Distributed photovoltaic power generation power intelligent prediction method and system

Abstract

The invention relates to the technical field of distributed photovoltaic power prediction, in particular to a distributed photovoltaic power generation power intelligent prediction method and system, and the method comprises the steps: mapping a to-be-predicted area into a two-dimensional plane coordinate system; obtaining two-dimensional plane coordinate points corresponding to the installation geographic positions of the first distributed photovoltaic to the Nth distributed photovoltaic, and respectively recording the two-dimensional plane coordinate points as first to Nth positions; fitting to obtain a wind vector distribution function and a current irradiance distribution function; acquiring actual active power of the first distributed photovoltaic to the Nth distributed photovoltaic; and performing spatio-temporal deduction on the spatial distribution of the actual active power, and calculating to obtain predicted output active power of the first distributed photovoltaic to the Nth distributed photovoltaic at the to-be-predicted time point. According to the method, multi-dimensional data such as wind speed and wind direction and factors such as photovoltaic autonomous voltage regulation characteristics are comprehensively considered, so that distributed photovoltaic output power prediction in high-altitude and light cumulative cloud areas with rich sunlight is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of distributed photovoltaic power prediction, in particular to a distributed photovoltaic power intelligent prediction method and system. BACKGROUND

[0002] In the field of distributed photovoltaic power generation, accurate prediction of distributed photovoltaic power is crucial for grid dispatching, energy management and stable operation of the system. Existing technologies usually use time series analysis, machine learning or statistical methods for power prediction based on historical power generation data, weather forecast information or measured data from a single site.

[0003] However, for distributed photovoltaics deployed in high-altitude areas with abundant sunshine, there are two technical problems that need to be solved in existing prediction methods: first, in high-altitude areas with abundant sunshine, stable stratocumulus clouds often appear, and their spatial distribution and rapid movement can cause rapid and uneven changes in irradiance within the area. Most existing prediction methods rely on sparse weather data from the site, lack the ability to capture and deduce the temporal and spatial dynamic changes in irradiance caused by moving cloud shadows, and are difficult to achieve short-term advance prediction of distributed photovoltaic park power. Second, existing prediction models generally ignore the influence of photovoltaic controller adjustment on output characteristics. Specifically, when the intensity of light increases, the active power output of photovoltaics increases, causing the voltage at the photovoltaic grid connection point to rise. Once the node voltage reaches the allowed upper limit, the photovoltaic inverter will automatically switch to a different operating mode, such as adjusting the output power factor, and the active power output characteristics will change nonlinearly. Existing prediction methods do not include the voltage dynamic adjustment mechanism in the model, resulting in a significant deviation of the power prediction results near the voltage critical point from the actual value, and the inability to reflect the true photovoltaic output under grid operating condition constraints.

[0004] Therefore, in view of the deficiencies of the prior art, an intelligent prediction method that can integrate weather spatiotemporal deduction and grid interaction dynamic response characteristics is needed to improve the accuracy of short-term prediction of distributed photovoltaic power generation. SUMMARY

[0005] (1) Technical problem to be solved The purpose of the present application is to provide a distributed photovoltaic power intelligent prediction method and system to achieve short-term advance prediction of distributed photovoltaic park power in high-altitude areas with abundant sunshine and stable stratocumulus weather.

[0006] (2) Technical solution To achieve the above purpose, the present application provides a distributed photovoltaic power intelligent prediction method, which comprises the following steps: S1, mapping the to-be-predicted area to a two-dimensional plane coordinate system according to latitude and longitude data of the to-be-predicted area to be predicted; numbering all distributed photovoltaics in the to-be-predicted area, and marking them as a first distributed photovoltaic to an nth distributed photovoltaic respectively. N S2, obtaining wind speed and wind direction data measured by an anemometer uniformly distributed in the to-be-predicted area, fitting to obtain a wind vector distribution function, and obtaining current irradiance data measured by an irradiance meter uniformly distributed in the to-be-predicted area, fitting to obtain a current irradiance distribution function. N S3, obtaining actual active power of the first distributed photovoltaic to the nth distributed photovoltaic, and marking them as a first actual active power to an nth actual active power respectively. N S4, according to the wind vector distribution function, the current irradiance distribution function, the first actual active power to the nth actual active power, the first position to the nth position, performing space-time deduction on the spatial distribution of the actual active power, and calculating to obtain predicted output active power of the first distributed photovoltaic to the nth distributed photovoltaic at a to-be-predicted time point. N S5, the first distributed photovoltaic to the nth distributed photovoltaic have the same rated active power and rated apparent power.

[0007] S2, obtaining wind speed and wind direction data measured by an anemometer uniformly distributed in the to-be-predicted area, fitting to obtain a wind vector distribution function, and obtaining current irradiance data measured by an irradiance meter uniformly distributed in the to-be-predicted area, fitting to obtain a current irradiance distribution function.

[0008] S3, obtaining actual active power of the first distributed photovoltaic to the nth distributed photovoltaic, and marking them as a first actual active power to an nth actual active power respectively. N S4, according to the wind vector distribution function, the current irradiance distribution function, the first actual active power to the nth actual active power, the first position to the nth position, performing space-time deduction on the spatial distribution of the actual active power, and calculating to obtain predicted output active power of the first distributed photovoltaic to the nth distributed photovoltaic at a to-be-predicted time point. N S5, the first distributed photovoltaic to the nth distributed photovoltaic have the same rated active power and rated apparent power.

[0009] S2, obtaining wind speed and wind direction data measured by an anemometer uniformly distributed in the to-be-predicted area, fitting to obtain a wind vector distribution function, and obtaining current irradiance data measured by an irradiance meter uniformly distributed in the to-be-predicted area, fitting to obtain a current irradiance distribution function. N N S3, obtaining actual active power of the first distributed photovoltaic to the nth distributed photovoltaic, and marking them as a first actual active power to an nth actual active power respectively. N S4, according to the wind vector distribution function, the current irradiance distribution function, the first actual active power to the nth actual active power, the first position to the nth position, performing space-time deduction on the spatial distribution of the actual active power, and calculating to obtain predicted output active power of the first distributed photovoltaic to the nth distributed photovoltaic at a to-be-predicted time point.

[0010] Further, the method for obtaining wind speed and wind direction data measured by an anemometer uniformly distributed in the to-be-predicted area, fitting to obtain a wind vector distribution function, and obtaining current irradiance data measured by an irradiance meter uniformly distributed in the to-be-predicted area, fitting to obtain a current irradiance distribution function comprises: using a spatial interpolation algorithm to fit the discrete wind speed and wind direction data measured by the anemometer uniformly distributed in the to-be-predicted area into a continuous wind vector distribution function covering the two-dimensional plane coordinate system; the wind vector distribution function comprises a horizontal axis wind speed distribution function and a vertical axis wind speed distribution function ; the horizontal axis wind speed distribution function represents the component of the wind speed at each two-dimensional plane coordinate point in the horizontal axis direction of the two-dimensional plane coordinate system; and the vertical axis wind speed distribution function represents the component of the wind speed at each two-dimensional plane coordinate point in the vertical axis direction of the two-dimensional plane coordinate system. represents the horizontal coordinate value of the two-dimensional plane coordinate point. ​The ordinate value of a two-dimensional plane coordinate point.

[0011] The current irradiance data measured by the irradiance meter is fitted into a continuous current irradiance distribution function covering the two-dimensional plane coordinate system by using a spatial interpolation algorithm. .

[0012] Further, the spatial distribution of the actual active power is spatio-temporally extrapolated according to the wind vector distribution function of the prediction slice, the current irradiance distribution function, the first actual active power to the first N actual active power, the first position to the first N position, and the method for calculating the predicted output active power of the first distributed photovoltaic to the first N distributed photovoltaic at the to-be-predicted time point comprises: The predicted irradiance distribution function is calculated according to the wind vector distribution function of the prediction slice and the current irradiance distribution function .

[0013] The actual output power factor of the first distributed photovoltaic to the first N distributed photovoltaic is obtained; the distributed photovoltaic whose actual output power factor is between the pre-set lower limit of the power factor and the upper limit of the power factor is screened out from the first distributed photovoltaic to the first N distributed photovoltaic, and is marked as a constant power factor operation photovoltaic; the actual active power corresponding to the constant power factor operation photovoltaic is obtained, and is recorded as a first active power fitting sequence; the two-dimensional plane coordinate point corresponding to the installation geographical position of the constant power factor operation photovoltaic is obtained, and is substituted into the current irradiance distribution function to calculate the current irradiance data corresponding to the installation geographical position of the constant power factor operation photovoltaic, and is recorded as a first current irradiance fitting sequence; the irradiation power linear correlation function is fitted by using the least square method fitting method with the first current irradiance fitting sequence as the independent variable and the first active power fitting sequence as the dependent variable.

[0014] The distributed photovoltaic whose actual output power factor is not between the pre-set lower limit of the power factor and the upper limit of the power factor is screened out from the first distributed photovoltaic to the first N distributed photovoltaic, and is marked as a variable power factor operation photovoltaic; the actual active power corresponding to the variable power factor operation photovoltaic is obtained, and is recorded as a second active power fitting sequence; the two-dimensional plane coordinate point corresponding to the installation geographical position of the variable power factor operation photovoltaic is obtained, and is substituted into the current irradiance distribution function , the current irradiance data corresponding to the installation geographic position of the photovoltaic operating at the variable power factor is calculated, and is recorded as a second current irradiance fitting sequence; the second current irradiance fitting sequence is taken as an independent variable, the second active power fitting sequence is taken as a dependent variable, and a polynomial fitting method is used to fit to obtain an irradiation power nonlinear correlation function.

[0015] The first distributed photovoltaic is connected to the second distributed photovoltaic in sequence. N The two-dimensional plane coordinate point corresponding to the installation geographic position of the distributed photovoltaic is substituted into the predicted irradiance distribution function to obtain the first predicted irradiance to the second predicted irradiance. N The predicted irradiance; the first predicted irradiance to the second predicted irradiance is input into the irradiation power linear correlation function, and the first initial predicted active power to the second initial predicted active power is output and obtained. N The predicted irradiance; the first predicted irradiance to the second predicted irradiance is input into the irradiation power linear correlation function, and the first initial predicted active power to the second initial predicted active power is output and obtained. N The predicted irradiance; the first predicted irradiance to the second predicted irradiance is input into the irradiation power linear correlation function, and the first initial predicted active power to the second initial predicted active power is output and obtained.

