photovoltaic module historical operating data includes current data and voltage data.

[0014] Further, the specific method of step S2 is:

[0015] Taking the ratio of the current and irradiance of the photovoltaic module in the same sampling point as the current state index; taking the voltage value of the photovoltaic module as the voltage state index.

[0016] Further, the specific method of step S3 is:

[0017] According to the formula:

[0018]

[0019]

[0020] M a =d·m

[0021] Obtaining the statistical modeling result corresponding to the a-th photovoltaic array when the set state index is x ; wherein π is a constant; exp represents the exponential function with natural constant e as the base; h is the bandwidth; M a represents the total number of state indexes of the a-th photovoltaic array in the total d days; m is the total number of state indexes in a day; σ is the standard deviation of the M a state indexes corresponding to the a-th photovoltaic array; when x changes, the statistical modeling result corresponding to the photovoltaic array changes.

[0022] Further, the specific method of step S4 includes the following sub-steps:

[0023] S4-1, setting the confidence level as (1-α); wherein 0<α<1;

[0024] ​S4-2, define the sum of the probabilities of all values less than or equal to x in the statistical modeling result as a cumulative distribution value, obtain multiple cumulative distribution values when x changes, and then obtain a cumulative distribution function;

[0025] S4-3, according to the formula:

[0026] P (x1 < X < x2) = P {G (a1) < X < G (a2)} = 1 - a

[0027] a2-a1=1-a

[0028] obtain the lower limit x1 and the upper limit x2 of the confidence interval; wherein P (x1 < X < x2) is the probability that the random variable X is distributed between x1 < X < x2; G (a1) = x1, G (a1) is the inverse function of the cumulative distribution function of x1; G (a2) = x2, G (a2) is the inverse function of the cumulative distribution function of x2; P {G (a1) < X < G (a2)} is the probability that the random variable X is distributed between x1 < X < x2;

[0029] S4-4, obtain the sampling points covered by the lower limit x1 to the upper limit x2 of the confidence interval;

[0030] S4-5, construct a current state index fitting curve and a voltage state index fitting curve based on the current state characteristics and voltage state characteristics of the covered sampling points, respectively;

[0031] S4-6, take the abscissa of the peak of the current state index fitting curve as the current state position feature; take the ordinate of the peak of the current state index fitting curve as the current state height feature; take the width covered by the upper and lower boundaries of the current state index fitting curve as the current state scale feature;

[0032] take the abscissa of the peak of the voltage state index fitting curve as the voltage state position feature; take the ordinate of the peak of the voltage state index fitting curve as the voltage state height feature; take the width covered by the upper and lower boundaries of the voltage state index fitting curve as the voltage state scale feature.

[0033] Further, the specific method of step S5 includes the following sub-steps:

[0034] S5-1, arrange the current state position features of different photovoltaic modules in the same area from small to large and divide them into 4 parts on average, and obtain the dividing value of the first part of current state position feature data and the second part of current state position feature data obtain the dividing value of the third part of current state position feature data and the fourth part of current state position feature data

[0035] S5-2, according to the formula:

[0036]

[0037]

[0038] Obtaining the upper end demarcation point of the current state height characteristic data peak position and the lower end demarcation point

[0039] S5-3, arranging the current state height characteristics of different photovoltaic components in the same area from small to large and dividing them into four parts, obtaining the demarcation value of the first current state height characteristic data and the second current state height characteristic data obtaining the demarcation value of the third current state height characteristic data and the fourth current state height characteristic data

[0040] S5-4, according to the formula:

[0041]

[0042]

[0043] Obtaining the upper end demarcation point of the current state height characteristic data peak position and the lower end demarcation point

[0044] S5-5, arranging the current state scale characteristics of different photovoltaic components in the same area from small to large and dividing them into four parts, obtaining the demarcation value of the first current state scale characteristic data and the second current state scale characteristic data obtaining the demarcation value of the third current state scale characteristic data and the fourth current state scale characteristic data

[0045] S5-6, according to the formula:

[0046]

[0047]

[0048] Obtaining the upper end demarcation point of the current state scale characteristic data peak position and the lower end demarcation point

[0049] S5-7, obtaining the upper end demarcation point of the voltage state characteristic data peak position and the lower end demarcation point Obtaining the upper end demarcation point of the voltage state height characteristic data peak position and the lower end demarcation point Acquiring the upper end demarcation point of the voltage state scale feature data top peak position and the lower end demarcation point

[0050] S5-8, taking as the current abnormal operating state threshold; taking and as the voltage abnormal operating state threshold.

