A method for estimating global output of wide-area distributed photovoltaics

Abstract

The present invention relates to a method for estimating the global output of wide-area distributed photovoltaics, step S1: obtaining the measured output of each distributed photovoltaic site in the same period of time within a wide area, cleaning the data, calculating the total cluster output in the corresponding time period, and obtaining the output of each site and the total cluster output sequence; step S2: calculating the mutual correlation coefficient of the output sequence of each distributed photovoltaic site and the cluster output according to the historical data obtained in step S1; step S3: calculating the autocorrelation coefficient of the output sequence of each distributed photovoltaic site and the total cluster output sequence; step S4: comprehensively considering the size of the mutual correlation coefficient, the difference between the site autocorrelation coefficient and the cluster autocorrelation coefficient, selecting n sites as sample sites, and then calculating the weight coefficient of each sample site; step S5: using the least squares estimation method to calculate the weight coefficient of each sample site; step S6: using the real-time collected data of the sample site as input to estimate the cluster output. The present invention has a small requirement for the amount of historical data of photovoltaic output, high prediction accuracy, and stronger adaptability.

Description

Technical Field

[0001] The present invention relates to the field of distributed photovoltaic prediction, and in particular to a method for estimating global output of wide-area distributed photovoltaics. Background Art

[0002] Distributed photovoltaic power generation has outstanding advantages such as environmental friendliness, low production cost, and simple installation and maintenance, and has developed rapidly in recent years. As the scale of distributed photovoltaic installations continues to expand, its inherent randomness and uncertainty have had a huge impact on the dispatching and operation of the power grid. However, the dispatching agency currently does not have real-time data on distributed photovoltaic power generation, and cannot accurately perceive the actual operating status of the power grid when making dispatching decisions. At the same time, distributed photovoltaics are connected to medium and low voltage power grids, with scattered access points and no dedicated communication network. Comprehensive monitoring of distributed photovoltaics has poor economic efficiency and low information security, making it difficult to establish a photovoltaic monitoring system over a wide area. In this regard, a feasible solution is to use the global estimation method to infer the distributed photovoltaic output of each region based on the key sample data of photovoltaic power generation and meteorological data monitoring, so as to provide data support for the operation of the power grid. Summary of the invention

[0003] In view of this, an object of the present invention is to provide a method for estimating the global output of wide-area distributed photovoltaics, thereby avoiding the problems of poor economy and low safety caused by establishing a wide-area monitoring system.

[0004] To achieve the above object, the present invention adopts the following technical solution:

[0005] A method for estimating global output of wide-area distributed photovoltaic power generation, comprising the following steps:

[0006] Step S1: Obtain the measured output of each distributed photovoltaic site in the wide area at the same time period, clean the data, calculate the total output of the cluster in the corresponding time period, and obtain the output of each site and the total output sequence of the cluster;

[0007] Step S2: Calculate the correlation coefficient between the output sequence of each distributed photovoltaic site and the cluster output according to the historical data obtained in step S1;

[0008] Step S3: Calculate the autocorrelation coefficient of the output sequence of each distributed photovoltaic site and the total output sequence of the cluster;

[0009] Step S4: comprehensively considering the size of the mutual correlation coefficient, the difference between the site autocorrelation coefficient and the cluster autocorrelation coefficient, selecting n sites as sample sites, and then calculating the weight coefficient of each sample site;

[0010] Step S5: Calculate the weight coefficient of each sample station using the least squares estimation method;

[0011] Step S6: Using the real-time collected data of the sample station as input, cluster output estimation is performed.

[0012] Furthermore, the correlation coefficient R between the output sequence of each distributed photovoltaic site and the cluster output FA The calculation is as follows:

[0013]

[0014] Where P At is the total active power output of the photovoltaic cluster in period t; P Ft is the active power output of PV site F belonging to cluster A in period t; t = 1, 2, …, T; R FA is the correlation coefficient between the actual output of PV site F and the actual total output of all plants in cluster A; is the average output of PV site F during period T; is the average output of PV cluster A during period T.