[0016] The grid simulation calculation model of the to-be-predicted area is established, and the first initial predicted active power to the second initial predicted active power is substituted into the grid simulation calculation model of the to-be-predicted area for simulation calculation. N The first distributed photovoltaic to the second distributed photovoltaic in the to-be-predicted area is filtered out according to the simulation value of the effective value of the grid node bus voltage. N The predicted variable power factor operating photovoltaic in the first distributed photovoltaic to the second distributed photovoltaic is calculated according to the irradiation power nonlinear correlation function, and the variable power factor predicted active power of the predicted variable power factor operating photovoltaic is calculated. N The variable power factor predicted active power is used to replace the initial predicted active power in the first initial predicted active power to the second initial predicted active power, and the final predicted output active power of the first distributed photovoltaic to the second distributed photovoltaic is iteratively calculated. N The predicted output active power of the first distributed photovoltaic to the second distributed photovoltaic.

[0017] Further, the method for calculating the predicted irradiance distribution function according to the wind vector distribution function of the predicted area and the current irradiance distribution function comprises the following steps. The predicted irradiance distribution function is calculated according to the wind vector distribution function of the predicted area and the current irradiance distribution function. The calculation method of the predicted irradiance distribution function is as follows. According to the values of x , y , a two-dimensional plane coordinate system is divided into two areas, which are respectively recorded as a first area and a second area. and ; wherein , respectively represent the minimum value and the maximum value of the horizontal coordinate of the two-dimensional plane coordinate point obtained in advance in the to-be-predicted area. , respectively represent the minimum and maximum values of the longitudinal coordinate of the two-dimensional plane coordinate points in the to-be-predicted region obtained in advance; represents a preset time interval, i.e., the difference between the prediction time point and the current time point; represents a preset cloud movement ratio coefficient, which is obtained through historical data statistics; the cloud movement ratio coefficient reflects the ratio of the moving speed of the cirrus cloud to the wind speed when the cirrus cloud moves under the action of the wind; the second region is the remaining area after the first region is excluded from the to-be-predicted region.

[0018] In the first region, The calculation formula of is: In the second region, The value of is simplified to be equal to .

[0019] Further, the power grid simulation calculation model of the to-be-predicted region is established, and the first initial predicted active power to the nth initial predicted active power is substituted into the power grid simulation calculation model of the to-be-predicted region for simulation calculation. N The simulation value of the effective value of the bus voltage of the power grid node is obtained, and the first distributed photovoltaic to the nth distributed photovoltaic is screened out. N The variable power factor prediction active power of the predicted variable power factor operation photovoltaic is calculated according to the irradiation power nonlinear correlation function, and the variable power factor prediction active power is used to replace the initial predicted active power in the first initial predicted active power to the nth initial predicted active power. N The final predicted output active power of the first distributed photovoltaic to the nth distributed photovoltaic is obtained through iterative calculation. N The method for obtaining the predicted output active power of the first distributed photovoltaic to the nth distributed photovoltaic comprises the following steps: Obtain the power grid topology data and line parameters of the to-be-predicted region, and establish a power grid simulation calculation model of the to-be-predicted region; in the power grid simulation calculation model of the to-be-predicted region, the active power of the first distributed photovoltaic to the nth distributed photovoltaic is set to the first initial predicted active power to the nth initial predicted active power respectively. N The power factor of the first distributed photovoltaic to the nth distributed photovoltaic is set to the average value of the lower limit of the power factor and the upper limit of the power factor. N The power factor of the first distributed photovoltaic to the nth distributed photovoltaic is set to the average value of the lower limit of the power factor and the upper limit of the power factor. N The power factor of the first distributed photovoltaic to the nth distributed photovoltaic is set to the average value of the lower limit of the power factor and the upper limit of the power factor. N The simulation value of the effective value of the bus voltage of the power grid node connected to the first distributed photovoltaic to the nth distributed photovoltaic is calculated by using the power flow simulation calculation method.

[0020] The simulation value of the effective value of the bus voltage of the power grid node connected to the first distributed photovoltaic to the nth distributed photovoltaic is calculated by using the power flow simulation calculation method. N ​The simulation value of the effective value of the bus voltage of the grid node connected by the distributed photovoltaic is screened, and the distributed photovoltaic whose simulation value is not between the lower limit voltage and the upper limit voltage is marked as a predicted variable power factor operation photovoltaic. N The predicted irradiance corresponding to the predicted variable power factor operation photovoltaic is screened in the predicted irradiance, and is marked as a variable power factor operation predicted irradiance; the variable power factor operation predicted irradiance is input into an irradiation power nonlinear correlation function respectively, and a variable power factor predicted active power is output. N In the initial predicted active power, the initial predicted active power corresponding to the predicted variable power factor operation photovoltaic is replaced with the variable power factor predicted active power; the active power after the replacement is taken as the active power of the current round, and the power factor of the predicted variable power factor operation photovoltaic is set as a preset hysteresis power factor value for voltage regulation, to obtain a distributed photovoltaic power condition of the current round.

[0021] The iteration step is that the power flow simulation calculation is re-executed, the simulation value of the effective value of the bus voltage of the grid node connected by the first distributed photovoltaic to the last distributed photovoltaic is obtained, the predicted variable power factor operation photovoltaic is re-screened, and the variable power factor predicted active power is calculated. N The iteration step is that the power flow simulation calculation is re-executed, the simulation value of the effective value of the bus voltage of the grid node connected by the first distributed photovoltaic to the last distributed photovoltaic is obtained, the predicted variable power factor operation photovoltaic is re-screened, and the variable power factor predicted active power is calculated.

[0022] The iteration termination condition is that the simulation value of the effective value of the bus voltage of the grid node connected by the first distributed photovoltaic to the last distributed photovoltaic is between the lower limit voltage and the upper limit voltage, or the iteration number reaches a preset maximum iteration number. N The iteration termination condition is that the simulation value of the effective value of the bus voltage of the grid node connected by the first distributed photovoltaic to the last distributed photovoltaic is between the lower limit voltage and the upper limit voltage, or the iteration number reaches a preset maximum iteration number.

[0023] Based on the same inventive concept, in another aspect, the application also provides a distributed photovoltaic power intelligent prediction system, which comprises: A data reading module is configured to map a to-be-predicted area to a two-dimensional plane coordinate system according to pre-set latitude and longitude data of the to-be-predicted area; and all distributed photovoltaics in the to-be-predicted area are numbered and marked as a first distributed photovoltaic to a last distributed photovoltaic. N The first distributed photovoltaic to the last distributed photovoltaic are obtained. N The installation geographical positions of the first distributed photovoltaic to the last distributed photovoltaic correspond to two-dimensional plane coordinate points, which are marked as a first position to a last position. N The first distributed photovoltaic to the last distributed photovoltaic are obtained. N The rated active power and the rated apparent power of the first distributed photovoltaic to the last distributed photovoltaic are the same; and the plane shape of the to-be-predicted area is a rectangle.

[0024] The data fitting module, connected to the data reading module, is used to acquire wind speed and direction data measured by an anemometer evenly distributed in the area to be predicted, and to fit the wind vector distribution function. It also acquires the current irradiance data measured by a radiometer evenly distributed in the area to be predicted, and to fit the current irradiance distribution function.

[0025] The power reading module, connected to the data fitting module, is used to obtain power data from the first distributed photovoltaic power generation unit to the second distributed photovoltaic power generation unit. N The actual active power of distributed photovoltaic power is denoted as the first actual active power to the second. N Actual active power.

[0026] The power prediction module, connected to the power reading module, is used to predict the wind vector distribution function of the predicted area, the current irradiance distribution function, and the first actual active power up to the second... N Actual active power, first position to the first N The location is used to perform spatiotemporal extrapolation of the spatial distribution of the actual active power, and the results are calculated to obtain the time points from the first distributed photovoltaic power station to the second distributed photovoltaic power station at the time to be predicted. N Predicted output active power of distributed photovoltaic power.

[0027] Furthermore, the data fitting module includes: The wind vector distribution function fitting module is used to fit the discrete wind speed and direction data obtained from anemometers uniformly distributed across the area to be predicted into a continuous wind vector distribution function covering the two-dimensional plane coordinate system using a spatial interpolation algorithm; the wind vector distribution function includes the horizontal axis wind speed distribution function. and vertical axis wind speed distribution function The horizontal axis wind speed distribution function represents the component of the wind speed at each two-dimensional plane coordinate point in the horizontal direction of the two-dimensional plane coordinate system; the vertical axis wind speed distribution function represents the component of the wind speed at each two-dimensional plane coordinate point in the vertical direction of the two-dimensional plane coordinate system. This represents the x-coordinate value of a point in a two-dimensional plane. This represents the ordinate value of a point in a two-dimensional plane.

[0028] The current irradiance distribution function fitting module, connected to the wind vector distribution function fitting module, is used to fit the discrete, uniformly distributed radiometer measurements of the area to be predicted into a continuous current irradiance distribution function covering the two-dimensional plane coordinate system using a spatial interpolation algorithm. .

[0029] Furthermore, the power prediction module includes: The module for calculating the predicted irradiance distribution function is used to calculate the predicted irradiance distribution function based on the wind vector distribution function of the predicted area and the current irradiance distribution function. .

[0030] The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic.

[0031] The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic.

[0032] The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic.