[0051] Further, the specific method of step S6 includes the following sub-steps:

[0052] S6-1, judging whether the voltage state position feature of the photovoltaic module is less than If yes, it is determined that the voltage of the photovoltaic module is abnormal, and step S6-4 is entered; otherwise, step S6-2 is entered.

[0053] S6-2, judging whether the voltage state height feature of the photovoltaic module is less than If yes, step S6-3 is entered; otherwise, it is determined that the voltage of the photovoltaic module is normal, and step S6-7 is entered.

[0054] S6-3, judging whether the voltage state scale feature of the photovoltaic module is greater than If yes, it is determined that the voltage of the photovoltaic module is abnormal, and step S6-4 is entered; otherwise, it is determined that the voltage of the photovoltaic module is normal, and step S6-7 is entered.

[0055] S6-4, judging whether the current state position feature of the photovoltaic module is less than If yes, it is determined that both the current and the voltage of the photovoltaic module are abnormal; otherwise, step S6-5 is entered.

[0056] S6-5, judging whether the current state height feature of the photovoltaic module is less than If yes, step S6-6 is entered; otherwise, it is determined that the current of the photovoltaic module is normal but the voltage is abnormal.

[0057] S6-6, judging whether the current state scale feature of the photovoltaic module is greater than If yes, it is determined that both the current and the voltage of the photovoltaic module are abnormal; otherwise, it is determined that the current of the photovoltaic module is normal but the voltage is abnormal.

[0058] S6-7, judging whether the current state position feature of the photovoltaic module is less than If yes, it is determined that both the current and the voltage of the photovoltaic module are abnormal; otherwise, step S6-8 is entered.

[0059] S6-8, judging whether the current state height feature of the photovoltaic module is less than If yes, step S6-9 is entered; otherwise, it is determined that the current of the photovoltaic module is normal but the voltage is abnormal.

[0060] S6-9, judging whether the current state scale feature of the photovoltaic module is greater than If yes, it is determined that the photovoltaic module current is abnormal but the voltage is normal; otherwise, it is determined that the photovoltaic module voltage and current are normal.

[0061] The beneficial effects of the present application are:

[0062] 1. The method considers the unstable characteristics of the photovoltaic array output, establishes the state indicators of current and voltage from the perspective of probability model, then establishes the probability density distribution model of the state indicators (statistical modeling), sets the confidence to obtain the confidence interval, determines the dividing point, and determines the threshold range of real-time operation data, finally determines the real-time state through the fault diagnosis process, which can accurately identify the abnormal photovoltaic module.

[0063] 2. The established state indicators consider that the distribution characteristics of current and power in the photovoltaic array are similar to the distribution characteristics of irradiance, and effectively reduce the data change rate.

[0064] 3. The probability density distribution model of the state indicators does not add any assumption to the data distribution, but studies the distribution characteristics from the data samples themselves, which can effectively solve the problem of inaccurate parameter fitting caused by the uncertainty of photovoltaic array output parameters.

[0065] 4. The array fault diagnosis method based on current and voltage threshold can effectively identify different faults, and can adjust the threshold interval by setting the confidence of the model, which can be used for early warning and related behavior execution. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 is the flowchart of the method;

[0067] Figure 2 is the time series distribution of 7-day historical data in the embodiment;

[0068] Figure 3 is the statistical modeling result distribution diagram corresponding to the current of the three components in the embodiment;

[0069] Figure 4 is the statistical modeling result distribution diagram corresponding to the voltage of the three components in the embodiment. DETAILED DESCRIPTION

[0070] The specific embodiments of the present application are described below to facilitate the understanding of the present application for those skilled in the art, but it should be clear that the present application is not limited to the scope of the specific embodiments, and for those skilled in the art, it is obvious that various changes are within the spirit and scope of the present application defined and determined by the appended claims, and all the inventions utilizing the concept of the present application are within the scope of protection.

[0071] As shown in the figure, the photovoltaic module anomaly recognition method based on statistical characteristics includes the following steps: Figure 1

[0072] S1, collect and clean the historical operation data and historical environment data of the photovoltaic module;

[0073] S2, respectively based on the current and voltage data of the photovoltaic module, establish the current state index and the voltage state index;

[0074] S3, respectively for the current state index distribution and the voltage state index distribution, statistical modeling is carried out, and the statistical modeling result is obtained;

[0075] S4, according to the statistical modeling result, respectively for the current state index and the voltage state index, feature extraction is carried out, and the current state feature and the voltage state feature are obtained;

[0076] S5, according to the current state feature distribution and the voltage state feature distribution of different photovoltaic modules in the same area, the current abnormal operation state threshold and the voltage abnormal operation state threshold are calculated respectively;

[0077] S6, based on the abnormal operation state threshold, the photovoltaic module anomaly is recognized.

