A method for extracting photovoltaic arrays with abnormal operating status in photovoltaic power stations

Abstract

The present invention discloses a method for extracting photovoltaic arrays with abnormal operation status in a photovoltaic power station, and the method comprises the following steps: S1. extracting a reference photovoltaic array; S2. using PR value as a characteristic variable, and using a method based on difference sorting to screen photovoltaic arrays with suspected abnormal operation status; S3. extracting photovoltaic arrays with abnormal operation status by analyzing the distance between power data points of the photovoltaic arrays with suspected abnormal operation status and the reference photovoltaic array. The present invention extracts photovoltaic arrays with abnormal operation status in a photovoltaic power station by using PR value as a characteristic variable, a method based on difference sorting, and analyzing the distance between power data points of the photovoltaic arrays with suspected abnormal operation status and the reference photovoltaic array, thereby compensating for the problem of inaccurate calculation of PR value of photovoltaic arrays caused by the quality problem of operation data of actual photovoltaic power stations, and can effectively extract photovoltaic arrays with abnormal operation status within the analysis period.

Description

Technical Field

[0001] The invention belongs to the technical field of photovoltaic power generation, and in particular relates to a method for extracting a photovoltaic array with abnormal operating status in a photovoltaic power station. Background Art

[0002] As an important component of renewable energy, solar energy has the characteristics of being inexhaustible, available everywhere, without moving parts, noiseless, pollution-free, with a short construction period and low operating costs, making it the most ideal renewable energy source at present. In recent years, the solar photovoltaic power generation industry has developed rapidly under the support of policies and technological progress, and its scale has been greatly improved. The number of solar power stations continues to grow rapidly around the world. According to statistics from the International Renewable Energy Agency (IRENA), my country's cumulative installed capacity of photovoltaic power generation in 2020 was 253.834GW, and the newly installed capacity of photovoltaic power generation has shown explosive growth in the past decade. Faced with the rapid growth in the number of photovoltaic power stations, the importance of photovoltaic power station operation and maintenance issues has become increasingly prominent.

[0003] The operation and maintenance mode of traditional photovoltaic power stations has the problems of little monitoring data, data cannot be reported in time, communication network construction and maintenance are limited, there is no intelligent analysis, and the on-site operation and maintenance personnel are not skilled enough. It can no longer meet the operation and maintenance needs of large-scale photovoltaic power stations. Therefore, it is imperative to carry out intelligent operation and maintenance of photovoltaic power stations with remote monitoring and control, power generation prediction services, and evaluation of power station assets and economic benefits. An important prerequisite for realizing intelligent operation and maintenance of photovoltaic power stations is to accurately and automatically analyze the operating status of photovoltaic modules. An important part of analyzing the operating status of photovoltaic modules is to extract the abnormal arrays of photovoltaic power station operation status. To extract the photovoltaic arrays with abnormal operating status of photovoltaic power stations, we must first determine the method of judging the abnormality of photovoltaic arrays, that is, to analyze the operating status of photovoltaic arrays.

[0004] At present, the analysis of the operating status of photovoltaic arrays is mainly calculated by the system efficiency (Performance Ratio, PR) indicator specified in the IEC 61724 standard. The photovoltaic array system efficiency (PR) is defined as the ratio of the actual output power of the photovoltaic array to the theoretical power generation of the photovoltaic array. The photovoltaic array PR quantitatively describes the impact of comprehensive factors including stain shading, photovoltaic module power attenuation, series-parallel mismatch and other factors on the operating status of the photovoltaic array.