[0033] The iteration calculation module is connected with the initial predicted active power calculation module, and is configured to establish a power grid simulation calculation model of the to-be-predicted area, and to perform simulation calculation on the initial predicted active power to the first initial predicted active power to the last initial predicted active power. N The initial predicted active power is substituted by the variable power factor predicted active power in the initial predicted active power, and the final predicted active power of the first distributed photovoltaic to the last distributed photovoltaic is obtained through iteration calculation. N The variable power factor predicted active power of the predicted variable power factor operation photovoltaic is calculated according to the nonlinear correlation function of the irradiance power, and the variable power factor predicted active power is obtained. N The initial predicted active power is substituted by the variable power factor predicted active power in the initial predicted active power, and the final predicted active power of the first distributed photovoltaic to the last distributed photovoltaic is obtained through iteration calculation. N The predicted output active power of the distributed photovoltaic.

[0034] Further, the predicted irradiance distribution function calculation module comprises: The segmented calculation module is configured to obtain the predicted irradiance distribution function through segmented calculation according to the wind vector distribution function of the to-be-predicted area and the current irradiance distribution function. The calculation method of the predicted irradiance distribution function is: According to the values of x , y , a two-dimensional plane coordinate system is divided into two areas, which are respectively recorded as a first area and a second area; the first area is and ; wherein , respectively represent the minimum and maximum values of the horizontal coordinates of the two-dimensional plane coordinate points in the to-be-predicted area obtained in advance; , respectively represent the minimum and maximum values of the vertical coordinates of the two-dimensional plane coordinate points in the to-be-predicted area obtained in advance; represents a time interval set in advance, that is, the difference between the prediction time point and the current time point; represents a cloud movement proportion coefficient set in advance, which is obtained through historical data statistics; the cloud movement proportion coefficient reflects the ratio of the moving speed of the wind-driven thin cumulus to the wind speed; the second area is the remaining area after the first area is deducted from the to-be-predicted area.

[0035] In the first area, The calculation formula of ; In the second area, The value of is simplified to be equal to

[0036] Furthermore, the iterative calculation module includes: The initial simulation module is used to acquire the power grid topology data and line parameters of the area to be predicted, and to establish a power grid simulation calculation model for the area to be predicted. In the power grid simulation calculation model of the area to be predicted, the first distributed photovoltaic system to the second... N The active power of distributed photovoltaic systems is set as the first initial predicted active power to the second... N Initial predicted active power; [Transferring the first distributed photovoltaic power to the second...] N The power factor of distributed photovoltaic (PV) systems is set as the average of the lower and upper power factor limits. Power flow simulation methods are used to calculate the power factors from the first distributed PV system to the [number missing]th [system / system / etc.]. N Simulated values ​​of the effective value of the bus voltage of the grid node connected to the distributed photovoltaic system.

[0037] The iterative simulation module, connected to the initial simulation module, is used to simulate the first distributed photovoltaic system to the second... N In distributed photovoltaic (PV) systems, those PV systems whose simulated effective values ​​of the bus voltage at connected grid nodes are not within the preset lower to upper voltage limits are marked as PV systems operating under predicted variable power factor (VPF) conditions. N The predicted irradiance corresponding to photovoltaic systems operating at the predicted variable power factor (VPF) is selected from the predicted irradiance and labeled as VPF predicted irradiance. These VPF predicted irradiances are then input into a nonlinear correlation function for irradiance power, and the output is the predicted active power at the VPF. From the first initial predicted active power to the... N In the initial predicted active power, the initial predicted active power corresponding to the photovoltaic system operating with the predicted variable power factor is replaced with the active power predicted with the variable power factor; the replaced active power is used as the active power of the current cycle, and the power factor of the photovoltaic system operating with the predicted variable power factor is set to a preset lag power factor value for voltage regulation to obtain the distributed photovoltaic power condition of the current cycle; the iterative steps are repeated until the iteration termination condition is met.

[0038] The iterative steps are as follows: re-execute the power flow simulation calculation to obtain the results from the first distributed photovoltaic power generation to the [missing information]. N Simulated values ​​of the effective values ​​of the bus voltage of the grid nodes connected to the distributed photovoltaic system; re-selected photovoltaic systems operating with predicted variable power factor and calculated the predicted active power of the variable power factor; updated the power conditions of the distributed photovoltaic system in the current cycle.

[0039] The iteration termination condition is: from the first distributed photovoltaic to the... N The simulated values ​​of the effective voltage of the grid node bus connected to the distributed photovoltaic system are all between the lower voltage limit and the upper voltage limit, or the number of iterations reaches the preset maximum number of iterations.

[0040] (3) Beneficial effects Compared with the prior art, the present application has the beneficial effects that: 1. According to the measured data of the wind measuring instrument and the irradiance meter uniformly distributed in the block, the spatial distribution function of the wind vector and the irradiance is fitted by using a spatial interpolation algorithm, and the irradiance distribution is deduced in space and time based on the wind field, so that the dynamic change of irradiance caused by the movement of the altocumulus cloud can be accurately captured, and the high-precision prior prediction of the distributed photovoltaic short-term power in the high-altitude area is realized.

[0041] 2. By establishing a power grid simulation model, the initial power prediction result is substituted into the power flow calculation to identify the photovoltaic that needs to switch the operation mode due to voltage out-of-limit, and the active power of the photovoltaic is iteratively corrected by using a nonlinear correlation function, so that the influence of the inverter autonomous adjustment on the output is accurately reflected, and the accuracy of the power prediction near the voltage critical point is improved. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 It is a flow chart of a distributed photovoltaic power intelligent prediction method according to Embodiment 1 of the present application. Figure 2 It is a module composition schematic diagram of a distributed photovoltaic power intelligent prediction system according to Embodiment 2 of the present application. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0044] Before examples are given, the application scenario of the concept of the present application needs to be described. The present application is applied to the short-term power prior prediction in a distributed photovoltaic park in a high-altitude area where the altocumulus weather dominates. The photovoltaic devices in the distributed photovoltaic park adopt the same model, have the same rated active power and rated apparent power, and are all equipped with voltage regulation functions.

[0045] Embodiment 1: As shown in the figure, the present embodiment provides a distributed photovoltaic power intelligent prediction method, which comprises the following steps: Figure 1 S1, according to the longitude and latitude data of the to-be-predicted block area preset in advance, the to-be-predicted block area is mapped into a two-dimensional plane coordinate system; all distributed photovoltaics in the to-be-predicted block area are numbered and respectively marked as the first distributed photovoltaic to the nth distributed photovoltaic. S2, the first distributed photovoltaic to the nth distributed photovoltaic are acquired; the first distributed photovoltaic to the nth distributed photovoltaic are respectively marked as the first photovoltaic to the nth photovoltaic. N S3, the first photovoltaic to the nth photovoltaic are respectively marked as the first irradiance to the nth irradiance; the first irradiance to the nth irradiance are respectively acquired; the first irradiance to the nth irradiance are respectively marked as the first wind vector to the nth wind vector; the first wind vector to the nth wind vector are respectively acquired. NThe two-dimensional plane coordinate points corresponding to the installation geographical positions of the distributed photovoltaics are respectively denoted as a first position to an nth position N The first distributed photovoltaic to the nth distributed photovoltaic N The rated active power and the rated apparent power of the distributed photovoltaics are the same.

[0046] S2, wind speed and wind direction data measured by anemometers uniformly distributed in the to-be-predicted area are obtained, and a wind vector distribution function is fitted; current irradiance data measured by irradiance meters uniformly distributed in the to-be-predicted area are obtained, and a current irradiance distribution function is fitted.

[0047] S3, the first distributed photovoltaic to the nth distributed photovoltaic N The actual active power of the distributed photovoltaics is denoted as a first actual active power to an nth actual active power N The actual active power.

[0048] S4, according to the wind vector distribution function, the current irradiance distribution function, the first actual active power to the nth actual active power, the first position to the nth position, and the spatial distribution of the actual active power, a time-space deduction of the spatial distribution of the actual active power is performed, and the predicted output active power of the first distributed photovoltaic to the nth distributed photovoltaic at the to-be-predicted time point is calculated. N The actual active power. N The actual active power. N The actual active power.

[0049] Exemplarily, according to pre-set latitude and longitude data of the to-be-predicted area, a map projection transformation method is used to map the to-be-predicted area into a two-dimensional plane rectangular coordinate system. The to-be-predicted area is rectangular in shape to simplify coordinate mapping and subsequent calculation. If the actual plane shape of the distributed photovoltaic park is not a regular rectangle, the to-be-predicted area is obtained by expanding the area around the distributed photovoltaic park, so that the to-be-predicted area is rectangular in shape. All the distributed photovoltaics in the to-be-predicted area are uniformly numbered, denoted as a first distributed photovoltaic to an nth distributed photovoltaic. The coordinate points corresponding to the installation geographical positions of each distributed photovoltaic in the constructed two-dimensional plane coordinate system are obtained, respectively denoted as a first position to an nth position. Since the to-be-predicted area belongs to the same park management institution, in order to obtain the cost advantage of batch procurement, the first to the nth distributed photovoltaics are of the same model, and the rated active power and the rated apparent power are the same. N The actual active power. N The actual active power. N The actual active power.

[0050] Real-time wind speed and direction data from multiple anemometers uniformly distributed within the area to be predicted, and current irradiance data from multiple radiometers are acquired. Using the monitoring data collected by the spatially discrete anemometers and radiometers, a spatial interpolation algorithm is employed for fitting, resulting in a continuous spatial distribution function covering the entire two-dimensional coordinate system. Specifically, the discrete wind speed and direction data are fitted into a continuous wind vector distribution function, which describes the wind vector at any coordinate point in the area; the discrete irradiance data are fitted into a continuous current irradiance distribution function, which describes the light intensity at any coordinate point in the area.

[0051] The photovoltaic power generation monitoring system obtains real-time data from the first distributed photovoltaic power generation system to the second distributed photovoltaic power generation system. N The actual active power output of distributed photovoltaic power at the current moment is denoted as the first actual active power up to the second. N Actual active power. The first actual active power to the second... N Actual active power reflects the real power generation capacity of each distributed photovoltaic system under current meteorological and grid conditions.

