Photovoltaic power abnormal data detection method and system

Abstract

The application discloses a photovoltaic power abnormal data detection method and system, comprising the following steps: acquiring a discrete sampling sequence of real-time active power; performing Fourier transform on the sampling data based on the discrete sampling sequence of real-time active power, and extracting frequency domain features; performing first data distortion judgment through waveform comparison based on the frequency domain features; performing second data distortion judgment through first numerical value comparison based on the first data distortion judgment; quantifying the abnormal degree of the numerical value by calculating a distortion coefficient based on the sampling point where the data distortion occurs; and finally determining whether the data of the current sampling point is distorted or not through second numerical value comparison based on the distortion coefficient.

Description

Technical Field

[0001] This invention relates to the field of new energy prediction technology, and in particular to a method and system for detecting abnormal photovoltaic power data. Background Technology

[0002] In the field of photovoltaic power generation, power forecasting for photovoltaic power plants is crucial for achieving efficient energy management and optimizing grid operation. Current power forecasting methods involve multiple influencing factors such as wind speed, wind direction, temperature, humidity, air pressure, and solar irradiance, as well as real-time power data. All current forecasting systems require the collection of this information.

[0003] The quality of historical photovoltaic power data has a significant impact on research and application. However, due to human, natural, and random factors, a large number of outliers inevitably exist in historical photovoltaic power data, posing a significant obstacle to data quality and subsequent utilization. Therefore, it is necessary to process the collected historical data, remove unqualified data, and update the power data using interpolation methods. Summary of the Invention

[0004] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] Therefore, the present invention provides a method and system for detecting abnormal photovoltaic power data to solve the problem of abnormal raw data in photovoltaic prediction systems.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] In a first aspect, the present invention provides a method for detecting abnormal photovoltaic power data, comprising:

[0009] Obtain a discrete sampling sequence of real-time active power;

[0010] Based on the discrete sampling sequence of real-time active power, Fourier transform is performed on the sampled data to extract frequency domain features.

[0011] Based on frequency domain characteristics, the first data distortion judgment is made by waveform comparison;

[0012] Based on the first data distortion judgment, a second data distortion judgment is made by comparing the first numerical value;

[0013] Based on the sampling points where data distortion occurs, the degree of abnormality of the values ​​is quantified by calculating the distortion coefficient;

[0014] Based on the distortion coefficient, a second numerical comparison is used to finally determine whether the data at the current sampling point has been distorted.

[0015] As a preferred embodiment of the photovoltaic power anomaly data detection method described in this invention, wherein:

[0016] The discrete sampling sequence for obtaining real-time active power includes sampling the real-time active power at fixed sampling intervals to obtain discrete sampling sequences for each signal.

[0017] As a preferred embodiment of the photovoltaic power anomaly data detection method described in this invention, wherein:

[0018] The initial data distortion assessment through waveform comparison includes the following steps:

[0019] Calculate the similarity between the frequency domain features of the current sampling point and the frequency domain features of the normal waveform to obtain the waveform similarity;

[0020] If the waveform similarity of the current sampling point is higher than the set threshold, then the data of the current sampling point has not been distorted.

[0021] If the waveform similarity of the current sampling point is lower than the set threshold, then the data at the current sampling point will be distorted.

[0022] As a preferred embodiment of the photovoltaic power anomaly data detection method described in this invention, wherein:

[0023] The second data distortion judgment based on the first numerical comparison includes the following steps:

[0024] The active power value at the current sampling point is denoted as A. k The active power values ​​of the previous two sampling points are denoted as A. k-1 and A k-2 ;

[0025] Calculate A k A k-1 and A k-2 The maximum value of the absolute values ​​of the three sampled values, max(|A k |,|A k-1 |,|A k-2 |);

[0026] Compare the maximum absolute value of the three samples with 10% of the average power of the previous day;

[0027] If the maximum absolute value of the three sampled values ​​is less than 10% of the average power of the previous day, then the data at the current sampling point is considered to be free from distortion.

[0028] If the maximum absolute value of the three sampled values ​​is greater than 10% of the average power of the previous day, then the data at the current sampling point is considered to be distorted.

[0029] As a preferred embodiment of the photovoltaic power anomaly data detection method described in this invention, wherein:

[0030] The process of quantifying the degree of abnormality of the numerical values ​​by calculating the distortion coefficients includes calculating three distortion coefficients of the sampled values ​​at different sampling times of the current channel using three formulas. The three distortion coefficients are expressed as follows:

[0031]

[0032]

[0033] Where k1 is the first distortion coefficient, A is calculated. k and A k-2 The average value of A k-1 The difference between them; k2 is the second distortion coefficient, calculate A k and A k-1 The average value of A k-2 The difference between them; k3 is the third distortion coefficient, calculate A k-1 and A k-2 The average value of A k The differences between them.

