Photovoltaic power station operation data cleaning method and system

Abstract

The present invention relates to the technical field of data processing, and proposes a photovoltaic power station operation data cleaning method and system, including: collecting photovoltaic power station operation data, determining the photovoltaic power station impact data sequence of the photovoltaic power generation sequence in the data time period, determining the environmental impact intensity of the photovoltaic power generation sequence, and screening the affected sequence; according to the linear relationship between the photovoltaic power station operation data reflected by the affected sequence and the photovoltaic power station impact data sequence of the affected sequence, and the discrete degree of the data contained in the affected sequence, the environmental impact degree of the affected sequence is obtained, and according to the environmental impact degrees of all affected sequences, the environmental interference degree of the photovoltaic power station operation data is obtained; and the photovoltaic power station operation data is cleaned according to the environmental interference degree of the photovoltaic power station operation data. The present invention solves the problem that the photovoltaic power station operation data affected by environmental factors is easily misjudged as abnormal data, resulting in insufficient accuracy of the data after data cleaning.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for cleaning photovoltaic power station operation data. Background Art

[0002] Photovoltaic power station operation data is used to record and analyze the performance and operating status of photovoltaic power stations. Photovoltaic power station operation data usually includes power generation data, environmental monitoring data and equipment status data. Analyzing photovoltaic power station operation data can monitor potential equipment failures, identify factors that affect the operating efficiency of photovoltaic power stations, and maintain the smooth operation of photovoltaic power stations. Photovoltaic power station operation data is updated in real time and the amount of data is extremely large. Before analyzing photovoltaic power station operation data, it is necessary to clean the photovoltaic power station operation data, identify and correct errors, duplications or incomplete data in the photovoltaic power station operation data, improve data quality, and ensure the accuracy and reliability of the analysis results of photovoltaic power station operation data.

[0003] The isolation forest algorithm can be used to identify outliers in the operation data of photovoltaic power plants to achieve the purpose of data cleaning. However, environmental changes will affect the operation data of photovoltaic power plants. For example, when the solar radiation intensity, humidity and temperature in the environment change, the photovoltaic power generation and the working efficiency of photovoltaic modules will change accordingly. Therefore, when the isolation forest algorithm is directly used to identify outliers in the operation data of photovoltaic power plants, the impact of environmental factors on the operation data of photovoltaic power plants is often ignored, and the operation data of photovoltaic power plants affected by environmental factors is easily misjudged as abnormal data, making the accuracy and reliability of the photovoltaic power plant operation data after data cleaning insufficient. Summary of the invention

[0004] The present invention provides a photovoltaic power station operation data cleaning method and system to solve the problem that photovoltaic power station operation data affected by environmental factors is easily misjudged as abnormal data, resulting in insufficient accuracy of the data after data cleaning. The technical solution adopted is as follows:

[0005] In a first aspect, an embodiment of the present invention provides a method for cleaning photovoltaic power station operation data, the method comprising the following steps:

[0006] Collecting photovoltaic power station operation data at different collection times within a local time period during the operation of the photovoltaic power station, the photovoltaic power station operation data including photovoltaic power generation and photovoltaic power station impact data, the photovoltaic power station impact data including temperature data, humidity data and light intensity data; dividing the data time period according to the data change trend of photovoltaic power generation, and determining the photovoltaic power station impact data sequence of the photovoltaic power generation sequence of the data time period according to the same type of photovoltaic power station operation data at all collection times within the same data time period;

[0007] According to the correlation between the photovoltaic power generation sequence and the photovoltaic power station impact data sequence, as well as the degree of fluctuation of the data contained in the photovoltaic power station impact data sequence, the environmental impact intensity of the photovoltaic power generation sequence is determined, and the affected sequence is screened out from the photovoltaic power generation sequence according to the environmental impact intensity of all photovoltaic power generation sequences;

[0008] According to the linear relationship between the photovoltaic power station operation data reflected by the affected sequence and the photovoltaic power station impact data sequence of the affected sequence, and the discrete degree of the data contained in the affected sequence, the environmental impact degree of the affected sequence is obtained, and according to the environmental impact degrees of all affected sequences, the environmental interference degree of the photovoltaic power station operation data is obtained;

[0009] The photovoltaic power station operation data is cleaned according to the environmental interference degree of the photovoltaic power station operation data.

[0010] Furthermore, the specific method of dividing the data time periods according to the data change trend of photovoltaic power generation includes:

[0011] The mutation point detection is performed on the photovoltaic power generation to obtain the mutation point in the photovoltaic power generation, and the total time period for collecting the photovoltaic power station operation data is divided into multiple data time periods according to the mutation point.