[0052] Using the wind vector distribution function as the dynamic factor for extrapolating the spatial movement of meteorological conditions (especially cloud shadows), spatiotemporal extrapolation is performed on the current irradiance distribution and actual active power distribution. This simulation calculates the impact of the spatial evolution of meteorological conditions at a future, predicted time point, on the first distributed photovoltaic power generation system to the second... N The new meteorological environment and power grid operation status that each distributed photovoltaic (PV) unit will face at its installation location are then used to calculate the predicted active power output of each PV unit at that future time point (i.e., the time point to be predicted).

[0053] Furthermore, the method for obtaining wind speed and direction data uniformly distributed in the area to be predicted by an anemometer, fitting a wind vector distribution function, and obtaining current irradiance data uniformly distributed in the area to be predicted by a radiometer, fitting a current irradiance distribution function includes: A spatial interpolation algorithm is used to fit the wind speed and direction data obtained from discrete anemometers uniformly distributed across the area to be predicted into a continuous wind vector distribution function covering the two-dimensional plane coordinate system; the wind vector distribution function includes a horizontal axis wind speed distribution function. and vertical axis wind speed distribution function The horizontal axis wind speed distribution function represents the component of the wind speed at each two-dimensional plane coordinate point in the horizontal direction of the two-dimensional plane coordinate system; the vertical axis wind speed distribution function represents the component of the wind speed at each two-dimensional plane coordinate point in the vertical direction of the two-dimensional plane coordinate system. This represents the x-coordinate value of a point in a two-dimensional plane. This represents the ordinate value of a point in a two-dimensional plane.

[0054] A spatial interpolation algorithm is used to fit the current irradiance data obtained from radiometer measurements, which are discretely and uniformly distributed across the area to be predicted, into a continuous current irradiance distribution function covering the two-dimensional plane coordinate system. .

[0055] For example, the data measured by anemometers is processed using a spatial interpolation algorithm. This algorithm, based on discrete wind speed and direction data collected by multiple anemometers distributed throughout the area to be predicted, uses mathematical methods to estimate the wind vector at any point in a two-dimensional coordinate system where no anemometer is located, thereby fitting and generating a continuous wind vector distribution function. Specifically, the wind vector distribution function consists of two component functions: one is a horizontal axis wind speed distribution function, used to describe the magnitude of the wind speed component at any coordinate point along the horizontal axis of the coordinate system; the other is a vertical axis wind speed distribution function, used to describe the magnitude of the wind speed component at any coordinate point along the vertical axis. These two mutually perpendicular component functions can comprehensively characterize the spatial distribution and directional features of the wind field throughout the entire area.

[0056] The data measured by radiometers is also processed using a spatial interpolation algorithm. This algorithm, based on discrete current irradiance data collected by multiple radiometers uniformly distributed within the area to be predicted, estimates the irradiance value at any coordinate point within the area through interpolation and extrapolation, thereby fitting and generating a continuous current irradiance distribution function. This function can accurately reflect the spatial non-uniformity of light intensity within the area caused by factors such as cloud cover.

[0057] Furthermore, the method based on the predicted wind vector distribution function of the area, the current irradiance distribution function, and the first actual active power to the second... N Actual active power, first position to the first N The location is used to perform spatiotemporal extrapolation of the spatial distribution of the actual active power, and the results are calculated to obtain the time points from the first distributed photovoltaic power station to the second distributed photovoltaic power station at the time to be predicted. N Methods for predicting the output active power of distributed photovoltaic systems include: The predicted irradiance distribution function is calculated based on the wind vector distribution function and the current irradiance distribution function of the predicted area. .

[0058] Obtain the first distributed photovoltaic to the first N The actual output power factor of distributed photovoltaic (PV) power; from the first distributed PV to the second NThe distributed photovoltaic with the actual output power factor in the pre-set power factor lower limit to the power factor upper limit is screened out, and is marked as a constant power factor operation photovoltaic; the actual active power corresponding to the constant power factor operation photovoltaic is obtained, and is marked as a first active power fitting sequence; the two-dimensional plane coordinate point corresponding to the installation geographical position corresponding to the constant power factor operation photovoltaic is obtained, and is substituted into a current irradiance distribution function , the current irradiance data corresponding to the installation geographical position corresponding to the constant power factor operation photovoltaic is calculated, and is marked as a first current irradiance fitting sequence; the first current irradiance fitting sequence is taken as an independent variable, the first active power fitting sequence is taken as a dependent variable, and a least square fitting method is used to fit to obtain an irradiation power linear correlation function.

[0059] The first distributed photovoltaic to the N distributed photovoltaic is screened out, and is marked as a variable power factor operation photovoltaic; the actual active power corresponding to the variable power factor operation photovoltaic is obtained, and is marked as a second active power fitting sequence; the two-dimensional plane coordinate point corresponding to the installation geographical position corresponding to the variable power factor operation photovoltaic is obtained, and is substituted into a current irradiance distribution function , the current irradiance data corresponding to the installation geographical position corresponding to the variable power factor operation photovoltaic is calculated, and is marked as a second current irradiance fitting sequence; the second current irradiance fitting sequence is taken as an independent variable, the second active power fitting sequence is taken as a dependent variable, and a polynomial fitting method is used to fit to obtain an irradiation power nonlinear correlation function.

[0060] The two-dimensional plane coordinate point corresponding to the installation geographical position corresponding to the first distributed photovoltaic to the N distributed photovoltaic is substituted into a predicted irradiance distribution function to obtain a first predicted irradiance to a N predicted irradiance; the first predicted irradiance to the N predicted irradiance is input into the irradiation power linear correlation function, and a first initial predicted active power to a N initial predicted active power is output.

[0061] A power grid simulation calculation model of a to-be-predicted area is established, the first initial predicted active power to the N initial predicted active power is substituted into the power grid simulation calculation model of the to-be-predicted area for simulation calculation, and a first distributed photovoltaic to a N distributed photovoltaic with a predicted variable power factor operation photovoltaic is screened out according to a simulation value of an effective value of a power grid node bus voltage, a variable power factor predicted active power of the predicted variable power factor operation photovoltaic is calculated according to the irradiation power nonlinear correlation function, and the first initial predicted active power to the NThe initial predicted active power is replaced by the predicted active power with variable power factor, and the final first to the N The predicted output active power of the distributed photovoltaic.

[0062] Exemplarily, according to the wind vector distribution function and the current irradiance distribution function, a predicted irradiance distribution function for prediction at a future time point is calculated. The core of this method is to deduce the spatial variation of the irradiance distribution in the future by using the physical law of the wind field driving the cloud layer to move.

[0063] According to the actual operation characteristics of the photovoltaic inverter, it is divided into two operation modes and is differentially modeled. Specifically, the actual output power factor of all distributed photovoltaics is obtained, and the photovoltaics whose actual power factor is in a preset normal operation interval (i.e. between the lower limit of power factor and the upper limit of power factor) are marked as constant power factor operation photovoltaics. The lower limit of power factor is set to 0.88 (leading), and the upper limit of power factor is set to 0.90 (leading). When the actual power factor is between the lower limit of power factor and the upper limit of power factor, the distributed photovoltaic outputs active power and reactive power at the same time. In this operation mode, the power factor of the distributed photovoltaic is theoretically a constant value, but considering the errors that may occur in actual operation, the lower limit of power factor and the upper limit of power factor are set to include the possible errors. The current actual active power of these photovoltaics and the current irradiance data corresponding to the installation position thereof are collected to form a first active power fitting sequence and a first current irradiance fitting sequence, respectively. Subsequently, a least squares fitting method is used to establish a linear correlation function between irradiance and power, with irradiance as the independent variable and active power as the dependent variable. This function reflects the basic linear relationship between photovoltaic output and irradiance intensity in the normal operation state.

[0064] For photovoltaics whose actual power factor deviates from the normal operation interval, they are marked as variable power factor operation photovoltaics. This indicates that they may be in a special voltage regulation mode due to local voltage out-of-limit. Similarly, the data of these photovoltaics are collected to form a second active power fitting sequence and a second current irradiance fitting sequence. Since the power output characteristics in this operation mode are more complex, a polynomial fitting method that can better fit nonlinear relationships is used to establish a nonlinear correlation function between irradiance and power.

[0065] Then, initial power prediction is performed. The installation position coordinates of each distributed photovoltaic are substituted into the predicted irradiance distribution function to obtain the corresponding future predicted irradiance. By inputting all the predicted irradiance values into the aforementioned linear correlation function between irradiance and power, the first initial predicted active power to the N The initial predicted active power. This step assumes that all photovoltaics are in constant power factor operation mode at the predicted time.

[0066] Finally, the power grid simulation is introduced for iterative correction. The power grid simulation calculation model of the to-be-predicted area is established, and the initial predicted active power set is taken as the load condition to be substituted into the model for power flow calculation. By analyzing the bus voltage of each photovoltaic access point obtained through simulation, it can be determined which photovoltaics will lead to voltage out-of-limit under the initial predicted power, and these photovoltaics are marked as predicted variable power factor operation photovoltaics. For these photovoltaics, the irradiance power nonlinear correlation function is used to recalculate the active power (i.e. variable power factor predicted active power) according to the predicted irradiance, and the original initial predicted value is replaced. Subsequently, the updated power set and the adjusted power factor setting are used to re-calculate the power flow, and the iteration is repeated until all node voltages are restored to the allowable range or the maximum number of iterations is reached, thereby obtaining the final converged predicted output active power which is more consistent with the actual operation constraints of the power grid.

[0067] Further, the method of calculating the predicted irradiance distribution function according to the wind vector distribution function of the to-be-predicted area and the current irradiance distribution function comprises: calculating the predicted irradiance distribution function according to the wind vector distribution function of the to-be-predicted area and the current irradiance distribution function in segments ; the calculation method of the predicted irradiance distribution function is: dividing a two-dimensional plane coordinate system into two areas according to the values of x , y , and marking the two areas as a first area and a second area respectively; the first area is and ; wherein , respectively represent the minimum and maximum values of the horizontal coordinates of the two-dimensional plane coordinate points in the to-be-predicted area obtained in advance; , respectively represent the minimum and maximum values of the vertical coordinates of the two-dimensional plane coordinate points in the to-be-predicted area obtained in advance; represents a pre-set time interval, i.e. the difference between the prediction time point and the current time point; represents a pre-set cloud movement proportion coefficient, which is obtained through historical data statistics; the cloud movement proportion coefficient reflects the ratio of the moving speed of the wind-driven thin cumulus cloud to the wind speed; the second area is the remaining area after the first area is deducted from the to-be-predicted area.