[0078] The specific method of step S1 is to collect the current and voltage in the historical operation data of the photovoltaic module and the environmental temperature and irradiance in the historical environment data, delete the current, voltage and irradiance data corresponding to the night and the daytime irradiance less than 100; wherein the historical operation data of the photovoltaic module includes current data and voltage data.

[0079] The specific method of step S2 is to take the ratio of current and irradiance of the photovoltaic module in the same sampling point as the current state index; take the voltage value of the photovoltaic module as the voltage state index.

[0080] The specific method of step S3 for statistical modeling of state index is: according to the formula:

[0081]

[0082]

[0083] Ma=d·m ​

[0084] obtaining the statistical modeling result corresponding to the a-th photovoltaic array when the state indicator is x wherein π is a constant; exp represents an exponential function with a natural constant e as the base; h is a bandwidth; Ma represents the total number of state indicators of the a-th photovoltaic array in a total of d days; m is the total number of state indicators in a day; σ is a standard deviation of the Ma state indicators corresponding to the a-th photovoltaic array; and the statistical modeling result corresponding to the photovoltaic array when x changes follows the change.

[0085] The specific method for feature extraction of the state indicator according to the statistical modeling result in step S4 includes the following sub-steps:

[0086] S4-1, setting a confidence level as (1-α); wherein 0<α<1;

[0087] S4-2, defining the sum of the probabilities of all values less than or equal to x in the statistical modeling result as a cumulative distribution value, obtaining a plurality of cumulative distribution values when x changes, and further obtaining a cumulative distribution function;

[0088] S4-3, according to the formula:

[0089] P(x1<X”x2)=P{G(α1)<X”G(α2)}=1-α

[0090] α2-α1=1-α

[0091] obtaining the lower limit x1 and the upper limit x2 of the confidence interval; wherein P(x1<X”x2) is the probability that the random variable X is distributed between x1<X”x2; G(α1)=x1, G(α1) is the inverse function of the cumulative distribution function of x1; G(α2)=x2, G(α2) is the inverse function of the cumulative distribution function of x2; P{G(α1)<X”G(α2)} is the probability that the random variable X is distributed between x1<X”x2;

[0092] S4-4, obtaining the sampling points covered by the lower limit x1 to the upper limit x2 of the confidence interval;

[0093] S4-5, constructing a current state indicator fitting curve and a voltage state indicator fitting curve based on the current state features and the voltage state features of the covered sampling points, respectively;

[0094] S4-6, taking the abscissa of the peak of the current state indicator fitting curve as the current state position feature, taking the ordinate of the peak of the current state indicator fitting curve as the current state height feature, and taking the width covered by the upper and lower boundaries of the current state indicator fitting curve as the current state scale feature;

[0095] The abscissa of the top of the voltage state index fitting curve is taken as the voltage state position feature; the ordinate of the top of the voltage state index fitting curve is taken as the voltage state height feature; and the width covered by the upper and lower bounds of the voltage state index fitting curve is taken as the voltage state scale feature.

[0096] The specific method of step S5 includes the following sub-steps:

[0097] S5-1, arrange the current state position features of different photovoltaic modules in the same area from small to large and divide them into 4 parts averagely, obtain the demarcation value of the first and second current state position feature data obtain the demarcation value of the third and fourth current state position feature data

[0098] S5-2, according to the formula:

[0099]

[0100]

[0101] obtain the upper demarcation point of the top of the current state feature data and the lower demarcation point

[0102] S5-3, arrange the current state height features of different photovoltaic modules in the same area from small to large and divide them into 4 parts averagely, obtain the demarcation value of the first and second current state height feature data obtain the demarcation value of the third and fourth current state height feature data

[0103] S5-4, according to the formula:

[0104]

[0105]

[0106] obtain the upper demarcation point of the top of the current state height feature data and the lower demarcation point

[0107] S5-5, arrange the current state scale features of different photovoltaic modules in the same area from small to large and divide them into 4 parts averagely, obtain the demarcation value of the first and second current state scale feature data obtain the demarcation value of the third and fourth current state scale feature data

[0108] S5-6, according to the formula:

[0109]

[0110]

[0111] Obtaining the upper end demarcation point of the peak position of the current state scale feature data and the lower end demarcation point

[0112] S5-7, obtaining the upper end demarcation point of the peak position of the voltage state feature data by using the same method as steps S5-1 to S5-6 and the lower end demarcation point Obtaining the upper end demarcation point of the peak position of the voltage state height feature data and the lower end demarcation point Obtaining the upper end demarcation point of the peak position of the voltage state scale feature data and the lower end demarcation point

[0113] S5-8, taking as the current abnormal operating state threshold; taking and as the voltage abnormal operating state threshold.