[0012] Further, the photovoltaic power station impact data sequence of the photovoltaic power generation sequence in the data time period is determined based on the photovoltaic power station operation data of the same type at all collection moments in the same data time period, including the specific method of:

[0013] Arranging the photovoltaic power generation at all collection moments in the same data time period in the order of collection time to determine the photovoltaic power generation sequence of the same data time period;

[0014] The photovoltaic power station operation data of the same type corresponding to the same data time period as the photovoltaic power generation sequence are arranged in the order of collection time to obtain the photovoltaic power station impact data sequence of the photovoltaic power generation sequence.

[0015] Furthermore, the environmental impact intensity of the photovoltaic power generation sequence is obtained by:

[0016] Any photovoltaic power generation sequence is recorded as a target photovoltaic power generation sequence, and any photovoltaic power station impact data sequence of the target photovoltaic power generation sequence is recorded as a target photovoltaic power station impact data sequence;

[0017] The sum of the correlations between the target photovoltaic power generation sequence and all photovoltaic power station influencing data sequences is recorded as the first correlation of the target photovoltaic power generation sequence, and the ratio of the correlation between the target photovoltaic power generation sequence and the target photovoltaic power station influencing data sequence to the first correlation of the target photovoltaic power generation sequence is recorded as the correlation weight between the target photovoltaic power generation sequence and the target photovoltaic power station influencing data sequence;

[0018] The standard deviation of all data contained in the target photovoltaic power station impact data sequence is recorded as the first standard deviation of the target photovoltaic power generation sequence, the mean of the first standard deviations of all photovoltaic power station impact data sequences of the target photovoltaic power generation sequence is recorded as the second standard deviation of the target photovoltaic power generation sequence, and the ratio of the first standard deviation to the second standard deviation of the target photovoltaic power generation sequence is recorded as the weight of the degree of change of the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence;

[0019] The environmental impact intensity of the photovoltaic power generation sequence is obtained according to the correlation weight between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence, as well as the change degree weight between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence.

[0020] Further, the environmental impact intensity of the photovoltaic power generation sequence is obtained according to the correlation weight between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence, and the change degree weight between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence, including the specific method of:

[0021]

[0022] In the formula, Indicates The environmental impact intensity of each photovoltaic power generation sequence; Indicates The PV power generation sequence and The first of the photovoltaic power generation series The correlation weight of the data series affected by each photovoltaic power station; The number of PV power plant impact data series representing the PV power generation series; Indicates The photovoltaic power generation sequence and The first of the photovoltaic power generation series The weight of the degree of change of the data series affected by each photovoltaic power station.

[0023] Furthermore, the method of selecting the affected sequence from the photovoltaic power generation sequence according to the environmental impact intensity of all photovoltaic power generation sequences includes the following specific methods:

[0024] The average value of the environmental impact intensity of all photovoltaic power generation sequences is recorded as the first division threshold, and the photovoltaic power generation sequences with environmental impact intensity greater than the first division threshold are recorded as affected sequences.

[0025] Furthermore, the environmental impact degree of the affected sequence is obtained by:

[0026] Calculate the coefficient of variation of all data contained in the photovoltaic power station impact data sequence of each affected sequence respectively;

[0027] Taking temperature data, humidity data and light intensity data as independent variables and photovoltaic power generation as dependent variable, multiple linear regression is performed on temperature data, humidity data, light intensity data and photovoltaic power generation according to the data contained in the affected sequence and the photovoltaic power station impact data sequence of the affected sequence to obtain the fitting equation and obtain the coefficients of temperature data, humidity data and light intensity data in the fitting equation;

[0028] The product of the absolute value of the coefficient of the data corresponding to the photovoltaic power station impact data sequence of the affected sequence in the fitting equation and the coefficient of variation of all data contained in the photovoltaic power station impact data sequence of the affected sequence is recorded as the impact degree of the photovoltaic power station impact data sequence of the affected sequence, and the cumulative sum of the impact degrees of all photovoltaic power station impact data sequences of the affected sequence is recorded as the environmental impact degree of the affected sequence.

[0029] Furthermore, the environmental interference degree of the photovoltaic power station operation data is obtained by:

[0030] The sum of the environmental impacts of all affected sequences is recorded as the environmental interference degree of the PV power station operation data.