[0068] in the first area, the calculation formula is: ; in the second area, the value is simplified to be equal to Equal.

[0069] Exemplarily, according to the wind vector distribution function and the preset time interval, a two-dimensional plane coordinate system is dynamically divided into two pieces with different calculation logics: a first piece and a second piece. This partition processing mode takes into account the calculation accuracy and efficiency.

[0070] The first piece refers to a set of coordinate points satisfying a certain condition. This condition ensures that, under the driving of the wind vector, the presumed cloud layer (or meteorological disturbance) source point is still located within the rectangular boundary of the piece to be predicted after the time interval. Specifically, for a given coordinate point, according to the wind vector component at the point and the prediction time interval, an "upstream" position can be reversely calculated. If this upstream position does not exceed the geographical range of the piece (i.e., its horizontal and vertical coordinates are between the minimum and maximum values of the piece coordinates), the point is classified into the first piece. In this piece, the calculation of the predicted irradiance distribution function has a clear physical meaning: it assumes that the irradiance condition at a certain upstream position at the current time will be "advected" to the target position of the current prediction under the driving of the wind field after the time interval. Therefore, the predicted irradiance value of the point can be calculated by substituting the coordinates of the upstream position into the current irradiance distribution function. In order to more accurately describe the relationship between cloud movement and wind field, an empirical cloud movement proportionality coefficient is introduced in the calculation, which is obtained by analyzing historical meteorological data and is used to represent the proportionality relationship between the movement speed of the cirrus cloud and the ground wind speed in the actual environment.

[0071] The second piece refers to the area remaining after the first piece is deducted from the piece to be predicted, which is usually located at the edge or downwind boundary of the piece. In this piece, for a given coordinate point, the "upstream" position calculated by the above method is located outside the piece, and valid data cannot be obtained from the current irradiance distribution function. Therefore, as a simplified and reasonable approximation, the value of the predicted irradiance distribution function in this piece is set to be equal to the value of the current irradiance distribution function at the point. This is essentially assuming that for these positions, the irradiance condition remains unchanged or follows the current value due to the lack of upstream data within a short time prediction scale. This processing mode ensures the prediction accuracy of the core area while simplifying the processing of the boundary problem, enhancing the engineering practicability of the method. It should be noted that, in order to ensure the calculation has high enough accuracy, , The movement distance should be less than a pre-set movement threshold.

[0072] Further, the power grid simulation calculation model of the piece to be predicted is established, and the first initial predicted active power is calculated to the first predicted active power of the piece to be predicted. NThe initial predicted active power is substituted into the power grid simulation calculation model of the to-be-predicted area for simulation calculation, and the simulation value of the effective value of the node bus voltage of the power grid is screened to obtain the first distributed photovoltaic to the last distributed photovoltaic N The predicted variable power factor operation photovoltaic in the distributed photovoltaic is calculated according to the irradiance power nonlinear correlation function to obtain the variable power factor predicted active power of the predicted variable power factor operation photovoltaic, and the variable power factor predicted active power is substituted into the first initial predicted active power to the last initial predicted active power N In the initial predicted active power, the variable power factor predicted active power is substituted for the initial predicted active power, and the final first distributed photovoltaic to the last distributed photovoltaic is obtained through iterative calculation N The method for predicting the output active power of the distributed photovoltaic comprises the following steps: The power grid topology data and line parameters of the to-be-predicted area are obtained, and a power grid simulation calculation model of the to-be-predicted area is established; in the power grid simulation calculation model of the to-be-predicted area, the first distributed photovoltaic to the last distributed photovoltaic N The active power of the distributed photovoltaic is set as the first initial predicted active power to the last initial predicted active power N The initial predicted active power; the first distributed photovoltaic to the last distributed photovoltaic N The power factor of the distributed photovoltaic is set as the average value of the lower limit of the power factor and the upper limit of the power factor, and the first distributed photovoltaic to the last distributed photovoltaic is calculated by using the power flow simulation calculation method N The simulation value of the effective value of the node bus voltage of the power grid connected to the distributed photovoltaic.

[0073] The first distributed photovoltaic to the last distributed photovoltaic N The distributed photovoltaic connected to the simulation value of the effective value of the node bus voltage of the power grid is screened out, and the distributed photovoltaic whose simulation value of the effective value of the node bus voltage is not between the pre-set lower limit of the voltage and the upper limit of the voltage is marked as the predicted variable power factor operation photovoltaic; the first predicted irradiance to the last predicted irradiance N The predicted irradiance corresponding to the predicted variable power factor operation photovoltaic is screened out in the first predicted irradiance to the last predicted irradiance, and is marked as the variable power factor operation predicted irradiance; the variable power factor operation predicted irradiance is input into the irradiance power nonlinear correlation function respectively, and the variable power factor predicted active power is output; in the first initial predicted active power to the last initial predicted active power N In the initial predicted active power, the initial predicted active power corresponding to the predicted variable power factor operation photovoltaic is replaced by the variable power factor predicted active power; the active power after the replacement is used as the active power of the current round, and the power factor of the predicted variable power factor operation photovoltaic is set as the pre-set lag power factor value for voltage regulation to obtain the distributed photovoltaic power condition of the current round; the iteration step is repeatedly executed until the iteration termination condition is met.

[0074] The iteration step is: re-executing the power flow simulation calculation, obtaining the first distributed photovoltaic to the last distributed photovoltaic NSimulated values ​​of the effective values ​​of the bus voltage of the grid nodes connected to the distributed photovoltaic system; re-selected photovoltaic systems operating with predicted variable power factor and calculated the predicted active power of the variable power factor; updated the power conditions of the distributed photovoltaic system in the current cycle.

[0075] The iteration termination condition is: from the first distributed photovoltaic to the... N The simulated values ​​of the effective voltage of the grid node bus connected to the distributed photovoltaic system are all between the lower voltage limit and the upper voltage limit, or the number of iterations reaches the preset maximum number of iterations.

[0076] For example, firstly, a power grid simulation model for the area to be predicted is established. Specifically, key data such as the power grid topology, line impedance parameters, and transformer turns ratio of the area to be predicted need to be obtained to establish a power grid simulation model that accurately reflects the electrical characteristics of the region. In the initial simulation model, all distributed photovoltaic (PV) systems (from the first to the second generation) are included. N The active power of distributed photovoltaic (PV) is set to the first initial predicted active power up to the [missing value]. N Initial predicted active power. Simultaneously, a uniform initial power factor is set for all photovoltaic (PV) systems. This value is typically the average of the lower and upper power factor limits, for example, 0.89 (leading), to simulate a common initial grid-connected state. Based on this, power flow simulation methods are used to solve for the steady-state operating conditions of the entire grid, thereby obtaining simulated values ​​of the effective voltage of the grid node bus connected to each distributed PV system.

[0077] Then, the simulated bus voltage values ​​for each node are compared with the pre-set lower and upper voltage limits according to grid safety operation standards. Distributed photovoltaic (PV) systems with simulated bus voltage values ​​exceeding the limits (i.e., below the lower voltage limit or above the upper voltage limit) are selected and marked as "predicted variable power factor (VPF) operating PV" for this round. These PV systems are considered to trigger the inverter's control logic and switch operating modes due to voltage issues at the predicted time point. Next, from the first predicted irradiance to the... N In the predicted irradiance, the future irradiance values ​​corresponding to these "predicted variable power factor (VPF) photovoltaic systems" are identified and labeled as "variable power factor (VPF) predicted irradiance." These irradiance values ​​are input into the previously established nonlinear correlation function for irradiance power to calculate the "variable power factor predicted active power," which should have been output without considering voltage constraints but is actually limited due to changes in its operating mode. Then, iterative updates are performed. From the first initial predicted active power to the... NThe calculated "variable power factor predicted active power" is used to replace the initial predicted value of the corresponding photovoltaic in the initial predicted active power set. This updated active power set constitutes the new boundary condition for the current round of simulation. At the same time, in order to more realistically simulate the support of these photovoltaics to the grid voltage in subsequent simulations, the power factor of these "predicted variable power factor operating photovoltaics" is set to the preset lagging power factor value (for example, 0.90 lagging) for voltage regulation, so that they can emit reactive power to help raise the local voltage. The above steps constitute a complete iteration step. This iteration step is repeated: that is, the power flow simulation calculation is performed again with the updated distributed photovoltaic power conditions (active power and power factor of part of the photovoltaics), the new bus voltage simulation value is obtained, and the predicted variable power factor operating photovoltaics are selected again, the new variable power factor predicted active power is calculated, and the power setting is updated.

[0078] This iteration process continues until the preset iteration termination condition is met. The termination condition includes two cases: one is the ideal case, that is, after several iterations, the first to the N The simulation values of the effective values of all bus voltages connected by the distributed photovoltaics are restored to the safe range between the voltage lower limit and the voltage upper limit; the second is to set a maximum number of iterations (for example, 50 times) as a safety guarantee to prevent infinite loops due to calculation oscillation or non-convergence. When either condition is met, the iteration stops, and the active power set obtained at this time is the final active power set that takes into account both meteorological changes and voltage safety constraints of the first to the N The predicted output active power of the distributed photovoltaics.

[0079] Embodiment 2: based on the same inventive concept, as Figure 2 shown, the embodiment also provides a distributed photovoltaic power intelligent prediction system, which comprises: A data reading module is used to map the to-be-predicted area to a two-dimensional plane coordinate system according to the latitude and longitude data of the to-be-predicted area that is set in advance; all distributed photovoltaics in the to-be-predicted area are numbered and are respectively denoted as the first distributed photovoltaic to the nth distributed photovoltaic. N Distributed photovoltaics; the installation geographic positions of the first distributed photovoltaic to the nth distributed photovoltaic correspond to two-dimensional plane coordinate points, which are respectively denoted as the first position to the nth position. N Distributed photovoltaics; the installation geographic positions of the first distributed photovoltaic to the nth distributed photovoltaic correspond to two-dimensional plane coordinate points, which are respectively denoted as the first position to the nth position. N Distributed photovoltaics; the installation geographic positions of the first distributed photovoltaic to the nth distributed photovoltaic correspond to two-dimensional plane coordinate points, which are respectively denoted as the first position to the nth position. N The rated active power of the first distributed photovoltaic to the nth distributed photovoltaic is the same as the rated apparent power; and the plane shape of the to-be-predicted area is a rectangle.