[0005] However, due to data quality issues such as missing operating data of actual PV power stations and large measurement errors, it is inaccurate to analyze the operating status of PV arrays based solely on the PR calculation results of PV arrays. Figure 1As shown in the figure, the calculation results of the monthly PR value of the PV array show abnormally high values ​​(PR value greater than 1) that violate the law, which can easily lead to misjudgment of the operating status of the PV array, thereby affecting the extraction results of the PV array with abnormal operating status of the PV power station. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides a method for extracting photovoltaic arrays with abnormal operating status in photovoltaic power stations. The method uses a PR value as a characteristic variable, a difference sorting method, and an analysis of the distance between the power data points of the suspected photovoltaic array with abnormal operating status and the reference photovoltaic array to extract the photovoltaic array with abnormal operating status in the photovoltaic power station. The method can make up for the problem of inaccurate calculation of the PR value of the photovoltaic array caused by the quality problem of the actual photovoltaic power station operation data, and can effectively extract the photovoltaic array with abnormal operating status within the analysis period.

[0007] To achieve the above object, the technical solution adopted by the present invention is:

[0008] A method for extracting photovoltaic arrays with abnormal operating status in a photovoltaic power station comprises the following steps:

[0009] S1. Extract reference photovoltaic array;

[0010] S2. Taking PR value as the characteristic variable, a method based on difference sorting is used to screen the photovoltaic arrays with suspected abnormal operating status;

[0011] S3. Extract the photovoltaic array with abnormal operating status by analyzing the distance between the power data points of the photovoltaic array suspected to have abnormal operating status and the reference photovoltaic array.

[0012] The step S1 of extracting the reference photovoltaic array is to use the ranking of the monthly PR values ​​of the photovoltaic arrays in each month, and the top 5% photovoltaic arrays are regarded as photovoltaic arrays in good operating status and serve as the reference photovoltaic arrays.

[0013] Furthermore, the extraction of the reference photovoltaic array is specifically performed as follows:

[0014] First, calculate the PR value of n photovoltaic arrays in the photovoltaic power station for m months, that is:

[0015]

[0016] After the PR values ​​of n photovoltaic arrays for m months are calculated, the PR values ​​of the n photovoltaic arrays in the same month are sorted from small to large to obtain the ranking of the PR values ​​of the n photovoltaic arrays for m months, that is, to determine an arrangement p1, p2, ..., p n , so that the relationship K p1 ≤K p2 ≤...≤K pnExists, according to this relationship, we get the PR value ranking matrix:

[0017]

[0018] After obtaining the PR value ranking of each photovoltaic array in each month, the PR value ranking of the same photovoltaic array in each month is summed up, that is:

[0019]

[0020] The top 5% photovoltaic arrays in terms of ranking are taken as the reference photovoltaic arrays, denoted as A stand , and take out the PR value matrix and voltage U of the reference photovoltaic array stand 、Current I stand , Power data P stand .

[0021] Furthermore, in step S2, the PR value is used as a characteristic variable, and a method based on difference sorting is used to screen the photovoltaic arrays with suspected abnormal operating status. The specific operation is as follows:

[0022] The PR values ​​of the reference photovoltaic arrays extracted in step S1 are averaged on a monthly basis, that is, the PR value matrix of k reference photovoltaic arrays is: Taking the average value by column and transposing it, we get:

[0023] The remaining nk photovoltaic arrays of all photovoltaic arrays in the photovoltaic power station except the reference photovoltaic array determined in step S1 are defined as photovoltaic arrays to be inspected. The PR value of the photovoltaic array to be inspected in each month is subtracted from the average PR value of the reference photovoltaic array in the same month to obtain the difference between the PR values ​​of the photovoltaic array to be inspected and the reference photovoltaic array in each month, that is:

[0024]

[0025] The calculated monthly differences between the PR values ​​of the photovoltaic array to be inspected and the reference photovoltaic array are sorted from small to large to obtain the ranking of the PR value differences between each photovoltaic array to be inspected and the reference photovoltaic array, that is, to determine an arrangement p1, p2, ..., p n-k , so that the relationship K p1 ≤K p2 ≤...≤K pn-k Exists, according to this relationship, we get the difference ranking matrix:

[0026]

[0027] The calculated rankings of the PR value differences between each PV array to be inspected and the reference PV array are summed up, namely:

[0028]

[0029] The calculated ranking sum of the PR value differences between each PV array to be inspected and the reference PV array is sorted, that is, an arrangement p1, p2, ..., p1 of 1, 2, ..., nk is determined. n-k , so that the relationship K p1 ≤K p2 ≤…≤K pn -k exists, according to this relationship:

[0030]

[0031] The first 10% of PV arrays are taken as the suspected abnormal PV arrays, denoted as A susp , and take out its voltage U susp 、Current I susp With power P susp data.