[0031] Furthermore, the data cleaning of the photovoltaic power station operation data according to the environmental interference degree of the photovoltaic power station operation data includes the following specific methods:

[0032] According to the environmental interference degree of the photovoltaic power station operation data, the number of adjusted isolated trees of the photovoltaic power station operation data is obtained. The calculation formula for the number of adjusted isolated trees of the photovoltaic power station operation data is:

[0033]

[0034] In the formula, Indicates the number of adjusted isolated trees for PV power plant operation data; Indicates the preset number of initial isolated trees; Indicates the preset product parameter; Indicates the environmental interference degree of photovoltaic power station operation data; represents the second partition threshold; Represents the rounding function;

[0035] The adjusted number of isolated trees of the photovoltaic power station operation data is used as the value of the number of isolated trees of the isolation forest algorithm, and anomaly detection is performed on the photovoltaic power station operation data to obtain abnormal data, remove the abnormal data, and complete the data cleaning of the photovoltaic power station operation data.

[0036] In a second aspect, an embodiment of the present invention further provides a photovoltaic power station operation data cleaning system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.

[0037] The beneficial effects of the present invention are:

[0038] The present application determines a photovoltaic power generation sequence and a photovoltaic power station impact data sequence of the photovoltaic power generation sequence based on the prominent characteristics of abnormal data that need to be processed during data cleaning, and at the same time, combines the prominent characteristics of data affected by environmental changes compared to unaffected data that changes according to data development trends, evaluates the degree to which the photovoltaic power generation sequence is affected by environmental changes, and screens out the affected sequence from the photovoltaic power generation sequence, where the affected sequence is the photovoltaic power generation sequence that is most affected by environmental changes among all photovoltaic power generation sequences; obtains the environmental impact degree of the affected sequence based on the linear relationship between the photovoltaic power station operation data reflected by the affected sequence and the photovoltaic power station impact data sequence of the affected sequence, and the discrete degree of the data contained in the affected sequence, and then obtains the environmental interference degree of the photovoltaic power station operation data, which is an evaluation of the degree of influence of environmental factors on the photovoltaic power generation contained in the photovoltaic power station operation data; finally, the photovoltaic power station operation data is cleaned based on the environmental interference degree of the photovoltaic power station operation data, so as to solve the problem that the photovoltaic power station operation data affected by environmental factors is easily misjudged as abnormal data, resulting in insufficient accuracy of the data after data cleaning. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0040] Figure 1 A schematic diagram of a process for cleaning photovoltaic power station operation data provided by an embodiment of the present invention;

[0041] Figure 2 A flowchart of obtaining an affected sequence provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] See also Figure 1 , which shows a flow chart of a photovoltaic power station operation data cleaning method provided by an embodiment of the present invention, the method comprising the following steps:

[0044] Step S001, collecting photovoltaic power station operation data at different collection times within a local time period during the operation of the photovoltaic power station, the photovoltaic power station operation data including photovoltaic power generation and photovoltaic power station impact data, the photovoltaic power station impact data including temperature data, humidity data and light intensity data; dividing the data time period according to the data change trend of photovoltaic power generation, and determining the photovoltaic power station impact data sequence of the photovoltaic power generation sequence of the data time period according to the same type of photovoltaic power station operation data at all collection times within the same data time period.

[0045] Ambient temperature, humidity and light intensity can affect the photoelectric conversion efficiency. Ambient temperature generally shows a trend of first increasing and then decreasing within a day. When the ambient temperature increases, the temperature of the photovoltaic module will rise accordingly, resulting in a decrease in conversion efficiency and a decrease in photovoltaic power generation. When the humidity is too high, it is easy for electrical equipment to get damp, causing photovoltaic equipment failure. Light intensity generally shows a trend of first increasing and then decreasing within a day. When the light intensity increases and the sunshine is sufficient, the photovoltaic power generation increases; when the light intensity decreases and the sunshine is insufficient, the photovoltaic power generation decreases. Therefore, this embodiment uses temperature data, humidity data and light intensity data as data to represent changes in environmental factors. On the basis of achieving the purpose of evaluating changes in environmental factors, the implementer can select other data to evaluate changes in environmental factors, and this application does not impose any special restrictions.

[0046] During the operation of the photovoltaic power station, a Logger4000 data collector is deployed in the monitoring system of the photovoltaic power station, and the photovoltaic power generation is collected using the Logger4000 data collector; a temperature sensor, a humidity sensor, and a light intensity sensor are set in the photovoltaic power station, and the temperature data, humidity data, and light intensity data are collected using the temperature sensor, humidity sensor, and light intensity sensor, respectively.

[0047] The temperature data, humidity data and light intensity data collected during the operation of the photovoltaic power station are recorded as the photovoltaic power station impact data. The photovoltaic power generation and photovoltaic power station impact data are recorded as photovoltaic power station operation data.