[0080] The data fitting module is connected with the data reading module, and is used for obtaining wind speed and wind direction data measured by an anemometer uniformly distributed in a to-be-predicted area, fitting to obtain a wind vector distribution function, and obtaining current irradiance data measured by an irradiance meter uniformly distributed in the to-be-predicted area, fitting to obtain a current irradiance distribution function.

[0081] The power reading module is connected with the data fitting module, and is used for obtaining the first distributed photovoltaic to the nth distributed photovoltaic actual active power. N The power reading module is connected with the data fitting module, and is used for obtaining the first distributed photovoltaic to the nth distributed photovoltaic actual active power. N The power reading module is connected with the data fitting module, and is used for obtaining the first distributed photovoltaic to the nth distributed photovoltaic actual active power.

[0082] The power prediction module is connected with the power reading module, and is used for performing space-time deduction on a spatial distribution of the actual active power according to the wind vector distribution function, the current irradiance distribution function, the first actual active power to the nth actual active power, the first position to the nth position of the first distributed photovoltaic to the nth distributed photovoltaic, and calculating to obtain the first distributed photovoltaic to the nth distributed photovoltaic predicted output active power at a to-be-predicted time point. N The power prediction module is connected with the power reading module, and is used for performing space-time deduction on a spatial distribution of the actual active power according to the wind vector distribution function, the current irradiance distribution function, the first actual active power to the nth actual active power, the first position to the nth position of the first distributed photovoltaic to the nth distributed photovoltaic, and calculating to obtain the first distributed photovoltaic to the nth distributed photovoltaic predicted output active power at a to-be-predicted time point. N The power prediction module is connected with the power reading module, and is used for performing space-time deduction on a spatial distribution of the actual active power according to the wind vector distribution function, the current irradiance distribution function, the first actual active power to the nth actual active power, the first position to the nth position of the first distributed photovoltaic to the nth distributed photovoltaic, and calculating to obtain the first distributed photovoltaic to the nth distributed photovoltaic predicted output active power at a to-be-predicted time point. N The power prediction module is connected with the power reading module, and is used for performing space-time deduction on a spatial distribution of the actual active power according to the wind vector distribution function, the current irradiance distribution function, the first actual active power to the nth actual active power, the first position to the nth position of the first distributed photovoltaic to the nth distributed photovoltaic, and calculating to obtain the first distributed photovoltaic to the nth distributed photovoltaic predicted output active power at a to-be-predicted time point.

[0083] Further, the data fitting module comprises: The wind vector distribution function fitting module is used for fitting discrete wind speed and wind direction data measured by an anemometer uniformly distributed in a to-be-predicted area into a continuous wind vector distribution function covering the two-dimensional plane coordinate system by using a spatial interpolation algorithm; the wind vector distribution function comprises a horizontal axis wind speed distribution function and a vertical axis wind speed distribution function ; the horizontal axis wind speed distribution function represents a component of wind speed at each two-dimensional plane coordinate point in the horizontal axis direction of the two-dimensional plane coordinate system; and the vertical axis wind speed distribution function represents a component of wind speed at each two-dimensional plane coordinate point in the vertical axis direction of the two-dimensional plane coordinate system. represents a horizontal coordinate value of a two-dimensional plane coordinate point. represents a vertical coordinate value of a two-dimensional plane coordinate point.

[0084] The current irradiance distribution function fitting module is connected with the wind vector distribution function fitting module, and is used for fitting discrete current irradiance data measured by an irradiance meter uniformly distributed in a to-be-predicted area into a continuous current irradiance distribution function covering the two-dimensional plane coordinate system by using a spatial interpolation algorithm. .

[0085] Further, the power prediction module comprises: The predicted irradiance distribution function calculation module is used for calculating a predicted irradiance distribution function according to the wind vector distribution function and the current irradiance distribution function of a prediction area.

[0086] The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic.

[0087] The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic.

[0088] The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic. N The irradiation power linear correlation function fitting module is connected with the predicted irradiance distribution function calculation module, and is configured to obtain the actual output power factor of the first distributed photovoltaic to the last distributed photovoltaic.

[0089] The iteration calculation module is connected with the initial predicted active power calculation module, and is configured to establish a power grid simulation calculation model of the to-be-predicted area, and to perform simulation calculation on the initial predicted active power to the first initial predicted active power to the last initial predicted active power. N The initial predicted active power is substituted by the variable power factor predicted active power in the initial predicted active power, and the final first initial predicted active power to the last initial predicted active power is obtained through iteration calculation. N The predicted variable power factor operating photovoltaic in the distributed photovoltaic is calculated according to the irradiance power nonlinear correlation function, and the variable power factor predicted active power of the predicted variable power factor operating photovoltaic is obtained. N The initial predicted active power is substituted by the variable power factor predicted active power in the initial predicted active power, and the final first initial predicted active power to the last initial predicted active power is obtained through iteration calculation. N The predicted output active power of the distributed photovoltaic.

[0090] Further, the predicted irradiance distribution function calculation module comprises: The segmented calculation module is configured to obtain the predicted irradiance distribution function through segmented calculation according to the wind vector distribution function of the to-be-predicted area and the current irradiance distribution function. The calculation method of the predicted irradiance distribution function is: According to the values of x and y, a two-dimensional plane coordinate system is divided into two areas, which are respectively recorded as a first area and a second area. x , y The first area is x and y ; wherein x and y respectively represent the minimum and maximum values of the horizontal coordinates of the two-dimensional plane coordinate points in the to-be-predicted area obtained in advance. , respectively represent the minimum and maximum values of the vertical coordinates of the two-dimensional plane coordinate points in the to-be-predicted area obtained in advance. represents a time interval set in advance, that is, the difference between the prediction time point and the current time point. represents a cloud movement proportion coefficient set in advance, which is obtained through historical data statistics; the cloud movement proportion coefficient reflects the ratio of the moving speed of the light cumulus to the wind speed under the condition that the light cumulus is driven by the wind; and the second area is the remaining area after the first area is deducted from the to-be-predicted area.

[0091] In the first area, The calculation formula of x ; In the second area, The value of x is simplified to be equal to x .

[0092] Furthermore, the iterative calculation module includes: The initial simulation module is used to acquire the power grid topology data and line parameters of the area to be predicted, and to establish a power grid simulation calculation model for the area to be predicted. In the power grid simulation calculation model of the area to be predicted, the first distributed photovoltaic system to the second... N The active power of distributed photovoltaic systems is set as the first initial predicted active power to the second... N Initial predicted active power; [Transferring the first distributed photovoltaic power to the second...] N The power factor of distributed photovoltaic (PV) systems is set as the average of the lower and upper power factor limits. Power flow simulation methods are used to calculate the power factors from the first distributed PV system to the [number missing]th [system / system / etc.]. N Simulated values ​​of the effective value of the bus voltage of the grid node connected to the distributed photovoltaic system.

[0093] The iterative simulation module, connected to the initial simulation module, is used to simulate the first distributed photovoltaic system to the second... N In distributed photovoltaic (PV) systems, those PV systems whose simulated effective values ​​of the bus voltage at connected grid nodes are not within the preset lower to upper voltage limits are marked as PV systems operating under predicted variable power factor (VPF) conditions. N The predicted irradiance corresponding to photovoltaic systems operating at the predicted variable power factor (VPF) is selected from the predicted irradiance and labeled as VPF predicted irradiance. These VPF predicted irradiances are then input into a nonlinear correlation function for irradiance power, and the output is the predicted active power at the VPF. From the first initial predicted active power to the... N In the initial predicted active power, the initial predicted active power corresponding to the photovoltaic system operating with the predicted variable power factor is replaced with the active power predicted with the variable power factor; the replaced active power is used as the active power of the current cycle, and the power factor of the photovoltaic system operating with the predicted variable power factor is set to a preset lag power factor value for voltage regulation to obtain the distributed photovoltaic power condition of the current cycle; the iterative steps are repeated until the iteration termination condition is met.

[0094] The iterative steps are as follows: re-execute the power flow simulation calculation to obtain the results from the first distributed photovoltaic power generation to the [missing information]. N Simulated values ​​of the effective values ​​of the bus voltage of the grid nodes connected to the distributed photovoltaic system; re-selected photovoltaic systems operating with predicted variable power factor and calculated the predicted active power of the variable power factor; updated the power conditions of the distributed photovoltaic system in the current cycle.

[0095] The iteration termination condition is: from the first distributed photovoltaic to the... N The simulated values ​​of the effective voltage of the grid node bus connected to the distributed photovoltaic system are all between the lower voltage limit and the upper voltage limit, or the number of iterations reaches the preset maximum number of iterations.

[0096] It should be noted that the specific manner in which the various modules perform operations in the system of the above embodiments has been described in detail in the embodiments relating to the method, and will not be described in detail here.