[0032] Furthermore, in step S3, the photovoltaic array with abnormal operation status is extracted by analyzing the distance between the power data points of the photovoltaic array suspected to have abnormal operation status and the reference photovoltaic array. The specific operation is as follows:

[0033] The suspected abnormal photovoltaic array screened out in step S2 is defined as the photovoltaic array to be analyzed. At a time interval of 15 minutes, the distance d between the power of each data point of the photovoltaic array to be analyzed and the power data mean of the reference photovoltaic array at the same time point is calculated. The calculation method is: if there are l data points, then the power data mean of k reference photovoltaic arrays is:

[0034]

[0035] According to the distance formula The distance matrix between the power of l data points of the s photovoltaic arrays to be analyzed in one month and the mean power data of the reference photovoltaic array power at the same time point is calculated as follows:

[0036]

[0037] After obtaining the distance matrix, calculate the average value of the power point distance between each photovoltaic array to be analyzed and the reference photovoltaic array in a single month, that is, average the distance matrix by column and transpose it to obtain:

[0038]

[0039] Then calculate the average value of the distance of each photovoltaic array to be analyzed, that is:

[0040] After the calculation is completed, the array whose distance mean is greater than the average value of each photovoltaic array to be analyzed is taken out. The taken out array is the photovoltaic array with abnormal operating status in the current analysis period, which is recorded as A error ,Right now

[0041] The power P of the photovoltaic array with abnormal operation status error 、Current I error , voltage U error Same parameter mean as reference PV array The comparison chart shows that further analysis of the cause of array abnormality is carried out based on the results shown in the image.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] The present invention provides a method for extracting photovoltaic arrays with abnormal operating status in a photovoltaic power station. The method solves the technical problem that the prior art uses the PR calculation results of the photovoltaic array to analyze the operating status of the photovoltaic array, which easily causes misjudgment of the operating status of the photovoltaic array, thereby affecting the extraction results of the photovoltaic arrays with abnormal operating status in the photovoltaic power station. The photovoltaic arrays with abnormal operating status in the photovoltaic power station are extracted by using the PR value as a characteristic variable, a method based on difference sorting, and an analysis of the distance between the power data points of the suspected photovoltaic array with abnormal operating status and the reference photovoltaic array. This makes up for the problem of inaccurate calculation of the PR value of the photovoltaic array caused by the quality problem of the actual photovoltaic power station operating data, and can effectively extract the photovoltaic arrays with abnormal operating status within the analysis period. In addition, the present invention only relies on the actual operating data of the photovoltaic power station for analysis, does not require a theoretical model, is relatively simple to calculate, and has a wider range of applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a comparison chart of the results of the PR calculation based solely on the photovoltaic array as described in the background technology.

[0045] Figure 2 The present invention is a flowchart of a method for extracting photovoltaic arrays with abnormal operating status according to an embodiment of the present invention.

[0046] Figure 3 This is a bar graph comparing the difference between the average power point distances between the photovoltaic array to be analyzed and the reference photovoltaic array according to an embodiment of the present invention.

[0047] Figure 4 This is a bar chart of the extraction results of the photovoltaic array with abnormal operating status in an embodiment of the present invention.

[0048] Figure 5 1 is a comparison diagram of the average output power of the abnormally operating array M53 and the reference photovoltaic array in an embodiment of the present invention.

