CN114021838A - Photovoltaic power station dust coverage prediction method - Google Patents

Abstract

A method for predicting dust coverage of a photovoltaic power station comprises the steps of carrying out quantitative calculation and analysis on the dust coverage of the power station by utilizing operation data existing in a monitoring system of the photovoltaic power station; the method comprises the steps of analyzing and predicting the dust coverage and estimating the loss electric quantity by determining the functional relation between the power generation loss rate and time t caused by the dust coverage of a power station, and simultaneously providing a dust coverage quantitative index to realize quantitative calculation of the dust coverage; the invention can predict the dust coverage of the power station; the dust that power station fortune dimension personnel can know whole power station in real time covers the condition, and then decides when to carry out photovoltaic panel and wash comparatively rationally.

Description

Photovoltaic power station dust coverage prediction method

Technical Field

The invention relates to the technical field of photovoltaic power generation, in particular to a method for predicting dust coverage of a photovoltaic power station.

Background

In recent years, the installed photovoltaic capacity of China is continuously increased, the development of photovoltaic power generation technology is changed day by day, and new technology and new method in the photovoltaic field are continuously emerged. The northern area is the area with the most photovoltaic power stations in China, the northern area has more sand storms and large dust, and the dust covered on the photovoltaic power station component seriously influences the power generation efficiency of the photovoltaic power station component.

At present, a reference standard is needed for a method for predicting dust coverage of a photovoltaic power station, such as regularly cleaning a string, and the like, and the dust coverage condition of the power station is monitored and the dust shielding loss rate is calculated by comparing current data of regularly cleaning a component with current data of an unwashed string. The method needs manual participation, cannot predict the covering condition of the dust, and mainly has the following defects:

1. the power loss caused by dust covering before and after the cleaning time point can only be calculated, and cannot be predicted. Because the irradiance and the current of the photovoltaic string are not in a linear relation, the difference value of the current values before and after the photovoltaic string is cleaned cannot effectively estimate the power generation loss caused by dust coverage.

2. Only the dust covering condition of the cluster can be monitored and the loss electric quantity can be analyzed, and the square matrix loss cannot be estimated. The influence degrees of dust on each group string are different, the average value measured and calculated by partial group strings is used for estimating the power generation loss of the whole photovoltaic power station, and the error ratio is large.

3. When the dust coverage condition is evaluated and analyzed, quantitative indexes are lacked, and the dust coverage condition of each group of strings and square matrixes cannot be intuitively reflected.

Disclosure of Invention

In order to overcome the defects of the prior art, the invention aims to provide a method for predicting the dust coverage of a photovoltaic power station, which determines the functional relation between the dust coverage and the loss electric quantity of the photovoltaic power station by using the existing operation data in a monitoring system of the photovoltaic power station, analyzes and predicts the dust coverage and estimates the loss electric quantity, and simultaneously provides a quantitative index of the dust coverage to realize quantitative calculation of the dust coverage.

In order to achieve the purpose, the technical scheme of the invention is as follows:

a method for predicting dust coverage of a photovoltaic power station comprises the following steps:

step 1: acquiring real-time current, voltage, generating capacity and irradiance data from a real-time database of an existing operation control system of a photovoltaic power station;

step 2: making a prediction of dust coverage, at a period T_jInner, T_jReading real-time data I of irradiance per minute in a real-time database at least by day_kThen accumulating the real-time data of irradiance per minute to obtain the total daily irradiation I_d，

Total daily exposure I_dConversion to equivalent daily hours of sunshine H_R-dCombining the formula: theoretical daily generated energy Q_L-dInstalled capacity C of power station_IX equivalent hours of sunshine H_R-dX the efficiency of the system, mu,

Q_L-d＝C_I(MW)×H_R-d(H)×μ

installed capacity C of power station_ICapacity C installed by power generation unit_I-zReplacing, calculating the daily generated energy Q of each theory_L-d-zAccumulating the theoretical daily generated energy to calculate the period T_jInternal theoretical generated energy Q_C-p-iAnd then according to the power decline empirical coefficient K of the photovoltaic module_P-s-zCalculating the period T_jGenerating capacity Q of inner unit kilowatt_C-Y-iThe power generation units are configured identically in the same photovoltaic power plant, so that Q is the same in the same photovoltaic power plant_C-Y-iThere is only one value:

Q_C-Y-i＝(1-K_P-s-z)×Q_C-p-i

and step 3: in an evaluation period T_jInner, T_jReading the current, voltage and real-time data per minute in the real-time database at least in units of days, and calculating the actual data per minuteTime power P_iFurther calculate the daily generated energy Q_zd-n：

Here T_iIs and P_iFor each minute, and then calculate the period T_jHistorical power generation Q_C-iIf minute-level Q is already in the real-time database_C-iThen, the following can be directly read for use:

and 4, step 4: in the period T_jInternal and external calculation unit kilowatt generating capacity Q_C-Tj-iInstalled capacity C_I-zRepresents:

Q_C-Tj-i＝Q_C-i/C_I-z(kw)

and 5: for unit kilowatt generating capacity Q_C-Tj-iSorting and determining the maximum value Q_C-max；

Step 6: calculating the unit kilowatt dust coverage power generation loss rate phi_C-Tj-i：

And 7: calculating the power generation loss rate value phi of dust coverage in at least three different periods_C-Tj-i-1、Φ_C-Tj-i-2、Φ_C-Tj-i-3。

And 8: fitting a function relation f (phi) of the power generation loss rate caused by the ith dust cover and the time t by using a least square method_iAnd t), calculating the ith dust coverage power generation loss rate at any time t according to the functional relation in the step 7, and taking the dust coverage power generation loss rate as a dust coverage quantitative evaluation index.

The invention has the advantages that:

carrying out quantitative calculation and analysis on the dust coverage of the power station; the dust coverage of the power station can be predicted by determining the functional relation between the power generation loss rate caused by the dust coverage of the power station and the time t; the dust that power station fortune dimension personnel can know whole power station in real time covers the condition, and then decides when to carry out photovoltaic panel and wash comparatively rationally.

Drawings

FIG. 1 is a flow chart of the method of the present invention.

Detailed Description

The present invention will be described in detail with reference to the accompanying drawings.

Referring to fig. 1, a method for predicting dust coverage of a photovoltaic power station includes the following steps:

step 1: acquiring data from a real-time database of an existing operation control system of a photovoltaic power station, firstly judging whether daily operation data of a photovoltaic string can be acquired or not, wherein the daily operation data of the photovoltaic string comprises real-time current, voltage, generated energy and irradiance data of each string, and if the daily operation data of the photovoltaic string can be acquired, predicting the coverage of string-level dust; otherwise, setting an inverter corresponding to a square matrix, and predicting the coverage of the square matrix dust.

The "string group" is a power generation unit formed by connecting a plurality of photovoltaic panels in series, and taking the photovoltaic power station 1 st of Jing county Longzhou as an example, the rated power of one photovoltaic panel is 260W, 22 photovoltaic panels are connected in series to form one "string group", the installed capacity of one "string group" is approximately 260W × 22 ═ 5720W (5.72KW), and approximately 176 string groups are arranged under one "square matrix".

Step 2: predicting the coverage of group cascade dust in the period T_jInner, T_jReading real-time data I of irradiance per minute in a real-time database at least by day_kThen accumulating the real-time data of irradiance per minute to obtain the total daily irradiation I_d，

Total daily exposure I_dConversion to daily, etcHours of effective sunshine H_R-dCombining the formula: theoretical daily generated energy Q_L-dInstalled capacity C of power station_IX equivalent hours of sunshine H_R-dX the efficiency of the system, mu,

Q_L-d＝C_I(MW)×H_R-d(H)×μ

installed capacity C of power station_ICapacity C of series-packed by groups_I-zReplacing, calculating the theoretical daily generated energy Q of each group string_L-d-zAccumulating the theoretical daily generated energy to calculate the period T of each string_jInternal theoretical generated energy Q_C-p-iAnd then according to the power decline empirical coefficient K of the photovoltaic module_P-s-zCalculating the period T_jEach internal group of series unit kilowatt power generation capacity Q_C-Y-iIn the same photovoltaic power plant, the configuration of the strings is the same, so that in the same photovoltaic power plant, Q_C-Y-iThere is only one value:

Q_C-Y-i＝(1-K_P-s-z)×Q_C-p-i

and step 3: in an evaluation period T_jInner, T_jReading real-time data of string current and string voltage per minute in the real-time database at least in units of days, and calculating real-time power P of each string per minute_iAnd further calculating the daily power generation quantity Q of each group string_zd-n：

Here T_iIs and P_iEvery minute correspondingly, and then calculates the period T of each group string_jHistorical power generation Q_C-iIf minute-level Q is already in the real-time database_C-iAnd the data can be directly read and used.

And 4, step 4: in the period T_jIn the interior, the unit kilowatt generating capacity Q of each group of strings is calculated_C-Tj-iFor assembling the capacity C_I-zRepresents:

Q_C-Tj-i＝Q_C-i/C_I-z(kw)

and 5: for each group of series unit kilowatt generating capacity Q_C-Tj-iSorting and determining the maximum value Q_C-max。

Step 6: calculating the dust covering power generation loss rate phi of each group of serial units of kilowatts_C-Tj-i：

And 7: calculating the power generation loss rate value phi of each group of dust coverage in three different periods_C-Tj-i-1、Φ_C-Tj-i-2、Φ_C-Tj-i-3。

And 8: fitting a function expression f (phi) of the power generation loss rate caused by the dust coverage of the ith group string and the time t by using a least square method_iAnd t), calculating the dust coverage power generation loss rate of the ith group string at any time t according to the functional relation in the step 7, and calculating and predicting the dust coverage of the group string by taking the 'dust coverage power generation loss rate' as a quantitative evaluation index of the dust coverage.

If the prediction of the dust coverage of the square matrix level is carried out, the period T is_jIn the method, the square matrix level related data is used for calculating and predicting the square matrix level dust coverage, and firstly, a power generation area below a photovoltaic power station is generally called as a square matrix. Take Jing Keng county Longzhou photovoltaic power plant 1 phase as an example: the installed capacity of the power station is 30MW, 30 photovoltaic square matrixes are arranged below the power station, and the installed capacity of each square matrix is 1MW (one square matrix comprises two inverters, and each inverter is 500 KW). The data of each square matrix can be directly read from the inverter. If only one inverter exists under one square matrix, the data of the inverter is the data of the square matrix; if two inverters exist under one square matrix, the generated energy and power data of the square matrix are the sum of the data of the two inverters.

The method specifically comprises the following steps:

(1): read period T_jHistory of internal real-time database variety party matrixGenerated energy Q_F-iCalculating the period T_jThe kilowatt generating capacity of each square array unit is as follows: q_F-T-i。

(2): sequencing kilowatt generating capacity of each square array unit and determining maximum value Q_F-max。

(3): calculating the period T_jThe theoretical generated energy of each square matrix in the system is as follows: q_F-P-i。

(4): according to the past accumulated power decline empirical coefficient K of the photovoltaic module_P-s-zCalculating the kilowatt-response generating capacity Q of each square matrix unit_F-Y-iThe square matrix configuration is generally the same in the same photovoltaic power plant, so that in the same photovoltaic power plant, Q_F-Y-iThere is only one value.

Q_F-Y-i＝(1-K_P-s-z)×Q_F-P-i

(5): calculating the dust covering power generation loss rate phi of each group of serial units of kilowatts_F-Tj-i：

(6): calculating the power generation loss rate value phi of each square matrix of dust coverage in three different periods_F-Tj-i-1、Φ_F-Tj-i-2、Φ_F-Tj-i-3。

(7): fitting a function relation f (phi) of the power generation loss rate caused by the dust coverage of the ith square matrix and the time t by using a least square method_iAnd t), calculating the dust coverage power generation loss rate of the ith square matrix at any time t according to the determined functional relation, and calculating and predicting the dust coverage of the square matrix by taking the 'dust coverage power generation loss rate' as a quantitative evaluation index of the dust coverage.