[0097] Finally, it should be noted that although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for part of the technical features, and any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for intelligent prediction of distributed photovoltaic power generation, characterized in that, The method includes the following steps: S1, based on the pre-set latitude and longitude data of the area to be predicted, map the area to be predicted into a two-dimensional plane coordinate system; number all distributed photovoltaic (PV) systems within the area to be predicted, denoted as the first distributed PV system to the next. N Distributed photovoltaic; obtaining the first distributed photovoltaic to the first N The two-dimensional plane coordinates corresponding to the geographical locations of distributed photovoltaic installations are denoted as the first location to the second location. N Location; from the first distributed photovoltaic to the first N The rated active power and rated apparent power of the distributed photovoltaic system are the same; the planar shape of the area to be predicted is rectangular. S2, obtain wind speed and wind direction data measured by an anemometer evenly distributed in the area to be predicted, and fit the wind vector distribution function; obtain current irradiance data measured by an irradiance meter evenly distributed in the area to be predicted, and fit the current irradiance distribution function. S3, obtain the first distributed photovoltaic to the... N The actual active power of distributed photovoltaic power is denoted as the first actual active power to the second. N Actual active power; S4, based on the predicted wind vector distribution function of the area, the current irradiance distribution function, and the first actual active power to the second... N Actual active power, first position to the first N The location is used to perform spatiotemporal extrapolation of the spatial distribution of the actual active power, and the results are calculated to obtain the time points from the first distributed photovoltaic power station to the second distributed photovoltaic power station at the time to be predicted. N Predicted output active power of distributed photovoltaic power.

2. The intelligent prediction method for distributed photovoltaic power generation as described in claim 1, characterized in that, The method for obtaining wind speed and direction data uniformly distributed in the area to be predicted by an anemometer, and fitting a wind vector distribution function thereon, and obtaining current irradiance data uniformly distributed in the area to be predicted by a radiometer, and fitting a current irradiance distribution function thereon, includes: A spatial interpolation algorithm is used to fit the wind speed and direction data obtained from discrete anemometers uniformly distributed across the area to be predicted into a continuous wind vector distribution function covering the two-dimensional plane coordinate system; the wind vector distribution function includes a horizontal axis wind speed distribution function. and vertical axis wind speed distribution function The horizontal axis wind speed distribution function represents the component of the wind speed at each two-dimensional plane coordinate point in the horizontal direction of the two-dimensional plane coordinate system; the vertical axis wind speed distribution function represents the component of the wind speed at each two-dimensional plane coordinate point in the vertical direction of the two-dimensional plane coordinate system. This represents the x-coordinate value of a point in a two-dimensional plane. This represents the ordinate value of a point in a two-dimensional plane. A spatial interpolation algorithm is used to fit the current irradiance data obtained from radiometer measurements, which are discretely and uniformly distributed across the area to be predicted, into a continuous current irradiance distribution function covering the two-dimensional plane coordinate system. .

3. The intelligent prediction method for distributed photovoltaic power generation as described in claim 2, characterized in that, The prediction is based on the wind vector distribution function of the predicted area, the current irradiance distribution function, and the first actual active power to the second... N Actual active power, first position to the first N The location is used to perform spatiotemporal extrapolation of the spatial distribution of the actual active power, and the results are calculated to obtain the time points from the first distributed photovoltaic power station to the second distributed photovoltaic power station at the time to be predicted. N Methods for predicting the output active power of distributed photovoltaic systems include: The predicted irradiance distribution function is calculated based on the wind vector distribution function and the current irradiance distribution function of the predicted area. ; Obtain the first distributed photovoltaic to the first N The actual output power factor of distributed photovoltaic (PV) power; from the first distributed PV to the second N In distributed photovoltaic (PV) systems, PV units whose actual output power factor falls between a pre-set lower and upper power factor limit are selected and marked as constant power factor operating PV units. The actual active power corresponding to the constant power factor operating PV units is obtained and denoted as the first active power fitting sequence. The two-dimensional plane coordinates corresponding to the installation location of the constant power factor operating PV units are obtained and substituted into the current irradiance distribution function. The current irradiance data corresponding to the installation geographical location of the photovoltaic system operating with constant power factor is calculated and denoted as the first current irradiance fitting sequence. The first current irradiance fitting sequence is used as the independent variable and the first active power fitting sequence is used as the dependent variable. The least squares fitting method is used to fit the linear correlation function of irradiance power. First distributed photovoltaic to the first N In distributed photovoltaic (PV) systems, those whose actual output power factor falls outside the preset lower to upper power factor limits are selected and marked as variable power factor (VPF) PV systems. The actual active power corresponding to the VPF PV system is obtained and denoted as the second active power fitting sequence. The two-dimensional plane coordinates corresponding to the installation location of the VPF PV system are obtained and substituted into the current irradiance distribution function. The current irradiance data corresponding to the installation geographical location of the photovoltaic system operating with variable power factor is calculated and denoted as the second current irradiance fitting sequence. The second current irradiance fitting sequence is used as the independent variable and the second active power fitting sequence is used as the dependent variable. The nonlinear correlation function of irradiance power is obtained by fitting the polynomial fitting method. To move the first distributed photovoltaic to the second N Substituting the two-dimensional plane coordinates corresponding to the installation location of distributed photovoltaics into the predicted irradiance distribution function yields the first predicted irradiance to the second. N Predicted irradiance; respectively, the first predicted irradiance to the second... N The predicted irradiance is obtained by inputting the irradiance power linear correlation function and outputting the first initial predicted active power up to the second. N Initial predicted active power; Establish a power grid simulation calculation model for the area to be predicted, and adjust the first initial predicted active power to the second... N The initial predicted active power is substituted into the power grid simulation model of the area to be predicted for simulation calculation. Based on the simulated values ​​of the effective values ​​of the grid node bus voltage, the first to the second distributed photovoltaic power generation projects are selected. N In distributed photovoltaic systems, the predicted variable power factor (VPF) photovoltaic system is calculated based on a nonlinear correlation function of irradiance power to obtain the predicted active power of the VPF system, and the predicted active power is calculated from the first initial predicted active power to the second... N In the initial predicted active power, the predicted active power using the variable power factor is used to replace the initial predicted active power, and the final active power from the first distributed photovoltaic power generation to the second is obtained through iterative calculation. N Predicted output active power of distributed photovoltaic power.

4. The intelligent prediction method for distributed photovoltaic power generation as described in claim 3, characterized in that, The predicted irradiance distribution function is calculated based on the wind vector distribution function of the predicted area and the current irradiance distribution function. The methods include: The predicted irradiance distribution function is obtained by calculating piecewise based on the wind vector distribution function of the predicted area and the current irradiance distribution function. The method for calculating the predicted irradiance distribution function is as follows: according to x , y The value of divides the two-dimensional plane coordinate system into two regions, denoted as the first region and the second region, respectively; the first region is... and ;in , These represent the minimum and maximum x-coordinates of the two-dimensional plane coordinate points within the area to be predicted, obtained in advance. , These represent the minimum and maximum ordinate values ​​of the two-dimensional plane coordinate points within the area to be predicted, obtained in advance. This represents a pre-set time interval, i.e., the difference between the predicted time point and the current time point; The cloud movement ratio is a pre-set cloud movement ratio obtained through historical data statistics; the cloud movement ratio reflects the ratio of the movement speed of cumulus clouds to the wind speed when wind drives the movement of cumulus clouds; the second area is the area remaining after deducting the first area from the area to be predicted. In the first area, The calculation formula is: ； In the second area, The value of is simplified to be the same as equal.

5. The intelligent prediction method for distributed photovoltaic power generation as described in claim 4, characterized in that, The establishment of a power grid simulation calculation model for the area to be predicted will be used to adjust the first initial predicted active power to the second... N The initial predicted active power is substituted into the power grid simulation model of the area to be predicted for simulation calculation. Based on the simulated values ​​of the effective values ​​of the grid node bus voltage, the first to the second distributed photovoltaic power generation projects are selected. N In distributed photovoltaic systems, the predicted variable power factor (VPF) photovoltaic system is calculated based on a nonlinear correlation function of irradiance power to obtain the predicted active power of the VPF system, and the predicted active power is calculated from the first initial predicted active power to the second... N In the initial predicted active power, the predicted active power using the variable power factor is used to replace the initial predicted active power, and the final active power from the first distributed photovoltaic power generation to the second is obtained through iterative calculation. N Methods for predicting the output active power of distributed photovoltaic systems include: Obtain the power grid topology data and line parameters of the area to be predicted, and establish a power grid simulation calculation model for the area to be predicted; in the power grid simulation calculation model of the area to be predicted, the first distributed photovoltaic power generation system to the second distributed photovoltaic power generation system will be integrated into the grid. N The active power of distributed photovoltaic systems is set as the first initial predicted active power to the second... N Initial predicted active power; [Transferring the first distributed photovoltaic power to the second...] N The power factor of distributed photovoltaic (PV) systems is set as the average of the lower and upper power factor limits. Power flow simulation methods are used to calculate the power factors from the first distributed PV system to the [number missing]th [system / system / etc.]. N Simulated values ​​of the effective value of the bus voltage of the grid node connected to the distributed photovoltaic system; First distributed photovoltaic to the first N In distributed photovoltaic (PV) systems, those PV systems whose simulated effective values ​​of the bus voltage at connected grid nodes are not within the preset lower to upper voltage limits are marked as PV systems operating under predicted variable power factor (VPF) conditions. N The predicted irradiance corresponding to photovoltaic systems operating at the predicted variable power factor (VPF) is selected from the predicted irradiance and labeled as VPF predicted irradiance. These VPF predicted irradiances are then input into a nonlinear correlation function for irradiance power, and the output is the predicted active power at the VPF. From the first initial predicted active power to the... N In the initial predicted active power, the initial predicted active power corresponding to the photovoltaic system operating with the predicted variable power factor is replaced with the active power predicted with the variable power factor; the replaced active power is used as the active power of the current cycle, and the power factor of the photovoltaic system operating with the predicted variable power factor is set to a preset lag power factor value for voltage regulation to obtain the distributed photovoltaic power condition of the current cycle; the iterative steps are repeated until the iteration termination condition is met. The iterative steps are as follows: re-execute the power flow simulation calculation to obtain the results from the first distributed photovoltaic power generation to the [missing information]. N Simulated values ​​of the effective values ​​of the bus voltage of the grid nodes connected to the distributed photovoltaic system; re-selected photovoltaic systems operating with predicted variable power factor and calculated the predicted active power of the variable power factor; updated the power conditions of the distributed photovoltaic system in the current round. The iteration termination condition is: from the first distributed photovoltaic to the... N The simulated values ​​of the effective voltage of the grid node bus connected to the distributed photovoltaic system are all between the lower voltage limit and the upper voltage limit, or the number of iterations reaches the preset maximum number of iterations.