[0049] Figure 61 is a comparison diagram of the voltage averages of the array M53 with abnormal operating status and the reference photovoltaic array in an embodiment of the present invention.

[0050] Figure 7 1 is a comparison diagram of the current average of the abnormal operating state array M53 and the reference photovoltaic array in the embodiment of the present invention. DETAILED DESCRIPTION

[0051] The specific implementation modes of the present invention will be further described below in conjunction with the accompanying drawings.

[0052] This example is based on the operating data of a photovoltaic power station from July to December 2017. The specific implementation process is as follows:

[0053] Step S1: Extracting reference PV arrays: First, calculate the PR values ​​of 553 PV arrays in the PV power station for 6 months.

[0054] Furthermore, the calculation formula of the PR value in step S1 is as follows:

[0055]

[0056]

[0057]

[0058] Where: E T is the electric energy generated by the array in the time period T, calculated according to the recorded power P of each combiner box. The input value of P is in vector form, with a time resolution of 10 minutes and a unit of kW. The integral operation is simplified to an interpolation operation; P e : The nominal capacity of the power station components installed, update the parameters according to the power station conditions; h T : It is the peak sunshine hours on the array surface during the T time period, calculated based on the instantaneous value of the array slope radiation H. The input value of H is in vector form, with a time resolution of 10 minutes and a unit of kW / m 2 , the integral operation is simplified to the interpolation operation.

[0059] After the PR value calculation is completed, the PR values ​​of the 553 photovoltaic arrays in the same month are sorted from small to large to obtain the PR value ranking of the 553 photovoltaic arrays for 6 months. After obtaining the PR value ranking of each photovoltaic array in each month, the PR value ranking of the same photovoltaic array in each month is summed up, and the top 5% of the sum of the rankings, that is, the top 10 photovoltaic arrays, are taken as reference photovoltaic arrays and their numbers are recorded, denoted as A. stand , take out the PR value matrix of the reference photovoltaic array and the voltage U stand 、Current I stand , Power data P stand .

[0060] The extraction results of the reference array are shown in Table 1.

[0061] Table 1

[0062]

[0063] Step S2: Using the PR value as a characteristic variable, a method based on difference sorting is used to screen photovoltaic arrays with suspected abnormal operating status;

[0064] Furthermore, the step S2 specifically includes the following steps:

[0065] Step S21: taking the average monthly PR value of the reference PV array extracted in step S1;

[0066] Step S22: The remaining 543 photovoltaic arrays of all photovoltaic arrays in the photovoltaic power station except the reference photovoltaic array determined in step S1 are defined as photovoltaic arrays to be inspected, and the PR value of the photovoltaic array to be inspected in each month is subtracted from the average PR value of the reference photovoltaic array in the same month to obtain the difference between the PR value of the photovoltaic array to be inspected and the reference photovoltaic array in each month;

[0067] Step S23: sorting the monthly differences between the PR values ​​of the photovoltaic array to be inspected and the reference photovoltaic array calculated in step S22 from small to large to obtain a ranking matrix of the differences between the photovoltaic array to be inspected and the reference photovoltaic array;

[0068] Step S24: summing the rankings of the PR value differences between each photovoltaic array to be inspected and the reference photovoltaic array calculated in step S23;

[0069] Step S25: Sort the ranking of the sum of the PR value differences between the photovoltaic arrays to be inspected and the reference photovoltaic arrays calculated in step S24, and take the top 10%, that is, the top 55 photovoltaic arrays as the suspected abnormal photovoltaic arrays, denoted as A susp , and take out its voltage U susp 、Current I susp With power P susp data;

[0070] Step S3: extracting the photovoltaic array with abnormal operation status by analyzing the distance between the power data points of the photovoltaic array suspected to have abnormal operation status and the reference photovoltaic array;

[0071] Furthermore, the specific implementation of step S3 is as follows:

[0072] Step S2 selects out the photovoltaic arrays suspected of abnormal operation status from the photovoltaic arrays to be inspected. Since there are photovoltaic arrays with smaller installed capacity among the photovoltaic arrays suspected of abnormal operation status, the result of comparing the photovoltaic arrays with smaller installed capacity with the reference photovoltaic array has no analytical significance. Therefore, firstly, 12 photovoltaic arrays with smaller installed capacity among the photovoltaic arrays suspected of abnormal operation status are deleted, and 43 photovoltaic arrays to be analyzed are obtained after deletion.