[0034] As a preferred embodiment of the photovoltaic power anomaly data detection method described in this invention, wherein:

[0035] The process of determining whether the data at the current sampling point has been distorted through a second numerical comparison includes the following steps:

[0036] The calculated three distortion coefficients k1, k2, and k3 are compared with the preset threshold k. set Comparison, k set Take 0.5;

[0037] If any distortion coefficient k i If the value is greater than 0.5, the current sampling point data is considered to be distorted, the unqualified data is removed, and the power data is updated using interpolation.

[0038] If all three distortion coefficients are less than or equal to 0.5, then the data at the current sampling point is considered to be undistorted.

[0039] As a preferred embodiment of the photovoltaic power anomaly data detection method described in this invention, wherein:

[0040] The process of removing unqualified data and updating power data using interpolation includes the following steps:

[0041] When it is determined that data distortion has occurred at a certain sampling point, the data at that sampling point is marked as abnormal data;

[0042] Remove data points marked as anomalies from the dataset;

[0043] The interpolation result is calculated using linear interpolation.

[0044] Replace the data points marked as abnormal with the interpolation results.

[0045] Secondly, the present invention provides an abnormal data detection system, comprising:

[0046] The acquisition module is used to acquire discrete sampling sequences of real-time active power;

[0047] The extraction module is used to perform Fourier transform on the sampled data based on the discrete sampled sequence of real-time active power to extract frequency domain features;

[0048] The comparison module is used to make the first judgment on data distortion based on frequency domain characteristics and waveform comparison;

[0049] The first numerical comparison module is used to perform a second data distortion judgment based on the first data distortion judgment and through the first numerical comparison.

[0050] The calculation module is used to quantify the degree of abnormality of the values ​​by calculating the distortion coefficient based on the sampling points where data distortion has occurred.

[0051] The second numerical comparison module is used to determine whether the data at the current sampling point is distorted based on the distortion coefficient and through second numerical comparison.

[0052] Thirdly, the present invention provides a computing device, comprising:

[0053] Memory, used to store programs;

[0054] A processor is configured to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the photovoltaic power anomaly data detection method.

[0055] Fourthly, the present invention provides a computer-readable storage medium, comprising: when the program is executed by a processor, the step of implementing the photovoltaic power anomaly data detection method.

[0056] The beneficial effects of this invention are as follows: Based on waveform recognition and correlation analysis, this invention can quickly identify and interpolate distorted sampled data. It samples at fixed intervals and judges whether the sampled values ​​are distorted by calculating the distortion coefficient between different sampling points. If the distortion coefficient between the sampled values ​​of different sampling points is too large, it is considered that the sampled data is distorted. The original data in the photovoltaic prediction system is used for monitoring and processing, thereby improving the reliability of the prediction system. Attached Figure Description

[0057] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0058] Figure 1 This is a schematic diagram of the basic process of a photovoltaic power anomaly data detection method provided in one embodiment of the present invention. Detailed Implementation

[0059] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0060] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0061] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0062] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0063] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0064] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integrated connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0065] Example 1

[0066] Reference Figure 1 As an embodiment of the present invention, a method for detecting abnormal photovoltaic power data is provided, comprising:

[0067] S1: Obtain the discrete sampling sequence of real-time active power;

[0068] In the embodiments of this application, the photovoltaic prediction system samples the data to be monitored (mainly real-time active power) at a fixed sampling interval T (T is taken as 15 minutes) to obtain discrete sampling sequences of each signal.

[0069] S2: Based on the discrete sampling sequence of real-time active power, perform Fourier transform on the sampled data to extract frequency domain features;

[0070] In the embodiments of this application, the extracted features may include amplitude, frequency, phase, waveform shape, etc.

[0071] S3: Based on frequency domain characteristics, the first data distortion judgment is made through waveform comparison;

[0072] In this embodiment of the application, the first data distortion judgment is made by waveform comparison, including the following steps:

[0073] The similarity between the frequency domain features of the current sampling point and the frequency domain features of the normal waveform is calculated to obtain the waveform similarity. Euclidean distance, Mahalanobis distance and other methods can be used to measure the similarity of waveforms.

[0074] If the waveform similarity of the current sampling point is higher than the set threshold, then the data of the current sampling point has not been distorted.

[0075] If the waveform similarity of the current sampling point is lower than the set threshold, then the data at the current sampling point will be distorted.