[0114] The specific method of step S6 includes the following sub-steps:

[0115] S6-1, judging whether the voltage state position feature of the photovoltaic module is less than If yes, it is determined that the voltage of the photovoltaic module is abnormal, and step S6-4 is entered; otherwise, step S6-2 is entered.

[0116] S6-2, judging whether the voltage state height feature of the photovoltaic module is less than If yes, step S6-3 is entered; otherwise, it is determined that the voltage of the photovoltaic module is normal, and step S6-7 is entered.

[0117] S6-3, judging whether the voltage state scale feature of the photovoltaic module is greater than If yes, it is determined that the voltage of the photovoltaic module is abnormal, and step S6-4 is entered; otherwise, it is determined that the voltage of the photovoltaic module is normal, and step S6-7 is entered.

[0118] S6-4, judging whether the current state position feature of the photovoltaic module is less than If yes, it is determined that both the current and the voltage of the photovoltaic module are abnormal; otherwise, step S6-5 is entered.

[0119] S6-5, judging whether the current state height feature of the photovoltaic module is less than If yes, go to step S6-6; otherwise, determine that the photovoltaic module current is normal but the voltage is abnormal.

[0120] S6-6, determine whether the current state scale feature of the photovoltaic module is greater than If yes, determine that the photovoltaic module current and voltage are both abnormal; otherwise, determine that the photovoltaic module current is normal but the voltage is abnormal.

[0121] S6-7, determine whether the current state position feature of the photovoltaic module is less than If yes, determine that the photovoltaic module current and voltage are both abnormal; otherwise, go to step S6-8.

[0122] S6-8, determine whether the current state height feature of the photovoltaic module is less than If yes, go to step S6-9; otherwise, determine that the photovoltaic module current is normal but the voltage is abnormal.

[0123] S6-9, determine whether the current state scale feature of the photovoltaic module is greater than If yes, determine that the photovoltaic module current is abnormal but the voltage is normal; otherwise, determine that the photovoltaic module voltage and current are both normal.

[0124] In the specific implementation process, the data used is derived from a certain photovoltaic test power station. Each branch in the array of the power station is composed of 20 modules in series, and the modules of four branches are equipped with sensors. The system collects current and voltage data every 10 minutes. One of the arrays is selected, and the historical data is cleaned. The current, voltage, power and irradiance data corresponding to the night and the day when the irradiance is lower than 100 W / m 2 Figure 2 The time series distribution of 7-day historical data is shown in the figure. As can be seen from the figure, the current and voltage fluctuate with the weather. The fluctuation is weak on the sunny days of the second day and the seventh day, and is violent under other weather conditions, but there is no obvious rule due to the randomness of weather changes. However, there is a significant difference between the night and the day. The curve basically remains in a certain interval during the day, and the relative change rate of the value is small. The current of the module and the current of the string are closely related to the change of the irradiance, and the curve trend presents a nearly consistent phenomenon, and the data relative change is large during the day.

[0125] As shown in Figure 3 and Figure 4 In the statistical modeling result distribution graph, comparison of a plurality of modules can find that the distribution of most modules is concentrated in a certain interval, but there are also some modules that are significantly different from others, mainly in that: ① the data distribution range is different; ② the abscissa data corresponding to the peak is greatly different from other modules.

[0126] ​The current state position characteristics of different photovoltaic modules in the same area are arranged from small to large and divided into 4 parts, and the obtained boundary values are shown in Table 1. The data of the first to the twentieth components in June are taken for testing, and the performance data of the three characteristic parameters of the 20 components are shown in Table 2.

[0127] Table 1: Boundary values

[0128]

[0129] Table 2: Performance data

[0130]

[0131]

[0132] From Table 2 and the identification method, the components with normal performance are Nos. 1, 2, 3, 5, 6, 7, 9, 10, 11, 13, 14, 15, 16, 17, 18 and 19; the components with abnormal current are Nos. 4 and 8; the components with abnormal voltage are Nos. 17 and 20; and the component with abnormal voltage and current is No. 12.