[0073] Calculate the distance matrix d between the power data points of the photovoltaic array to be analyzed every 15 minutes in September and the power data mean of the reference photovoltaic array at the same time point. The calculation method is as follows: first calculate the power data mean of 3394 data points of 10 reference photovoltaic arrays; then calculate the distance matrix d between the power of 3394 data points of 43 photovoltaic arrays to be analyzed in September and the power data mean of the reference photovoltaic array at the same time point; after obtaining the distance matrix d, calculate the average value of the distance between each photovoltaic array to be analyzed and the reference photovoltaic array power point in September, and then calculate the average value of the distance mean of each photovoltaic array to be analyzed;

[0074] After the calculation is completed, the array whose distance mean is greater than the average value of each photovoltaic array to be analyzed is taken out and its number is recorded. The taken out array is the photovoltaic array with abnormal operating status in the current analysis period, which is recorded as A. error ,Right now:

[0075]

[0076] The extraction results of the abnormal running status array are as follows Figure 4 As shown in the figure, the M53 photovoltaic array with the largest distance from the reference array power point is selected to make the output power P of the photovoltaic array error 、Current I error , voltage U error Same parameter mean as reference PV array The comparison chart of output power is shown in Figure 2. Figure 5 As shown, the voltage contrast is Figure 6 As shown, the current contrast is Figure 7 shown.

[0077] Depend on Figure 5 It can be seen that the output power of the M53 photovoltaic array is significantly lower than that of the reference photovoltaic array, which shows that the above method of screening out abnormal photovoltaic arrays is effective; Figure 6 It can be seen that the overall current of the M53 photovoltaic array in September is higher than that of the reference array. Figure 7 It can be seen that the voltage of the M53 photovoltaic array is significantly lower than that of the reference photovoltaic array. Figure 5 , 6The results shown in Figure 7 show that the reduction in the output power of the M53 photovoltaic array is caused by the reduction in its voltage. The reason for the voltage reduction of the M53 photovoltaic array may be a short circuit in the array. The specific cause needs further analysis and investigation.

[0078] The above is only a specific implementation of the invention, but the protection scope of the invention is not limited to it. Any changes or substitutions that are not conceived through creative work should be included in the protection scope of the invention. Therefore, the protection scope of the invention should be based on the protection scope defined in the claims.