[0076] S4: Based on the first data distortion judgment, a second data distortion judgment is made by comparing the first value;

[0077] In an optional embodiment, the first numerical comparison can be the standard deviation method, the moving average method, or the maximum value comparison method;

[0078] In an optional embodiment, the standard deviation method includes calculating the mean and standard deviation of the data sequence, setting a threshold k, typically 2 or 3, if a certain data point A k Satisfy |A k If -μ∣>k·σ, then the data distortion has occurred at that point.

[0079] In an optional embodiment, the moving average method includes calculating a moving average of data points, setting a threshold, and considering that a data point has data distortion if its deviation from the moving average exceeds the threshold.

[0080] In this application embodiment, the maximum value comparison method includes the following steps:

[0081] The active power value at the current sampling point is denoted as A. k The active power values ​​of the previous two sampling points are denoted as A. k-1 and A k-2 ;

[0082] Calculate A k A k-1 and A k-2 The maximum value of the absolute values ​​of the three sampled values, max(|A k |,|A k-1 |,|A k-2 |);

[0083] Compare the maximum absolute value of the three samples with 10% of the average power of the previous day;

[0084] If the maximum absolute value of the three sampled values ​​is less than 10% of the average power of the previous day, then the data at the current sampling point is considered to be free from distortion.

[0085] If the maximum absolute value of the three sampled values ​​is greater than 10% of the average power of the previous day, then the data at the current sampling point is considered to be distorted.

[0086] It should be noted that this invention not only considers numerical abrupt changes (by calculating the maximum absolute value) but also incorporates waveform features (by calculating distortion coefficients and waveform similarity), thereby improving the accuracy of anomaly detection. Through preliminary screening and further waveform comparison, false alarms and missed alarms can be effectively avoided, especially when data fluctuates significantly, this method is more accurate in identifying anomalies. It is applicable to various types of data, particularly periodic or regular data, and can better capture anomalies.

[0087] S5: Based on the sampling points where data distortion occurs, the degree of numerical abnormality is quantified by calculating the distortion coefficient;

[0088] The degree of abnormality of the numerical value is quantified by calculating the distortion coefficients. This involves calculating three distortion coefficients of the sampled values ​​at different sampling times of the current channel using three formulas. The three distortion coefficients are expressed as follows:

[0089]

[0090] Where k1 is the first distortion coefficient, A is calculated. k and A k-2 The average value of A k-1 The difference between them; k2 is the second distortion coefficient, calculate A k and A k-1 The average value of A k-2 The difference between them; k3 is the third distortion coefficient, calculate A k-1 and A k-2 The average value of A k The differences between them.

[0091] It should be noted that if A k and A k-2 The average value of A k-1 The difference between them is large, indicating that A k-1 It might be an outlier, if A k and A k-1 The average value of A k-2 The difference between them is large, indicating that A k-2 It might be an outlier, if A k-1 and A k-2 The average value of A k The difference between them is large, indicating that A k It might be an outlier.

[0092] S6: Based on the distortion coefficient, the data at the current sampling point is ultimately determined to be distorted by comparing the second numerical value.

[0093] In an optional embodiment, the second numerical comparison can be the Z-Score method, the isolated forest method, or the distortion coefficient comparison method.

[0094] In an optional embodiment, the Z-Score method includes calculating the mean μ and standard deviation σ of the data sequence for each data point A. k Calculate its Z-Score: Set a threshold z threshold If the Z-Score of a data point satisfies |Z k |>z threshold If the data at that point is distorted, then it is considered that data distortion has occurred.

[0095] In an optional embodiment, the isolated forest method includes training a model using the isolated forest algorithm, using the trained model to predict data points, determining whether they are outliers, setting a threshold for contamination to represent the proportion of outliers, typically between 0.01 and 0.1, and considering that a data point is considered to have data distortion if the model predicts that a data point is an outlier.

[0096] In this embodiment of the application, the distortion coefficient comparison method includes the following steps:

[0097] The calculated three distortion coefficients k1, k2, and k3 are compared with the preset threshold k. set Comparison, k set Take 0.5;

[0098] If any distortion coefficient k i If the value is greater than 0.5, i = 1, 2, 3, then the data at the current sampling point is considered to be distorted, the unqualified data is removed, and the power data is updated using interpolation.

[0099] If all three distortion coefficients are less than or equal to 0.5, then the data at the current sampling point is considered to be undistorted.

[0100] It should be noted that the present invention has significant advantages in terms of accuracy and robustness in anomaly detection, and is particularly suitable for application in complex and variable data environments.