[0133] In summary, the method considers the unstable characteristics of the photovoltaic array, establishes the state indicators of current and voltage from the perspective of probability model, then establishes the probability density distribution model of the state indicators, sets the confidence interval by setting the confidence, determines the threshold range of real-time operation data, and finally determines the real-time state through the fault diagnosis process. The abnormal identification of photovoltaic components can be accurately performed.

[0134] It is apparent to those skilled in the art that the present application is not limited to the details of the foregoing exemplary embodiments, and that the present application can be implemented in other particular forms without departing from the spirit or essential characteristics of the present application. Therefore, the embodiments should be considered in all respects as illustrative and not restrictive, and the scope of the present application is defined by the appended claims rather than the foregoing description, and it is intended to encompass all changes falling within the meaning and range of equivalents of the elements of the claims. Any reference signs in the claims should not be considered as limiting the claims involved.

[0135] In addition, it should be understood that although the present specification is described in terms of embodiments, not every embodiment contains only one independent technical solution, and the description of the specification is only for the sake of clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can be properly combined to form other embodiments that those skilled in the art can understand.

Claims

1. A statistical feature-based photovoltaic module anomaly identification method, characterized by, The method comprises the following steps: S1, collecting and cleaning historical operation data and historical environmental data of photovoltaic modules; S2, establishing current state indicators and voltage state indicators based on current and voltage data of the photovoltaic modules respectively; S3, performing statistical modeling on current state indicator distribution and voltage state indicator distribution respectively to obtain statistical modeling results; S4, extracting features of the current state indicators and the voltage state indicators according to the statistical modeling results to obtain current state features and voltage state features; S5, calculating current abnormal operation state thresholds and voltage abnormal operation state thresholds according to current state feature distribution and voltage state feature distribution of different photovoltaic modules in the same area; S6, identifying photovoltaic module abnormalities based on the abnormal operation state thresholds. The specific method for extracting features of the state indicators according to the statistical modeling results in step S4 comprises the following sub-steps: S4-1, setting a confidence level as (1-α); wherein 0<α<1; S4-2, defining a sum of probabilities of all values less than or equal to x in the statistical modeling results as a cumulative distribution value, obtaining multiple cumulative distribution values when x changes, and then obtaining a cumulative distribution function; S4-3, obtaining lower limit x1 and upper limit x2 of a confidence interval according to the formula: P(x1<X≤x2)=P{G(α1)<X≤G(α2)}=1-α α2-α1＝1-α wherein P(x1<X≤x2) is a probability that a random variable X is distributed between x1 and x2; G(α1)=x1, G(α1) is an inverse function of the cumulative distribution function of x1; G(α2)=x2, G(α2) is an inverse function of the cumulative distribution function of x2; P{G(α1)<X≤G(α2)} is a probability that a random variable X is distributed between x1 and x2; S4-4, obtaining sampling points covered by the lower limit x1 and the upper limit x2 of the confidence interval; S4-5, constructing a current state indicator fitting curve and a voltage state indicator fitting curve based on current state features and voltage state features of the covered sampling points; S4-6, taking a horizontal coordinate of a top peak of the current state indicator fitting curve as a current state position feature; taking a vertical coordinate of the top peak of the current state indicator fitting curve as a current state height feature; taking a width covered by upper and lower limits of the current state indicator fitting curve as a current state scale feature; taking a horizontal coordinate of a top peak of the voltage state indicator fitting curve as a voltage state position feature; taking a vertical coordinate of the top peak of the voltage state indicator fitting curve as a voltage state height feature; and taking a width covered by upper and lower limits of the voltage state indicator fitting curve as a voltage state scale feature. The specific method of step S5 comprises the following sub-steps: S5-2, according to the formula: S5-4, according to the formula: S5-6, according to the formula: S5-1, arrange the current state position features of different photovoltaic modules in the same area from small to large and average them into 4 parts, obtain the demarcation value of the first part of current state position feature data and the second part of current state position feature data obtain the demarcation value of the third part of current state position feature data and the fourth part of current state position feature data The specific method of step S1 is: Acquiring current state feature data top peak position upper end demarcation point and lower end demarcation point F1 I,D ; S5-3, arrange the current state height features of different photovoltaic modules in the same area from small to large and divide them