[0048] It can be understood that the light intensity sensor should be installed in an unobstructed position on the photovoltaic panel, and the temperature sensor and humidity sensor should be installed away from the inverter heat source device; the data collection time of different types of photovoltaic power station operation data should be kept consistent, and the different types of photovoltaic power station operation data should be normalized separately to eliminate the impact of dimension.

[0049] In this embodiment, the sampling interval for the photovoltaic power station impact data and photovoltaic power generation is 30 minutes, and continuous data is collected within one month. In actual application, as other implementation methods, the implementer can decide the sampling interval and collection time of the photovoltaic power station impact data and photovoltaic power generation according to the actual situation, and this application does not impose any special restrictions.

[0050] In a day, the ambient temperature, humidity and light intensity have certain rules. However, the external environment may change suddenly. For example, sudden rainfall and continuous cloudy weather will cause rapid changes in ambient temperature, humidity and light intensity. These rapidly changing photovoltaic power generation and photovoltaic power station impact data are likely to be misjudged as abnormal data. Therefore, when cleaning the photovoltaic power station operation data, special attention should be paid to these prominent data to determine whether they correspond to abnormal data or data affected by environmental changes.

[0051] The photovoltaic power generation is detected at mutation points to obtain mutation points in the photovoltaic power generation. The total time period for collecting photovoltaic power station operation data is divided into multiple data time periods according to the mutation points.

[0052] This embodiment adopts the CUSUM algorithm to detect mutation points in photovoltaic power generation. The expected value of the CUSUM algorithm is set to the average value of all photovoltaic power generation, and the threshold is twice the standard deviation of all photovoltaic power generation. It is a well-known technology to use the CUSUM algorithm to detect mutation points in data and will not be described in detail. In actual application, as other implementation methods, on the basis of achieving the purpose of detecting mutation points in data, the implementer may also use algorithms such as piecewise linear regression (Piece-wise Linear Regression), change point finder (Change Finder), PELT (Pruned Exact Linear Time) change point detection to detect mutation points in the sequence, and this application does not impose any special restrictions.

[0053] The photovoltaic power generation at all collection moments in the same data time period is arranged in the order of collection time to determine the photovoltaic power generation sequence of the same data time period. The photovoltaic power station operation data of the same type in the same data time period corresponding to the photovoltaic power generation sequence is arranged in the order of collection time to obtain the photovoltaic power station impact data sequence of the photovoltaic power generation sequence.

[0054] It can be understood that each photovoltaic power generation sequence corresponds to multiple photovoltaic power station impact data sequences, and the data types corresponding to these photovoltaic power station impact data sequences include temperature data, humidity data and light intensity data, a total of three types, that is, there are three photovoltaic power station impact data sequences for the photovoltaic power generation sequence.

[0055] At this point, the photovoltaic power generation sequence in each data time period and the photovoltaic power station impact data sequence of the photovoltaic power generation sequence are obtained.

[0056] Step S002, determining the environmental impact intensity of the photovoltaic power generation sequence according to the correlation between the photovoltaic power generation sequence and the photovoltaic power station impact data sequence, and the degree of fluctuation of the data contained in the photovoltaic power station impact data sequence, and screening out the affected sequences from the photovoltaic power generation sequence according to the environmental impact intensity of all photovoltaic power generation sequences.

[0057] The environmental impact intensity of the photovoltaic power generation sequence is determined based on the correlation between the photovoltaic power generation sequence and the photovoltaic power station impact data sequence, as well as the degree of fluctuation of the data contained in the photovoltaic power station impact data sequence.

[0058] Preferably, as an embodiment of the present application, any one photovoltaic power generation sequence is recorded as the target photovoltaic power generation sequence, and any one photovoltaic power station impact data sequence of the target photovoltaic power generation sequence is recorded as the target photovoltaic power station impact data sequence. The sum of the correlations between the target photovoltaic power generation sequence and all photovoltaic power station impact data sequences is recorded as the first correlation of the target photovoltaic power generation sequence, and the ratio of the correlation between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence to the first correlation of the target photovoltaic power generation sequence is recorded as the correlation weight between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence. The standard deviation of all data contained in the target photovoltaic power station impact data sequence is recorded as the first standard deviation of the target photovoltaic power generation sequence, the mean of the first standard deviation of all photovoltaic power station impact data sequences of the target photovoltaic power generation sequence is recorded as the second standard deviation of the target photovoltaic power generation sequence, and the ratio of the first standard deviation of the target photovoltaic power generation sequence to the second standard deviation is recorded as the change degree weight of the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence. The product obtained by multiplying the correlation weight of the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence by the change degree weight is recorded as the first product of the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence. The sum of the first products of the target photovoltaic power generation sequence and all photovoltaic power station impact data sequences is recorded as the environmental impact intensity of the target photovoltaic power generation sequence.

