JP2015192586A - Demand and supply control system and demand and supply control method for micro grid - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a demand and supply control system for micro grid capable of appropriately performing demand and supply control by easily obtaining a short term output prediction value having high accuracy of a photovoltaic power generation facility.SOLUTION: A demand and supply control system for micro grid comprises: illuminance sensors 10, 10n that are installed on a prediction target photovoltaic power generation facility PV-0 and surrounding photovoltaic power generation facilities PV-n disposed around the facility PV-0 in an electric power system, respectively and measure output of the respective photovoltaic power generation facilities; a photovoltaic power generation output prediction value calculation unit 20 that individually performs Fourier transformation on output of the PV-0 and output of the PV-n measured before predetermined time by the illuminance sensors 10, 10n, calculates the degree of correlation between the output of the PV-0 and output of the PV-n on the basis of phase difference between the output of the PV-0 and output of the PV-n after the Fourier transformation for each frequency by using a phase only correlation method, and determines the present output of the PV-n having the maximum degree of correlation with the output of the PV-0 to be an output prediction value of the PV-0; and a control unit 30 for performing demand and supply control of a micro grid using the determined output prediction value as a parameter for a control system.

Description

  The present invention relates to a supply and demand control system and method for a microgrid (small-scale power system) capable of supplying power and heat to a load in a region independently of a power system by a plurality of distributed power sources.

  In a microgrid, which is a small-scale energy network that satisfies local demand and supply, without depending on a large-scale power system, precise supply and demand control is performed to maintain the system. It is usually carried out by the following procedure.

  (1) Based on the weather forecast, the next day's 30-minute load forecast / photovoltaic power forecast is calculated by a statistical method.

  (2) Assign attributes (reference voltage, reference frequency, tracking fluctuation range, upper and lower limit values, tracking time, dead band, etc.) that follow fluctuations to each distributed power source (hereinafter also referred to as DER).

  (3) Based on the conditions (1) and (2), an optimal operation plan is created by a mathematical programming method, and an operation command is given to the DER.

  (4) When the DER has sufficient power as a base power source and adjustment power, automatic operation may be performed according to the following rules without the procedures of (1) and (3).

-Base power supply operates at a constant voltage and output.-Other power supplies follow the output within the set value range based on the voltage of the base power supply.

  (5) In general, sudden fluctuations are absorbed by a storage battery or a capacitor.

  The photovoltaic power generation prediction by the statistical method (1) has been performed as shown in FIG. 11, for example.

  FIG. 11 shows how the clouds move by the wind, the boundary between the clouds overlaps the sun, and covers the sun, and the amount of power generation output of the photovoltaic power generation (hereinafter also referred to as PV) equipment at that time. It shows the transition of.

  The photovoltaic power generation output greatly changes during a period from time t1 when the boundary of the cloud overlaps the sun (position P1) to time t2 when the cloud completely covers the sun (position P2).

  As a statistical analysis procedure, first, photovoltaic power generation output data between the times t1 and t2 is acquired (including trend data before a downward change). Next, the falling width and time are collected, and general statistical processing (average and deviation) is performed. The trend data before the descent is analyzed including the weather, temperature, wind conditions, and time (regression analysis and spectrum analysis). Then, pattern matching logic is created and the logic is verified using history data.

  In addition, the method of patent document 1 is proposed, for example as a method of estimating the output of a solar power generation facility, without using a weather forecast.

JP 2013-208050 A

  In the photovoltaic power generation output prediction method using the statistical analysis shown in FIG. 11, it is very difficult to predict the descending timing (time t1, t2) of the photovoltaic power generation output. A combination is necessary.

  In addition, when making predictions based on weather forecasts, only predictions can be made on a daily basis, and predictions can be made in the long term, but predictions in the short term are very difficult.

  The method of Patent Document 1 can evaluate the point that pattern matching is simplified and weather information is independent, but the purpose of pattern matching is only statistical prediction using history data. For this reason, it is difficult to predict the future output from the past output history data of the photovoltaic power generation facility to be predicted, and knowing the timing (time t1, t2) between the clouds as shown in FIG. Almost impossible.