6. A distributed photovoltaic power generation intelligent prediction system, characterized in that, The system includes: The data reading module is used to map the area to be predicted into a two-dimensional plane coordinate system based on the pre-set latitude and longitude data of the area to be predicted; and to number all distributed photovoltaic (PV) systems within the area to be predicted, designating them as the first distributed PV system to the next. N Distributed photovoltaic; obtaining the first distributed photovoltaic to the first N The two-dimensional plane coordinates corresponding to the geographical locations of distributed photovoltaic installations are denoted as the first location to the second location. N Location; from the first distributed photovoltaic to the first N The rated active power and rated apparent power of the distributed photovoltaic system are the same; the planar shape of the area to be predicted is rectangular. The data fitting module, connected to the data reading module, is used to acquire wind speed and wind direction data measured by an anemometer evenly distributed in the area to be predicted, and to fit the wind vector distribution function; and to acquire the current irradiance data measured by a radiometer evenly distributed in the area to be predicted, and to fit the current irradiance distribution function. The power reading module, connected to the data fitting module, is used to obtain power data from the first distributed photovoltaic power generation unit to the second distributed photovoltaic power generation unit. N The actual active power of distributed photovoltaic power is denoted as the first actual active power to the second. N Actual active power; The power prediction module, connected to the power reading module, is used to predict the wind vector distribution function of the predicted area, the current irradiance distribution function, and the first actual active power up to the second... N Actual active power, first position to the first N The location is used to perform spatiotemporal extrapolation of the spatial distribution of the actual active power, and the results are calculated to obtain the time points from the first distributed photovoltaic power station to the second distributed photovoltaic power station at the time to be predicted. N Predicted output active power of distributed photovoltaic power.

7. The distributed photovoltaic power generation intelligent prediction system as described in claim 6, characterized in that, The data fitting module includes: The wind vector distribution function fitting module is used to fit the discrete wind speed and direction data obtained from anemometers uniformly distributed across the area to be predicted into a continuous wind vector distribution function covering the two-dimensional plane coordinate system using a spatial interpolation algorithm; the wind vector distribution function includes the horizontal axis wind speed distribution function. and vertical axis wind speed distribution function The horizontal axis wind speed distribution function represents the component of the wind speed at each two-dimensional plane coordinate point in the horizontal direction of the two-dimensional plane coordinate system; the vertical axis wind speed distribution function represents the component of the wind speed at each two-dimensional plane coordinate point in the vertical direction of the two-dimensional plane coordinate system. This represents the x-coordinate value of a point in a two-dimensional plane. This represents the ordinate value of a point in a two-dimensional plane. The current irradiance distribution function fitting module, connected to the wind vector distribution function fitting module, is used to fit the discrete, uniformly distributed radiometer measurements of the area to be predicted into a continuous current irradiance distribution function covering the two-dimensional plane coordinate system using a spatial interpolation algorithm. .

8. The distributed photovoltaic power generation intelligent prediction system as described in claim 7, characterized in that, The power prediction module includes: The module for calculating the predicted irradiance distribution function is used to calculate the predicted irradiance distribution function based on the wind vector distribution function of the predicted area and the current irradiance distribution function. ; The irradiance power linear correlation function fitting module is connected to the predicted irradiance distribution function calculation module to obtain the first distributed photovoltaic power to the second... N The actual output power factor of distributed photovoltaic (PV) power; from the first distributed PV to the second N In distributed photovoltaic (PV) systems, PV units whose actual output power factor falls between a pre-set lower and upper power factor limit are selected and marked as constant power factor operating PV units. The actual active power corresponding to the constant power factor operating PV units is obtained and denoted as the first active power fitting sequence. The two-dimensional plane coordinates corresponding to the installation location of the constant power factor operating PV units are obtained and substituted into the current irradiance distribution function. The current irradiance data corresponding to the installation geographical location of the photovoltaic system operating with constant power factor is calculated and denoted as the first current irradiance fitting sequence. The first current irradiance fitting sequence is used as the independent variable and the first active power fitting sequence is used as the dependent variable. The least squares fitting method is used to fit the linear correlation function of irradiance power. The irradiance power nonlinear correlation function fitting module, connected to the irradiance power linear correlation function fitting module, is used to fit the first distributed photovoltaic power to the second... N In distributed photovoltaic (PV) systems, those whose actual output power factor falls outside the preset lower to upper power factor limits are selected and marked as variable power factor (VPF) PV systems. The actual active power corresponding to the VPF PV system is obtained and denoted as the second active power fitting sequence. The two-dimensional plane coordinates corresponding to the installation location of the VPF PV system are obtained and substituted into the current irradiance distribution function. The current irradiance data corresponding to the installation geographical location of the photovoltaic system operating with variable power factor is calculated and denoted as the second current irradiance fitting sequence. The second current irradiance fitting sequence is used as the independent variable and the second active power fitting sequence is used as the dependent variable. The nonlinear correlation function of irradiance power is obtained by fitting the polynomial fitting method. The initial predicted active power calculation module is connected to the irradiance power nonlinear correlation function fitting module, and is used to convert the first distributed photovoltaic power to the second... N Substituting the two-dimensional plane coordinates corresponding to the installation location of distributed photovoltaics into the predicted irradiance distribution function yields the first predicted irradiance to the second. N Predicted irradiance; respectively, the first predicted irradiance to the second... N The predicted irradiance is obtained by inputting the irradiance power linear correlation function and outputting the first initial predicted active power up to the second. N Initial predicted active power; The iterative calculation module, connected to the initial predicted active power calculation module, is used to establish a power grid simulation calculation model for the area to be predicted, and to calculate the first initial predicted active power to the second... N The initial predicted active power is substituted into the power grid simulation model of the area to be predicted for simulation calculation. Based on the simulated values ​​of the effective values ​​of the grid node bus voltage, the first to the second distributed photovoltaic power generation projects are selected. N In distributed photovoltaic systems, the predicted variable power factor (VPF) photovoltaic system is calculated based on a nonlinear correlation function of irradiance power to obtain the predicted active power of the VPF system, and the predicted active power is calculated from the first initial predicted active power to the second... N In the initial predicted active power, the predicted active power using the variable power factor is used to replace the initial predicted active power, and the final active power from the first distributed photovoltaic power generation to the second is obtained through iterative calculation. N Predicted output active power of distributed photovoltaic power.

9. The distributed photovoltaic power generation intelligent prediction system as described in claim 8, characterized in that, The module for calculating the predicted irradiance distribution function includes: The segmented calculation module is used to calculate the predicted irradiance distribution function in segments based on the wind vector distribution function of the predicted area and the current irradiance distribution function. The method for calculating the predicted irradiance distribution function is as follows: according to x , y The value of divides the two-dimensional plane coordinate system into two regions, denoted as the first region and the second region, respectively; the first region is... and ;in , These represent the minimum and maximum x-coordinates of the two-dimensional plane coordinate points within the area to be predicted, obtained in advance. , These represent the minimum and maximum ordinate values ​​of the two-dimensional plane coordinate points within the area to be predicted, obtained in advance. This represents a pre-set time interval, i.e., the difference between the predicted time point and the current time point; The cloud movement ratio is a pre-set cloud movement ratio obtained through historical data statistics; the cloud movement ratio reflects the ratio of the movement speed of cumulus clouds to the wind speed when wind drives the movement of cumulus clouds; the second area is the area remaining after deducting the first area from the area to be predicted. In the first area, The calculation formula is: ； In the second area, The value of is simplified to be the same as equal.

10. The distributed photovoltaic power generation intelligent prediction system as described in claim 9, characterized in that, The iterative calculation module includes: The initial simulation module is used to acquire the power grid topology data and line parameters of the area to be predicted, and to establish a power grid simulation calculation model for the area to be predicted. In the power grid simulation calculation model of the area to be predicted, the first distributed photovoltaic system to the second... N The active power of distributed photovoltaic systems is set as the first initial predicted active power to the second... N Initial predicted active power; [Transferring the first distributed photovoltaic power to the second...] N The power factor of distributed photovoltaic (PV) systems is set as the average of the lower and upper power factor limits. Power flow simulation methods are used to calculate the power factors from the first distributed PV system to the [number missing]th [system / system / etc.]. N Simulated values ​​of the effective value of the bus voltage of the grid node connected to the distributed photovoltaic system; The iterative simulation module, connected to the initial simulation module, is used to simulate the first distributed photovoltaic system to the second... N In distributed photovoltaic (PV) systems, those PV systems whose simulated effective values ​​of the bus voltage at connected grid nodes are not within the preset lower to upper voltage limits are marked as PV systems operating under predicted variable power factor (VPF) conditions. N The predicted irradiance corresponding to photovoltaic systems operating at the predicted variable power factor (VPF) is selected from the predicted irradiance and labeled as VPF predicted irradiance. These VPF predicted irradiances are then input into a nonlinear correlation function for irradiance power, and the output is the predicted active power at the VPF. From the first initial predicted active power to the... N In the initial predicted active power, the initial predicted active power corresponding to the photovoltaic system operating with the predicted variable power factor is replaced with the active power predicted with the variable power factor; the replaced active power is used as the active power of the current cycle, and the power factor of the photovoltaic system operating with the predicted variable power factor is set to a preset lag power factor value for voltage regulation to obtain the distributed photovoltaic power condition of the current cycle; the iterative steps are repeated until the iteration termination condition is met. The iterative steps are as follows: re-execute the power flow simulation calculation to obtain the results from the first distributed photovoltaic power generation to the [missing information]. N Simulated values ​​of the effective values ​​of the bus voltage of the grid nodes connected to the distributed photovoltaic system; re-selected photovoltaic systems operating with predicted variable power factor and calculated the predicted active power of the variable power factor; updated the power conditions of the distributed photovoltaic system in the current round. The iteration termination condition is: from the first distributed photovoltaic to the... N The simulated values ​​of the effective voltage of the grid node bus connected to the distributed photovoltaic system are all between the lower voltage limit and the upper voltage limit, or the number of iterations reaches the preset maximum number of iterations.

Images