[0015] Furthermore, the calculation of the output sequence P of each distributed photovoltaic site Ft 、Cluster total output sequence P At The first-order autocorrelation coefficient r k , specifically:

[0016]

[0017] Where P1, P2, …, P T is the output time series; is the output sample mean; k is the lag order of the selected autocovariance.

[0018] Furthermore, the step S4 is specifically as follows: according to the calculation results of step S2 and step S3, according to the mutual correlation coefficient R FA Sort all plants from large to small, and select the autocorrelation coefficient r from the first 2n plants Fk The cluster autocorrelation coefficient r Ak The n closest plants and stations are used as sample stations;

[0019] The cluster output estimation formula is as follows:

[0020]

[0021] In the formula, is the estimated output value of cluster A in period t; b i is the weight coefficient of the i-th sample site, P i,t is the actual output value of the i-th model station in period t.

[0022] Furthermore, the step S5 is specifically as follows:

[0023] 1) Select m time series measurement points, arrange the output of n sample stations at different times into a matrix form, recorded as X, and write the total output of the cluster in m time periods into a column vector form, recorded as y, that is:

[0024]

[0025] 2) Calculate formula (4) to obtain the weight coefficient b i The resulting vector b is:

[0026] b=(X T X) -1 X T y (4).

[0027] Furthermore, the step S6 is specifically as follows: taking the active data collected in real time by n sample stations as input, substituting it into formula (3) can obtain the global estimated output of photovoltaic cluster A at the current moment:

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. The present invention does not rely on a wide-area monitoring system, thus avoiding the problems of poor economy and low safety caused by establishing a wide-area monitoring system. At the same time, the method is based on the correlation analysis of photovoltaic cluster output, and realizes accurate estimation of the global output of distributed photovoltaics in a wide area through photovoltaic sample point selection, weight matching and amplification calculation;

[0030] 2. The present invention has small requirements on the amount of historical data of photovoltaic output, high prediction accuracy and stronger adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0032] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0033] Please refer to Figure 1 The present invention provides a method for estimating the global output of wide-area distributed photovoltaic power generation, comprising the following steps:

[0034] Step S1: Obtain the measured output of each distributed photovoltaic site in the wide area at the same time period, clean the data, calculate the total output of the cluster in the corresponding time period, and obtain the output of each site and the total output sequence of the cluster;

[0035] Step S2: Calculate the correlation coefficient between the output sequence of each distributed photovoltaic site and the cluster output according to the historical data obtained in step S1;

[0036] Step S3: Calculate the autocorrelation coefficient of the output sequence of each distributed photovoltaic site and the total output sequence of the cluster;

[0037] Step S4: comprehensively considering the size of the mutual correlation coefficient, the difference between the site autocorrelation coefficient and the cluster autocorrelation coefficient, selecting n sites as sample sites, and then calculating the weight coefficient of each sample site;

[0038] Step S5: Calculate the weight coefficient of each sample station using the least squares estimation method;

[0039] Step S6: Using the real-time collected data of the sample station as input, cluster output estimation is performed.

[0040] In this embodiment, step S1 is specifically as follows:

[0041] First, obtain the measured output P of each distributed photovoltaic site in the same period of time within a wide area Ft , and clean the data to calculate the total cluster output P in the corresponding time period At , get the output of each site and the total output sequence of the cluster;

[0042] Calculate the correlation coefficient R between the output sequence of each distributed photovoltaic site and the cluster output FA :

[0043]

[0044] Where P At is the total active power output of the photovoltaic cluster in period t; P Ft is the active power output of PV site F belonging to cluster A in period t; t = 1, 2, …, T; R FA is the correlation coefficient between the actual output of PV site F and the actual total output of all plants in cluster A; is the average output of PV site F during period T; is the average output of PV cluster A during period T.

[0045] Calculate the output sequence P of each distributed photovoltaic site Ft 、Cluster total output sequence P At The first-order autocorrelation coefficient r k , specifically:

[0046]

[0047] Where P1, P2, …, P T is the output time series; is the output sample mean; k is the lag order of the selected autocovariance.