  Therefore, it cannot be used for supply / demand control and fluctuation suppression control of a microgrid that is required to appropriately respond to fluctuations in the amount of output power.

  Here, the state of conventional supply and demand adjustment control in the microgrid will be described with reference to FIG. FIG. 12 shows the transition of the output of the photovoltaic power generation (solid line) that fluctuated in the same manner as FIG. 11 due to the cloud covering the sun, and the transition of the output of the distributed power source (DER) that is operated following the fluctuation (broken line). Is shown.

In FIG.
(1) Conventionally, since the short-term predicted value of the output of the photovoltaic power generation facility is not used for supply and demand control, when the DER detects a sudden change at time t1, it performs an operation that follows the change (variation mitigation operation). . For example, the reciprocating engine generator performs governor-free operation, and the storage battery performs absorber operation.

  (2) If the fluctuation range / gradient gradient of the output of the solar power generation facility exceeds the limit, the inertia of the reciprocating engine generator is over, and the charge / discharge capacity is over in the storage battery. When the load exceeds the output adjustment range of all operating DERs, the system stops at time tx (it took time to start DER during stopping and preparation for stopping DER during starting, and could not keep up) ).

  (3) Connection operation was difficult in this way.

  FIG. 13 shows a configuration example of a conventional cooperative control system for distributed power sources in a microgrid. The control system of FIG. 13 does not use the short-term predicted value of the output of the photovoltaic power generation facility. On the microgrid system side (the control unit), the history data of the past several years of the photovoltaic power generation output on the day before the control is performed. The load / PV predicted value is obtained based on the above, and a 24-hour load / PV curve is obtained from the predicted value. Then, an optimal operation plan is calculated based on the 24-hour load / PV curve, and 24-hour distributed power operation data (operation schedule) is created (daily control).

  On the distributed power controller side, on the day of control, the 24-hour distributed power operation data (schedule) is received, and each distributed power supply is individually managed according to the optimum operation plan while monitoring its own voltage, current, and frequency. Operate optimally.

  Each distributed power controller has a simple slope prediction function, and the gradient of change of active power, voltage, frequency, reactive power, etc. at each end of each distributed power source is maintained for a while. It is assumed (predicted) that the soldering is performed, and output control (for example, charging / discharging of the storage battery, etc.) based on the assumption is performed.

  When the predicted value obtained on the previous day deviates from the threshold value, a subsequent load / PV predicted value is obtained on the microgrid system side, a 24-hour load / PV curve is obtained from the predicted value, and the optimum based on the curve is obtained. The operation plan is recalculated, and 24-hour distributed power operation data (operation schedule) is created again (time control).

  The distributed power controller side receives the re-created 24-hour distributed power operation data (schedule) and individually monitors each distributed power source according to the optimum operation plan while monitoring its own voltage, current, and frequency. Operate optimally.

  Each distributed power source includes PV1, monogeneration 2 (power generation only), storage battery 3, cogeneration 4, heat storage device 5, power load 6, heat load 7, and the like.

  In the control system shown in FIG. 13, the dependence of the storage battery 3 used for system stability is high, and it is difficult to meet the demand for load fluctuations and peak shifts exceeding the capacity of the storage battery 3.

  The present invention solves the above-mentioned problems, and the object thereof is to easily obtain a short-term predicted output value of a highly accurate photovoltaic power generation facility and to appropriately cope with fluctuations in the output of the photovoltaic power generation facility. It is an object of the present invention to provide a microgrid supply and demand control system and method capable of performing accurate supply and demand control.