And (4) conclusion:

the strings and the square matrix belong to two-dimensional power generation units common to photovoltaic power plants. Because some photovoltaic power plants put into operation in the previous years are influenced by the current power plant construction standard and the intelligentization degree of photovoltaic power generation equipment, data of group string current and group string voltage cannot be acquired, so that the group string level dust coverage can not be predicted by using the method, and only the square matrix level dust coverage can be predicted; in recent years, the intelligent degree of installed photovoltaic power generation equipment of photovoltaic power plants in operation is gradually improved, and a plurality of power plants can acquire string current and string voltage data, so that the dust coverage of the photovoltaic power plants can be predicted from two dimensions of strings and matrixes by using the method. Because the data acquisition positions of the cluster and the square matrix are different, and the calculation process of the data is different:

the method comprises the steps that current and voltage data of a minute level are taken for string grouping, then power data of the minute level are obtained through calculation, the power data of the minute level are multiplied by corresponding time and then accumulated, and finally actual generated energy data of the string grouping are obtained;

the generated energy data required by the square matrix is directly read from the inverters, and if a plurality of inverters are arranged below the square matrix, the actual generated energy data of the square matrix can be obtained only by accumulating the generated energy data of the plurality of inverters, so that the prediction results of two dimensions are generally different.

Therefore, the power plant operation maintenance personnel comprehensively consider the prediction results of the two dimensions, select the optimal power plant photovoltaic panel dust cleaning time point, reduce the power plant operation maintenance cost and improve the generated energy to the maximum extent.

Claims (1)

1. A method for predicting dust coverage of a photovoltaic power station is characterized by comprising the following steps:

step 1: acquiring real-time current, voltage, generating capacity and irradiance data from a real-time database of an existing operation control system of a photovoltaic power station;

step 2: making a prediction of dust coverage, at a period T_jInner, T_jReading real-time data I of irradiance per minute in a real-time database at least by day_kThen accumulating the real-time data of irradiance per minute to obtain the total daily irradiation I_d，

Total daily exposure I_dConversion to equivalent daily hours of sunshine H_R-dCombining the formula: theoretical daily generated energy Q_L-dInstalled capacity C of power station_IX equivalent hours of sunshine H_R-dX the efficiency of the system, mu,

Q_L-d＝C_I(MW)×H_R-d(H)×μ

installed capacity C of power station_ICapacity C installed by power generation unit_I-zReplacing, calculating the daily generated energy Q of each theory_L-d-zAccumulating the theoretical daily generated energy to calculate the period T_jInternal theoretical generated energy Q_C-p-iAnd then according to the power decline empirical coefficient K of the photovoltaic module_P-s-zCalculating the period T_jGenerating capacity Q of inner unit kilowatt_C-Y-iThe power generation units are configured identically in the same photovoltaic power plant, so that Q is the same in the same photovoltaic power plant_C-Y-iThere is only one value:

Q_C-Y-i＝(1-K_P-s-z)×Q_C-p-i

and step 3: in an evaluation period T_jInner, T_jReading current, voltage and real-time data per minute in a real-time database at least in units of days, and calculating real-time power per minute P_iFurther calculate the daily generated energy Q_zd-n：

Here T_iIs and P_iFor each minute, and then calculate the period T_jHistorical power generation Q_C-iIf minute-level Q is already in the real-time database_C-iThen, the following can be directly read for use:

and 4, step 4: in the period T_jInternal and external calculation unit kilowatt generating capacity Q_C-Tj-iInstalled capacity C_I-zRepresents:

Q_C-Tj-i＝Q_C-i/C_I-z(kw)

and 5: for unit kilowatt generating capacity Q_C-Tj-iSorting and determining the maximum value Q_C-max；

Step 6: calculating the unit kilowatt dust coverage power generation loss rate phi_C-Tj-i：

And 7: calculating the power generation loss rate value phi of dust coverage in at least three different periods_C-Tj-i-1、Φ_C-Tj-i-2、Φ_C-Tj-i-3；

And 8: fitting a function relation f (phi) of the power generation loss rate caused by the ith dust cover and the time t by using a least square method_iAnd t), calculating the ith dust coverage power generation loss rate at any time t according to the functional relation in the step 7, and taking the dust coverage power generation loss rate as a dust coverage quantitative evaluation index.

Images