[0048] According to the calculation results of step S2 and step S3, according to the mutual correlation coefficient RFA Sort all plants from large to small, and select the autocorrelation coefficient r from the first 2n plants Fk The cluster autocorrelation coefficient r Ak The n closest plants and stations are used as sample stations;

[0049] The cluster output estimation formula is as follows:

[0050]

[0051] In the formula, is the estimated output value of cluster A in period t; b i is the weight coefficient of the i-th sample site, P i,t is the actual output value of the i-th model station in period t.

[0052] Weight coefficient b i It can be obtained by the least squares estimation method, the specific process is:

[0053] 1) Select m time series measurement points, arrange the output of n sample stations at different times into a matrix form, recorded as X, and write the total output of the cluster in m time periods into a column vector form, recorded as y, that is:

[0054]

[0055] 2) Calculate formula (4) to obtain the weight coefficient b i The resulting vector b is:

[0056] b=(X T X) -1 X T y (4).

[0057] In this embodiment, step S6 is specifically as follows: taking the active data collected in real time by n sample stations as input, substituting it into formula (3) can obtain the global estimated output of photovoltaic cluster A at the current moment: The above description is only a preferred embodiment of the present invention. All equivalent changes and modifications made according to the scope of the patent application of the present invention should fall within the scope of the present invention.

Claims

1. A method for estimating global output of wide-area distributed photovoltaics, characterized in that: The following steps are involved: Step S1: Obtain the measured output of each distributed photovoltaic site in the wide area at the same time period, clean the data, calculate the total output of the cluster in the corresponding time period, and obtain the output of each site and the total output sequence of the cluster; Step S2: Calculate the correlation coefficient between the output sequence of each distributed photovoltaic site and the cluster output according to the historical data obtained in step S1; Step S3: Calculate the autocorrelation coefficient of the output sequence of each distributed photovoltaic site and the total output sequence of the cluster; Step S4: comprehensively considering the size of the mutual correlation coefficient, the difference between the site autocorrelation coefficient and the cluster autocorrelation coefficient, selecting n sites as sample sites, and then calculating the weight coefficient of each sample site; Step S5: Calculate the weight coefficient of each sample station using the least squares estimation method; Step S6: using the real-time collected data of the sample station as input to estimate the cluster output; The step S4 is specifically as follows: according to the calculation results of step S2 and step S3, according to the mutual correlation coefficient R FA Sort all plants from large to small, and select the autocorrelation coefficient r from the first 2n plants Fk The cluster autocorrelation coefficient r Ak The n closest plants and stations are used as sample stations; The cluster output estimation formula is as follows: In the formula, is the estimated output value of cluster A in period t; b i is the weight coefficient of the i-th sample site, P i,t is the actual output value of the i-th sample station during period t; The step S5 is specifically as follows: 1) Select m time series measurement points, arrange the output of n sample stations at different times into a matrix form, recorded as X, and write the total output of the cluster in m time periods into a column vector form, recorded as y, that is: 2) Calculate formula (4) to obtain the weight coefficient b i The resulting vector b is: b=(X T X) -1 X T y (4) The step S6 is specifically as follows: taking the active data collected in real time by n sample stations as input, substituting it into formula (3) can obtain the global estimated output of photovoltaic cluster A at the current moment:

2. A method for estimating global output of wide-area distributed photovoltaic power according to claim 1, characterized in that: The correlation coefficient R between the output sequence of each distributed photovoltaic site and the cluster output FA The calculation is as follows: Where P At is the total active power output of the photovoltaic cluster in period t; P Ft is the active power output of PV site F belonging to cluster A in period t; t = 1, 2, …, T; R FA is the correlation coefficient between the actual output of PV site F and the actual total output of all plants in cluster A; is the average output of PV site F during period T; is the average output of PV cluster A during period T.

3. A method for estimating global output of wide-area distributed photovoltaic power according to claim 1, characterized in that: The calculation of the output sequence P of each distributed photovoltaic site Ft 、Cluster total output sequence P At The first-order autocorrelation coefficient r k , specifically: Where P1, P2, …, P T is the output time series; is the output sample mean; k is the lag order of the selected autocovariance.

Images