[0059] When the standard deviation of all data contained in the target photovoltaic power station impact data sequence is larger than the second standard deviation of the target photovoltaic power generation sequence, the weight of the change degree of the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence is larger. At this time, the more drastic the change degree of the data corresponding to the target photovoltaic power station impact data sequence is, the greater the influence degree of the data type corresponding to the target photovoltaic power station impact data sequence on the target photovoltaic power generation sequence is, and accordingly, the greater the environmental impact intensity of the target photovoltaic power generation sequence is. When the correlation between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence is larger than the first correlation of the target photovoltaic power generation sequence, the more consistent the change trend of the data corresponding to the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence is, at this time, the greater the influence degree of the data type corresponding to the target photovoltaic power station impact data sequence on the target photovoltaic power generation sequence is, and accordingly, the greater the environmental impact intensity of the target photovoltaic power generation sequence is.

[0060] Specifically, the calculation formula for the environmental impact intensity of the photovoltaic power generation sequence is:

[0061]

[0062]

[0063]

[0064] In the formula, Indicates The environmental impact intensity of each photovoltaic power generation sequence; Indicates The photovoltaic power generation sequence and The first of the photovoltaic power generation series The correlation weight of the data series affected by each photovoltaic power station; Indicates The photovoltaic power generation sequence and The first of the photovoltaic power generation series The correlation of the data series affected by each photovoltaic power station; The number of PV power plant impact data series representing the PV power generation series; Indicates The photovoltaic power generation sequence and The first of the photovoltaic power generation series The weight of the degree of change in the data series affected by each photovoltaic power station; Indicates The first of the photovoltaic power generation series Each PV power station affects the standard deviation of all data contained in the data series.

[0065] Preferably, the correlation between the photovoltaic power generation sequence and the photovoltaic power station impact data sequence can be the Pearson correlation coefficient of the photovoltaic power generation sequence and the photovoltaic power station impact data sequence. As other implementation methods, on the basis of achieving the purpose of measuring the correlation of the two sequences, the implementer can adopt other methods in the prior art such as cosine similarity to obtain the correlation between the two sequences, and this application does not impose any special restrictions.

[0066] At this point, the environmental impact intensity of all photovoltaic power generation sequences is obtained.

[0067] The average value of the environmental impact intensity of all photovoltaic power generation sequences is recorded as the first division threshold. The photovoltaic power generation sequence with an environmental impact intensity greater than the first division threshold is recorded as an affected sequence.

[0068] The affected sequence is the photovoltaic power generation sequence that is most affected by environmental changes among all photovoltaic power generation sequences.

[0069] At this point, the affected sequence is determined, and the affected sequence acquisition flow chart is as follows: Figure 2 shown.

[0070] Step S003, according to the linear relationship between the photovoltaic power station operation data reflected by the affected sequence and the photovoltaic power station impact data sequence of the affected sequence, and the discrete degree of the data contained in the affected sequence, the environmental impact degree of the affected sequence is obtained, and according to the environmental impact degrees of all affected sequences, the environmental interference degree of the photovoltaic power station operation data is obtained.

[0071] The isolation forest algorithm is an efficient anomaly detection method. The effect of anomaly detection is affected by the number of isolated trees that need to be manually set in the isolation forest algorithm. When the number of isolated trees selected is too small, the accuracy and stability of the anomaly detection results will be poor; when the number of isolated trees selected is too large, the computational cost will be too high, and at the same time, the overfitting problem is prone to occur. Therefore, this application selects the appropriate value of the number of isolated trees based on the determined affected sequence.

[0072] The environmental impact degree of the affected sequence is obtained based on the linear relationship between the photovoltaic power station operation data reflected by the affected sequence and the photovoltaic power station impact data sequence of the affected sequence, and the discrete degree of the data contained in the affected sequence.

[0073] The coefficient of variation of all data contained in the photovoltaic power station impact data sequence of each affected sequence is calculated separately.

[0074] Among them, the coefficient of variation is a normalized measure of the degree of discreteness of the data, which is a well-known definition in the field of mathematics. Specifically, the coefficient of variation of all data contained in the photovoltaic power station impact data sequence of the affected sequence is the ratio of the mean of all data contained in the photovoltaic power station impact data sequence to the standard deviation. Among them, the calculation of the coefficient of variation, mean and standard deviation is a well-known technology and will not be repeated here.