  The supply and demand control system for a microgrid according to claim 1 for solving the above-described problem is a supply and demand control system for a microgrid provided with at least a plurality of photovoltaic power generation facilities, and includes a photovoltaic power generation output in an electric power system. Solar power that is provided for each of the solar power generation equipment, and the solar power generation equipment that is provided in each of the solar power generation equipment and the surrounding solar power generation equipment disposed around the solar power generation equipment to be predicted The output of the power generation output measuring means and the output of the prediction target solar power generation equipment and the output of the surrounding solar power generation equipment measured by the solar power generation output measurement means a predetermined time before are respectively Fourier-transformed, and the prediction target sun after the Fourier transform Based on the phase difference between the output of the photovoltaic power generation facility and the output of the surrounding photovoltaic power generation facility for each frequency, the output of the target photovoltaic power generation facility and the output of the surrounding photovoltaic power generation facility are output by the phase-only correlation method. The predicted output of photovoltaic power generation is determined by determining the current output of the surrounding photovoltaic power generation facility that has the highest correlation with the output of the predicted photovoltaic power generation facility as the predicted output value of the predicted photovoltaic power generation facility. It is characterized by comprising: a calculating means; and a control means for performing supply and demand control of the microgrid using the output predicted value determined by the photovoltaic power generation output predicted value calculating means as a control system parameter.

  The microgrid supply and demand control system according to claim 2 is the microgrid supply and demand control system according to claim 1, wherein a plurality of the surrounding solar power generation facilities are arranged, and the solar power generation output predicted value calculation means is the solar power generation output Obtain the photovoltaic power output data for the past set time of the prediction target photovoltaic power generation facility and the surrounding photovoltaic power generation facility measured by the measuring means, perform Fourier transform on the obtained photovoltaic power generation output data, and after Fourier transform Based on the phase difference of each frequency of the photovoltaic power output data of the solar power generation output data, the similarity to the output data of the prediction target solar power generation facility is calculated for the output data of all the surrounding solar power generation facilities, the calculated From the photovoltaic power output measurement means, the current photovoltaic power output data of the surrounding photovoltaic power generation facility having the highest degree of similarity is used as the predicted output value of the target photovoltaic power generation facility. It is characterized in that Tokusuru.

  The microgrid supply and demand control method according to claim 7 is a microgrid supply and demand control method including at least a plurality of solar power generation facilities, and is a prediction target of solar power generation output in an electric power system. The photovoltaic power generation output measuring means provided in each of the prediction target photovoltaic power generation facilities and a plurality of surrounding solar power generation facilities arranged around the prediction target solar power generation facilities are output from each solar power generation facility. A solar power generation output measuring step for measuring the solar power generation output predicted value calculation means, and the prediction target solar power generation equipment and the surrounding solar power generation equipment for the past set time measured by the solar power generation output measurement means The solar power output data acquisition step of acquiring the solar power output data, and the solar power output predicted value calculation means, Fourier transform the acquired solar power output data, Similarity to calculate the similarity with the output data of the prediction target photovoltaic power generation facility for the output data of all the surrounding photovoltaic power generation facilities based on the phase difference for each frequency of the photovoltaic power generation output data after the lier conversion The calculation step and the photovoltaic power generation output predicted value calculation means acquire the current photovoltaic power generation output data of the surrounding photovoltaic power generation facility having the highest calculated similarity from the photovoltaic power generation output measurement means, A predicted value determining step for determining the predicted data as an output predicted value of the photovoltaic power generation facility to be predicted, and the control means using the output predicted value determined by the photovoltaic power generation output predicted value calculating means as a control system parameter And a control step for performing supply and demand control.

  In the above configuration, the photovoltaic power generation output predicted value calculation means of the present invention is configured so that the output of the surrounding photovoltaic power generation equipment having the highest correlation with the output data of the prediction target photovoltaic power generation equipment in a predetermined time is the current time zone. Is also estimated to be the same as the output of the solar photovoltaic power generation facility to be predicted, so this is determined as the predicted output value of the solar power generation facility to be predicted. As a result, a highly accurate output predicted value can be easily obtained.

  In the present invention, the output prediction value predicted based on the weather forecast and the historical data of the photovoltaic power generation output over the long term are not used as in the prior art, and the output data of the surrounding photovoltaic power generation facilities for the set time and the prediction target Since the output prediction value is determined based on the correlation degree of the output data of the photovoltaic power generation equipment, even if the output of the solar power generation equipment fluctuates rapidly due to sudden weather changes, a highly accurate output prediction value is obtained. Can be obtained.

  Moreover, since the control means performs supply and demand control of the microgrid based on the highly accurate predicted output value, it is possible to perform a preparation operation or a fluctuation mitigation operation corresponding to the output fluctuation.