[0101] In this embodiment, unqualified data is removed, and power data is updated using interpolation, including the following steps:

[0102] When it is determined that data distortion has occurred at a certain sampling point, the data at that sampling point is marked as abnormal data;

[0103] Remove data points marked as anomalies from the dataset;

[0104] The interpolation result is calculated using linear interpolation.

[0105] Replace the data points marked as abnormal with the interpolation results.

[0106] In this embodiment of the application, the interpolation result calculated by linear interpolation is expressed as follows:

[0107]

[0108] Where T is the sampling interval time, and t is the time from A k-1 To A k The time interval.

[0109] This embodiment also provides an abnormal data detection system, including:

[0110] The acquisition module is used to acquire discrete sampling sequences of real-time active power;

[0111] The extraction module is used to perform Fourier transform on the sampled data based on the discrete sampled sequence of real-time active power to extract frequency domain features;

[0112] The comparison module is used to make the first judgment on data distortion based on frequency domain characteristics and waveform comparison;

[0113] The first numerical comparison module is used to perform a second data distortion judgment based on the first data distortion judgment and through the first numerical comparison.

[0114] The calculation module is used to quantify the degree of abnormality of the values ​​by calculating the distortion coefficient based on the sampling points where data distortion has occurred.

[0115] The second numerical comparison module is used to determine whether the data at the current sampling point is distorted based on the distortion coefficient and through second numerical comparison.

[0116] Furthermore, this also includes:

[0117] Memory, used to store programs;

[0118] A processor is used to load the program to execute the photovoltaic power anomaly data detection method.

[0119] This embodiment also provides a computer-readable storage medium storing a program that, when executed by a processor, implements the photovoltaic power anomaly data detection method.

[0120] The storage medium proposed in this embodiment and the photovoltaic power anomaly data detection method proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0121] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0122] Example 2

[0123] This is an embodiment of the present invention, which provides a photovoltaic power anomaly data detection system, including an acquisition module, an extraction module, a comparison module, a first numerical comparison module, a calculation module, and a second numerical comparison module;

[0124] In this embodiment of the application, the acquisition module includes acquiring a discrete sampling sequence of real-time active power; sampling the real-time active power at a fixed sampling interval T (usually 15 minutes) to acquire the discrete sampling sequence.

[0125] In this embodiment, the extraction module includes a discrete sampling sequence based on real-time active power, performing Fourier transform on the sampled data to extract frequency domain features; extracting features of the sampled data, such as frequency, amplitude, phase, and waveform shape. Frequency domain features are extracted using methods such as Fourier transform (FFT) or wavelet transform (WT).

[0126] In this embodiment of the application, the comparison module includes a first data distortion judgment based on frequency domain features and waveform comparison; comparing the waveform of the current sampling point with the normal waveform, and using methods such as Euclidean distance and Mahalanobis distance to measure the similarity of the waveforms; if the similarity between the waveform of the current sampling point and the normal waveform is lower than a certain threshold, it is considered that data distortion has occurred at that point.

[0127] In this embodiment of the application, the first numerical comparison module is used to perform a second data distortion judgment based on the first data distortion judgment and through the first numerical comparison; calculate the maximum value of the absolute values ​​of three consecutive sampled values, and if the maximum value is lower than the set threshold value (usually 10% of the average power of the previous day), it is considered that no data distortion has occurred at the current point, and obviously normal data points are excluded to reduce the burden of subsequent calculations.

[0128] In this embodiment of the application, the calculation module includes quantifying the degree of abnormality of the numerical value by calculating the distortion coefficient based on the sampling points where data distortion occurs; and calculating three distortion coefficients between three consecutive sampling points, which reflect the degree of abnormality of the numerical changes between different sampling points.

[0129] In this embodiment, the second numerical comparison module includes a method based on distortion coefficients to determine whether the current sampling point data is distorted through a second numerical comparison. If any distortion coefficient exceeds a preset threshold (default 0.5, adjustable according to actual needs), the sampled data is considered distorted, and the system will issue an alarm. Once it is determined that the data of a certain sampling point is distorted, the data of that point should be removed from the dataset to avoid interfering with subsequent analysis or prediction. For the removed abnormal data points, linear interpolation, polynomial interpolation, or other advanced interpolation methods (such as spline interpolation) can be used to estimate the reasonable value of the point, thereby maintaining the continuity and integrity of the data sequence.

[0130] It should be noted that the present invention provides a different technical approach, which can quickly identify anomalies and perform interpolation, facilitating subsequent prediction.