[0075] Taking temperature data, humidity data and light intensity data as independent variables and photovoltaic power generation as dependent variable, multivariate linear regression is performed on temperature data, humidity data, light intensity data and photovoltaic power generation according to the data contained in the affected sequence and the photovoltaic power station impact data sequence of the affected sequence to obtain the fitting equation. According to the fitting equation, the coefficients of temperature data, humidity data and light intensity data in the fitting equation can be obtained.

[0076] Preferably, as an embodiment of the present application, the product of the absolute value of the coefficient of the data corresponding to the photovoltaic power station impact data sequence of the affected sequence in the fitting equation and the coefficient of variation of all data contained in the photovoltaic power station impact data sequence of the affected sequence is recorded as the influence degree of the photovoltaic power station impact data sequence of the affected sequence, and the cumulative sum of the influence degrees of all photovoltaic power station impact data sequences of the affected sequence is recorded as the environmental impact degree of the affected sequence.

[0077] Specifically, the calculation formula for the environmental impact of the affected sequence is:

[0078]

[0079] In the formula, Indicates The environmental impact of each affected sequence; Indicates The first affected sequence The coefficients of the data corresponding to the data series affected by each photovoltaic power station in the fitting equation; Indicates The first affected sequence The coefficient of variation of all data contained in the data sequence affected by each photovoltaic power station; The number of PV power plant impact data series representing the PV power generation series.

[0080] The environmental impact degree of the affected sequence reflects the degree of influence of environmental factors on the photovoltaic power generation contained in the affected sequence. In this embodiment, temperature data, humidity data and light intensity data are used as data reflecting changes in environmental factors. When the coefficient of variation of all data contained in the photovoltaic power station impact data sequence of the affected sequence is greater, the greater the change in environmental factors, and at the same time, when the absolute value of the coefficient corresponding to the photovoltaic power station impact data sequence of the affected sequence in the fitting equation is greater, the greater the degree of influence of environmental factors on the photovoltaic power generation contained in the affected sequence, at this time, the environmental impact degree of the affected sequence is greater.

[0081] It can be understood that when the coefficient of the data corresponding to the photovoltaic power station impact data sequence of the affected sequence in the fitting equation is greater than 0, the environmental factors have a positive correlation on the photovoltaic power generation contained in the affected sequence, that is, when the data corresponding to the photovoltaic power station impact data sequence increases, the photovoltaic power generation contained in the affected sequence increases accordingly; when the coefficient of the data corresponding to the photovoltaic power station impact data sequence of the affected sequence in the fitting equation is less than 0, the environmental factors have a negative correlation on the photovoltaic power generation contained in the affected sequence, that is, when the data corresponding to the photovoltaic power station impact data sequence decreases, the photovoltaic power generation contained in the affected sequence increases accordingly.

[0082] The sum of the environmental impacts of all affected sequences is recorded as the environmental interference degree of the PV power station operation data.

[0083] The environmental interference degree is an evaluation of the degree of influence of environmental factors on the photovoltaic power generation included in the photovoltaic power station operation data. The greater the influence of environmental factors on the photovoltaic power generation included in the photovoltaic power station operation data, the greater the environmental interference degree of the photovoltaic power station operation data.

[0084] At this point, the environmental interference degree of the photovoltaic power station operation data is obtained.

[0085] Step S004: cleaning the photovoltaic power station operation data according to the environmental interference degree of the photovoltaic power station operation data.

[0086] During the normal operation of the photovoltaic power station, the photovoltaic power station operation data is collected and the environmental interference degree of the photovoltaic power station operation data is obtained. The average value of all environmental interference degrees obtained during the normal operation of the photovoltaic power station is recorded as the second division threshold.

[0087] Preferably, in one embodiment of the present application, during the normal operation of the photovoltaic power station, the photovoltaic power station operation data is collected 100 times, that is, the environmental interference degree during the normal operation of 100 photovoltaic power stations is obtained. In the actual application process, as other implementation methods, the implementer can decide the number of environmental interferences during the normal operation of the photovoltaic power station according to the actual situation, and this application does not impose any special restrictions.

[0088] According to the difference between the environmental interference degree of the photovoltaic power station operation data and the second division threshold, the number of adjusted isolated trees of the photovoltaic power station operation data is obtained.

[0089] Preferably, in one embodiment of the present application, the method for obtaining the number of adjusted isolated trees of photovoltaic power station operation data is:

[0090]

[0091] In the formula, Indicates the number of adjusted isolated trees for PV power plant operation data; represents the preset number of initial isolated trees. In this embodiment, the value of the preset number of initial isolated trees is 100; represents a preset product parameter. In this embodiment, the value of the product parameter is 5; Indicates the environmental interference degree of photovoltaic power station operation data; represents the second partition threshold; Represents the rounding function, which is used to obtain the rounded value of the number in brackets.