into 4 parts averagely, obtain the demarcation value of the first part of current state height feature data and the second part of current state height feature data obtain the demarcation value of the third part of current state height feature data and the fourth part of current state height feature data wherein the historical operation data of the photovoltaic modules comprises current data and voltage data. Acquiring current state high feature data top peak position upper end demarcation point and lower end demarcation point F1 I,G ; S5-5, arrange the current state scale features of different photovoltaic modules in the same area from small to large and average them into 4 parts, obtain the demarcation value of the first part of current state scale feature data and the second part of current state scale feature data obtain the demarcation value of the third part of current state scale feature data and the fourth part of current state scale feature data The specific method of step S2 is: Obtaining current state scale feature data top peak position upper end demarcation point and lower end demarcation point F1 I,K ; S5-7, obtaining the upper end demarcation point of the top peak position of the voltage state feature data by using the same method as steps S5-1 to S5-6 and the lower end demarcation point F1 U,D , obtaining the upper end demarcation point of the top peak position of the voltage state height feature data and the lower end demarcation point F1 U,G , obtaining the upper end demarcation point of the top peak position of the voltage state size feature data and the lower end demarcation point F1 U,K ; S5-8, F1 I,D , F1 I,G , as a current abnormal operating condition threshold; F1 U,D , F1 U,G and as a voltage abnormal operating condition threshold. 2.The statistical feature based photovoltaic module anomaly identification method of claim 1, wherein, taking a ratio of the current of the photovoltaic module to the irradiance in the same sampling point as the current state indicator; and taking the voltage value of the photovoltaic module as the voltage state indicator. Collecting current and voltage in historical operation data of photovoltaic module, and environmental temperature and irradiance in historical environmental data, deleting current, voltage and irradiance data corresponding to night and irradiance lower than 100 W / m 2 during day; The specific method for performing statistical modeling on the state indicators in step S3 is: 3.The statistical feature based photovoltaic module anomaly identification method of claim 1, wherein, according to the formula: ​ 4.The statistical feature based photovoltaic module anomaly identification method of claim 1, wherein, ​ ​ M a = d · m statistical modeling result corresponding to the a-th photovoltaic array when the state indicator is set to x where π is a constant; exp represents an exponential function with the natural constant e as the base; h is the bandwidth; M a represents the total number of state indicators of the a-th photovoltaic array in a total of d days; m is the total number of state indicators in a day; σ is the standard deviation of the M a state indicators corresponding to the a-th photovoltaic array; when x changes, the statistical modeling result corresponding to the photovoltaic array follows the change. 5.The statistical feature based photovoltaic module anomaly identification method of claim 1, wherein, The specific method of step S6 includes the following sub-steps: S6-1, judging whether the voltage state position feature of the photovoltaic module is less than F1 U,D If yes, it is determined that the voltage of the photovoltaic module is abnormal, and step S6-4 is entered; otherwise, step S6-2 is entered. S6-2, determining whether the voltage state high feature of the photovoltaic module is less than F1 U,G If yes, go to step S6-3; otherwise, determine that the voltage of the photovoltaic module is normal, and go to step S6-7; S6-3, judging whether the voltage state scale feature of the photovoltaic module is greater than If yes, it is determined that the voltage of the photovoltaic module is abnormal, and step S6-4 is entered; otherwise, it is determined that the voltage of the photovoltaic module is normal, and step S6-7 is entered. S6-4, judging whether the current state position feature of the photovoltaic module is less than F1 I,D If yes, it is determined that the photovoltaic module is abnormal in both current and voltage; otherwise, step S6-5 is entered. S6-5, determining whether the current state height feature of the photovoltaic module is less than F1 I,G If yes, go to step S6-6; otherwise, determine that the current of the photovoltaic module is normal but the voltage is abnormal. S6-6, judging whether the current state scale feature of the photovoltaic module is greater than If yes, it is determined that the current and voltage of the photovoltaic module are both abnormal; otherwise, it is determined that the current of the photovoltaic module is normal but the voltage is abnormal; S6-7, judging whether the current state position feature of the photovoltaic module is less than F1 I,D If yes, it is determined that the photovoltaic module is abnormal in both current and voltage; otherwise, step S6-8 is entered. S6-8, determining whether the current state height feature of the photovoltaic module is less than F1 I,G If yes, go to step S6-9; otherwise, determine that the photovoltaic module current is normal but the voltage is abnormal; S6-9, judging whether the current state scale feature of the photovoltaic module is greater than If yes, it is determined that the photovoltaic module has abnormal current but normal voltage; otherwise, it is determined that the photovoltaic module has normal voltage and current.

Images