Claims

1. A method for extracting photovoltaic arrays with abnormal operating status in a photovoltaic power station, characterized in that: The steps include: S1. Extract reference photovoltaic array; S2. Using the PR value as the characteristic variable, a difference-based sorting method is used to screen photovoltaic arrays with suspected abnormal operating conditions; S3. Extracting abnormal operating state photovoltaic arrays by analyzing the distance between the power data points of the suspected abnormal operating state photovoltaic array and the reference photovoltaic array; wherein, The extraction of the reference photovoltaic array in step S1 is based on the ranking of the monthly PR values ​​of the photovoltaic arrays in each month. The top 5% photovoltaic arrays are considered to be in good operating condition and serve as the reference photovoltaic arrays. The specific operations for extracting the reference photovoltaic array in step S1 are as follows: First, calculate the PR value of n photovoltaic arrays in the photovoltaic power station for m months, that is: ； After the PR values ​​of n photovoltaic arrays for m months are calculated, the PR values ​​of n photovoltaic arrays in the same month are sorted from small to large to obtain the ranking of the PR values ​​of n photovoltaic arrays for m months, that is, to determine an arrangement p1, p2, ..., p1 of 1, 2, ..., n. n , so that the relationship K p1 ≤K p2 ≤…≤K pm Exists, according to this relationship, we can get the PR value ranking matrix: ； After obtaining the PR value ranking of each photovoltaic array in each month, the PR value ranking of the same photovoltaic array in each month is summed up, that is: ； The top 5% photovoltaic arrays in terms of the sum of rankings are taken as the reference photovoltaic arrays, denoted as A stand , and take out the PR value matrix of the reference photovoltaic array and the voltage U stand , current I stand , power data P stand ; In step S2, the PR value is used as a characteristic variable, and a method based on difference sorting is used to screen photovoltaic arrays with suspected abnormal operating status. The specific operation is as follows: The PR values ​​of the reference photovoltaic arrays extracted in step S1 are averaged monthly, that is, the PR value matrix of k reference photovoltaic arrays is: , taking the average value by column and transposing it to get: ; The remaining nk photovoltaic arrays of all photovoltaic arrays in the photovoltaic power station, excluding the reference photovoltaic array determined in step S1, are defined as photovoltaic arrays to be inspected. The PR value of the photovoltaic array to be inspected in each month is subtracted from the average PR value of the reference photovoltaic array in the same month to obtain the difference between the PR values ​​of the photovoltaic array to be inspected and the reference photovoltaic array in each month, that is: ； The calculated monthly differences between the PR values ​​of the photovoltaic array to be inspected and the reference photovoltaic array are sorted from small to large to obtain the ranking of the PR value differences between each photovoltaic array to be inspected and the reference photovoltaic array, that is, to determine A permutation p1, p2, ..., p n-k , so that the relationship K p1 ≤K p2 ≤…≤K pn-k Exists, according to this relationship, we get the difference ranking matrix: ； The calculated ranking of the PR value differences between each PV array to be tested and the reference PV array is summed up, that is: ； Sort the sum of the calculated rankings of the PR value differences between each photovoltaic array to be tested and the reference photovoltaic array, that is, determine an arrangement p1, p2, ..., p1 of 1, 2, ..., nk. n-k , so that the relationship K p1 ≤K p2 ≤…≤K pn-k Exists, according to this relationship: ； The top 10% of PV arrays are taken as the suspected abnormal PV arrays, denoted as A susp , and take out its voltage U susp , current I susp With power P susp data; In step S3, the abnormal operating state photovoltaic array is extracted by analyzing the distance between the power data points of the suspected abnormal operating state photovoltaic array and the reference photovoltaic array. The specific operation is as follows: The photovoltaic arrays suspected of abnormal status selected in step S2 are defined as the photovoltaic arrays to be analyzed. At a time interval of 15 minutes, the distance d between the power of each data point of the photovoltaic array to be analyzed and the mean power data of the reference photovoltaic array at the same time point is calculated. The calculation method is: if there are l data points, then the mean power data of k reference photovoltaic arrays is: ； According to the distance formula The distance matrix between the power of l data points of the s photovoltaic arrays to be analyzed in one month and the mean power data of the reference photovoltaic array at the same time point is calculated as follows: ； After obtaining the distance matrix, calculate the average value of the power point distance between each photovoltaic array to be analyzed and the reference photovoltaic array within a single month. That is, calculate the average value of the distance matrix by column and transpose it to obtain: ； Then calculate the average value of the distance of each photovoltaic array to be analyzed, that is: ; After the calculation is completed, the array with the average distance greater than the average value of each photovoltaic array to be analyzed is taken out. The array taken out is the photovoltaic array with abnormal operating status in the current analysis period, which is recorded as A error ,Right now ; The power P of the photovoltaic array with abnormal operation status error , current I error , voltage U error Same parameter mean as reference PV array 、 、 The comparison chart shows that further analysis of the cause of array abnormality is carried out based on the results shown in the image.

Images