[0092] When the difference between the environmental interference degree of the photovoltaic power station operation data and the second division threshold is greater, the greater the adjustment of the preset initial number of isolated trees is required, and the greater the difference between the adjusted number of isolated trees of the photovoltaic power station operation data and the preset initial number of isolated trees is. When the environmental interference degree of the photovoltaic power station operation data is greater than the second division threshold, the greater the impact of environmental factors on the photovoltaic power generation contained in the affected sequence, and a larger value should be selected for the adjusted number of isolated trees of the photovoltaic power station operation data to improve the sensitivity of the isolation forest algorithm for anomaly detection; when the environmental interference degree of the photovoltaic power station operation data is smaller than the second division threshold, the smaller the impact of environmental factors on the photovoltaic power generation contained in the affected sequence, and a smaller value should be selected for the adjusted number of isolated trees of the photovoltaic power station operation data to reduce the amount of calculation for anomaly detection by the isolation forest algorithm and save computing resources.

[0093] At this point, the number of adjusted isolated trees of the photovoltaic power station operation data is obtained.

[0094] The adjusted number of isolated trees of the photovoltaic power station operation data is used as the value of the number of isolated trees of the isolation forest algorithm, and anomaly detection is performed on the photovoltaic power station operation data to obtain abnormal data.

[0095] Preferably, in this embodiment, the number of samples extracted each time, which is a parameter that needs to be manually set in the isolation forest algorithm, is set to 256. Using the isolation forest algorithm to perform anomaly detection on data is a well-known technology and will not be described in detail.

[0096] The abnormal data is regarded as the outlier in the photovoltaic power station operation data, and the abnormal data is eliminated to complete the data cleaning of the photovoltaic power station operation data.

[0097] Based on the same inventive concept as the above method, an embodiment of the present invention also provides a photovoltaic power station operation data cleaning system, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, the steps of any one of the above-mentioned photovoltaic power station operation data cleaning methods are implemented.

[0098] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A photovoltaic power station operation data cleaning method, characterized in that: The method comprises the following steps: Collecting photovoltaic power station operation data at different collection times within a local time period during the operation of the photovoltaic power station, the photovoltaic power station operation data including photovoltaic power generation and photovoltaic power station impact data, the photovoltaic power station impact data including temperature data, humidity data and light intensity data; dividing the data time period according to the data change trend of photovoltaic power generation, and determining the photovoltaic power station impact data sequence of the photovoltaic power generation sequence of the data time period according to the same type of photovoltaic power station operation data at all collection times within the same data time period; According to the correlation between the photovoltaic power generation sequence and the photovoltaic power station impact data sequence, as well as the degree of fluctuation of the data contained in the photovoltaic power station impact data sequence, the environmental impact intensity of the photovoltaic power generation sequence is determined, and the affected sequence is screened out from the photovoltaic power generation sequence according to the environmental impact intensity of all photovoltaic power generation sequences; According to the linear relationship between the photovoltaic power station operation data reflected by the affected sequence and the photovoltaic power station impact data sequence of the affected sequence, and the discrete degree of the data contained in the affected sequence, the environmental impact degree of the affected sequence is obtained, and according to the environmental impact degrees of all affected sequences, the environmental interference degree of the photovoltaic power station operation data is obtained; Data cleaning of photovoltaic power station operation data according to the environmental interference degree of photovoltaic power station operation data; The specific method for obtaining the environmental impact intensity of the photovoltaic power generation sequence is as follows: Any photovoltaic power generation sequence is recorded as a target photovoltaic power generation sequence, and any photovoltaic power station impact data sequence of the target photovoltaic power generation sequence is recorded as a target photovoltaic power station impact data sequence; The sum of the correlations between the target photovoltaic power generation sequence and all photovoltaic power station influencing data sequences is recorded as the first correlation of the target photovoltaic power generation sequence, and the ratio of the correlation between the target photovoltaic power generation sequence and the target photovoltaic power station influencing data sequence to the first correlation of the target photovoltaic power generation sequence is recorded as the correlation weight between the target photovoltaic power generation sequence and the target photovoltaic power station influencing data sequence; The standard deviation of all data contained in the target photovoltaic power station impact data sequence is recorded as the first standard deviation of the target photovoltaic power generation sequence, the mean of the first standard deviations of all photovoltaic power station impact data sequences of the target photovoltaic power generation sequence is recorded as the second standard deviation of the target photovoltaic power generation sequence, and the ratio of the first standard deviation to the second standard deviation of the target photovoltaic power generation sequence is recorded as the weight of the degree of change of the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence; The environmental impact intensity of the photovoltaic power generation sequence is obtained according to the correlation weight between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence, as well as the change degree weight between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence; According to the correlation weight between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence, and the weight of the degree of change between the target photovoltaic power generation sequence and the target photovoltaic power station impact data sequence, the environmental impact intensity of the photovoltaic power generation sequence is obtained, including the specific methods as follows: In the formula, Indicates The environmental impact intensity of each photovoltaic power generation sequence; Indicates The PV power generation sequence and The first of the photovoltaic power generation series The correlation weight of the data series affected by each photovoltaic power station; The number of PV power plant impact data series representing the PV power generation series; Indicates The PV power generation sequence and The first of the photovoltaic power generation series The weight of the degree of change of the data series affected by each photovoltaic power station.