[0131] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for detecting abnormal photovoltaic power data, characterized in that, include: Obtain a discrete sampling sequence of real-time active power; Based on the discrete sampling sequence of real-time active power, Fourier transform is performed on the sampled data to extract frequency domain features. Based on frequency domain characteristics, the first data distortion judgment is made by waveform comparison; Based on the first data distortion judgment, a second data distortion judgment is made by comparing the first numerical value; Based on the sampling points where data distortion occurs, the degree of abnormality of the values ​​is quantified by calculating the distortion coefficient; Based on the distortion coefficient, the second numerical comparison is used to finally determine whether the data at the current sampling point has been distorted; The initial data distortion assessment through waveform comparison includes the following steps: Calculate the similarity between the frequency domain features of the current sampling point and the frequency domain features of the normal waveform to obtain the waveform similarity; If the waveform similarity of the current sampling point is higher than the set threshold, then the data of the current sampling point has not been distorted. If the waveform similarity of the current sampling point is lower than the set threshold, then the data of the current sampling point will be distorted. The second data distortion judgment based on the first numerical comparison includes the following steps: The active power value at the current sampling point is denoted as A. k The active power values ​​of the previous two sampling points are denoted as A. k-1 and A k-2 ; Calculate A k A k-1 and A k-2 The maximum value of the absolute values ​​of the three sampled values, max(|A k |,|A k-1 |,|A k-2 |); Compare the maximum absolute value of the three samples with 10% of the average power of the previous day; If the maximum absolute value of the three sampled values ​​is less than 10% of the average power of the previous day, then the data at the current sampling point is considered to be free from distortion. If the maximum absolute value of the three sampled values ​​is greater than 10% of the average power of the previous day, then the data at the current sampling point is considered to be distorted. The process of quantifying the degree of abnormality of the numerical values ​​by calculating the distortion coefficients includes calculating three distortion coefficients of the sampled values ​​at different sampling times of the current channel using three formulas. The three distortion coefficients are expressed as follows: Where k1 is the first distortion coefficient, A is calculated. k and A k-2 The average value of A k-1 The difference between them; k2 is the second distortion coefficient, calculate A k and A k-1 The average value of A k-2 The difference between them; k3 is the third distortion coefficient, calculate A k-1 and A k-2 The average value of A k The differences between them.

2. The photovoltaic power anomaly data detection method as described in claim 1, characterized in that: The discrete sampling sequence for obtaining real-time active power includes sampling the real-time active power at fixed sampling intervals to obtain discrete sampling sequences for each signal.

3. The photovoltaic power anomaly data detection method as described in claim 2, characterized in that: The process of determining whether the data at the current sampling point has been distorted through a second numerical comparison includes the following steps: The calculated three distortion coefficients k1, k2, and k3 are compared with the preset threshold k. set Comparison, k set Take 0.5; If any distortion coefficient k i If the value is greater than 0.5, the current sampling point data is considered to be distorted, the unqualified data is removed, and the power data is updated using interpolation. If all three distortion coefficients are less than or equal to 0.5, then the data at the current sampling point is considered to be undistorted.

4. The photovoltaic power anomaly data detection method as described in claim 3, characterized in that: The process of removing unqualified data and updating power data using interpolation includes the following steps: When it is determined that data distortion has occurred at a certain sampling point, the data at that sampling point is marked as abnormal data; Remove data points marked as anomalies from the dataset; The interpolation result is calculated using linear interpolation. Replace the data points marked as abnormal with the interpolation results.

5. A system based on the photovoltaic power anomaly data detection method according to claim 1, characterized in that: The acquisition module is used to acquire discrete sampling sequences of real-time active power; The extraction module is used to perform Fourier transform on the sampled data based on the discrete sampled sequence of real-time active power to extract frequency domain features; The comparison module is used to make the first judgment on data distortion based on frequency domain characteristics and waveform comparison; The first numerical comparison module is used to perform a second data distortion judgment based on the first data distortion judgment and through the first numerical comparison. The calculation module is used to quantify the degree of abnormality of the values ​​by calculating the distortion coefficient based on the sampling points where data distortion has occurred. The second numerical comparison module is used to determine whether the data at the current sampling point is distorted based on the distortion coefficient and through second numerical comparison.

6. An electronic device, characterized in that, include: Memory, used to store programs; A processor for loading the program to perform the steps of the photovoltaic power anomaly data detection method as described in any one of claims 1-4.

7. A computer-readable storage medium storing a program, characterized in that, When the program is executed by the processor, it implements the steps of the photovoltaic power anomaly data detection method as described in any one of claims 1-4.

Images