2. A photovoltaic power station operation data cleaning method according to claim 1, characterized in that: The specific method of dividing the data time periods according to the data change trend of photovoltaic power generation includes: The mutation point detection is performed on the photovoltaic power generation to obtain the mutation point in the photovoltaic power generation, and the total time period for collecting the photovoltaic power station operation data is divided into multiple data time periods according to the mutation point.

3. A photovoltaic power station operation data cleaning method according to claim 1, characterized in that: The specific method of determining the photovoltaic power station impact data sequence of the photovoltaic power generation sequence in the data time period based on the same type of photovoltaic power station operation data at all collection moments in the same data time period includes: Arranging the photovoltaic power generation at all collection moments in the same data time period in the order of collection time to determine the photovoltaic power generation sequence of the same data time period; The photovoltaic power station operation data of the same type corresponding to the same data time period as the photovoltaic power generation sequence are arranged in the order of collection time to obtain the photovoltaic power station impact data sequence of the photovoltaic power generation sequence.

4. A photovoltaic power station operation data cleaning method according to claim 1, characterized in that: The method of selecting the affected sequence from the photovoltaic power generation sequence according to the environmental impact intensity of all photovoltaic power generation sequences includes the following specific methods: The average value of the environmental impact intensity of all photovoltaic power generation sequences is recorded as the first division threshold, and the photovoltaic power generation sequences with environmental impact intensity greater than the first division threshold are recorded as affected sequences.

5. A photovoltaic power station operation data cleaning method according to claim 1, characterized in that: The specific method for obtaining the environmental impact degree of the affected sequence is: Calculate the coefficient of variation of all data contained in the photovoltaic power station impact data sequence of each affected sequence respectively; Taking temperature data, humidity data and light intensity data as independent variables and photovoltaic power generation as dependent variable, multiple linear regression is performed on temperature data, humidity data, light intensity data and photovoltaic power generation according to the data contained in the affected sequence and the photovoltaic power station impact data sequence of the affected sequence to obtain the fitting equation and obtain the coefficients of temperature data, humidity data and light intensity data in the fitting equation; The product of the absolute value of the coefficient of the data corresponding to the photovoltaic power station impact data sequence of the affected sequence in the fitting equation and the coefficient of variation of all data contained in the photovoltaic power station impact data sequence of the affected sequence is recorded as the impact degree of the photovoltaic power station impact data sequence of the affected sequence, and the cumulative sum of the impact degrees of all photovoltaic power station impact data sequences of the affected sequence is recorded as the environmental impact degree of the affected sequence.

6. A photovoltaic power station operation data cleaning method according to claim 1, characterized in that: The specific method for obtaining the environmental interference degree of the photovoltaic power station operation data is as follows: The sum of the environmental impacts of all affected sequences is recorded as the environmental interference degree of the PV power station operation data.

7. A photovoltaic power station operation data cleaning method according to claim 1, characterized in that: The data cleaning of the photovoltaic power station operation data according to the environmental interference degree of the photovoltaic power station operation data includes the following specific methods: According to the environmental interference degree of the photovoltaic power station operation data, the number of adjusted isolated trees of the photovoltaic power station operation data is obtained. The calculation formula for the number of adjusted isolated trees of the photovoltaic power station operation data is: In the formula, Indicates the number of adjusted isolated trees for PV power plant operation data; Indicates the preset number of initial isolated trees; Indicates the preset product parameter; Indicates the environmental interference degree of photovoltaic power station operation data; represents the second partition threshold; Represents the rounding function; The adjusted number of isolated trees of the photovoltaic power station operation data is used as the value of the number of isolated trees of the isolation forest algorithm, and anomaly detection is performed on the photovoltaic power station operation data to obtain abnormal data, remove the abnormal data, and complete the data cleaning of the photovoltaic power station operation data.

8. A photovoltaic power station operation data cleaning 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 method according to any one of claims 1 to 7 are implemented.

Images