  The supply and demand control system for a microgrid according to claim 3 is the microgrid supply and demand control system according to claim 1 or 2, wherein the control means is a millisecond to a few seconds of an output predicted value determined by the photovoltaic power generation output predicted value calculating means. The output fluctuation slope is predicted based on the data, and the start / stop of the supply and demand adjustment control is determined according to the predicted slope of the output fluctuation.

  The microgrid supply and demand control method according to claim 8 is the microgrid supply and demand control method according to claim 7, wherein the control step outputs data based on data of millisecond to several seconds of the predicted output value determined by the predicted value determination step. The present invention is characterized in that a fluctuation slope is predicted, and start / stop of supply and demand adjustment control is determined according to the predicted slope of output fluctuation.

  According to the above configuration, the slope of the output fluctuation can be predicted from the data of the output predicted value in milliseconds to several seconds, and when the slope is predicted to be large, the load is cut off or distributed. The mold power supply can be turned on / off.

  According to a fourth aspect of the present invention, there is provided a supply and demand control system for a microgrid according to any one of the first to third aspects, wherein the control unit is configured to output an output predicted value determined by the photovoltaic power generation output predicted value calculating unit. The output fluctuation range is predicted based on data of several seconds to several minutes, and a supply and demand adjustment resource is selected according to the predicted output fluctuation range.

  The microgrid supply and demand control method according to claim 9 is the method according to claim 7 or 8, wherein the output fluctuation range is determined based on data of several seconds to several minutes of the predicted output value determined by the predicted value determination step. It is characterized in that a supply and demand adjustment resource is selected according to the predicted output fluctuation range.

  According to the above configuration, when the output fluctuation range can be predicted from data of the output predicted value of several seconds to several minutes, and the output fluctuation range is predicted to be large, switching of the interrupting load or The number of distributed power sources can be controlled.

  The supply and demand control system for a microgrid according to claim 5 is characterized in that, in any one of claims 1 to 4, the control means includes an output predicted value determined by the photovoltaic power generation output predicted value calculation means. In addition, an optimal operation plan for each distributed power source in the microgrid is created by a mathematical programming method based on the attribute that follows the output fluctuation, and a command for operation and stop is given to each distributed power source.

  The supply and demand control method for a microgrid according to claim 10 is the method according to any one of claims 7 to 9, wherein the control step includes an output predicted value determined by the predicted value determination step and an output fluctuation. It is characterized in that an optimal operation plan for each distributed power source in the microgrid is created by a mathematical programming method based on the following attributes, and an operation and stop command is given to each distributed power source.

  According to the said structure, since an operation plan can be performed in advance using the output predicted value of a photovoltaic power generation facility, optimal calculation time can be ensured and highly efficient optimal control can be implemented. In addition, it is possible to create an optimal operation plan for each distributed power source based on highly accurate predicted output data of photovoltaic power generation facilities, and each distributed power source can operate individually according to its role according to the operation plan. Can be executed.

  According to a sixth aspect of the present invention, there is provided the supply and demand control system for a microgrid according to the fifth aspect, wherein the optimum operation plan excludes a time restriction condition and a distributed power source on / off restriction condition divided every predetermined time period. It is characterized in that it is created by performing an optimization calculation by linear programming every predetermined time period using the optimized constraint condition equation and the objective function.

  The supply and demand control method for a microgrid according to claim 11 is the method according to claim 10, wherein the optimum operation plan excludes a time restriction condition and a distributed power source on / off restriction condition divided for each predetermined time period. It is characterized in that it is created by performing an optimization calculation by linear programming every predetermined time period using the optimized constraint condition equation and the objective function.

  According to the above configuration, the calculation is performed excluding the time constraint condition and the distributed power on / off constraint condition, which are integer elements in the optimization calculation, so that high-speed calculation is possible.

(1) According to the inventions described in claims 1 to 11, it is possible to easily obtain a short-term predicted output value of a highly accurate photovoltaic power generation facility, and to appropriately cope with fluctuations in the output of the photovoltaic power generation facility. Supply and demand control can be performed.

  In addition, it is possible to sufficiently cover a band of 1 / hundreds of a hundredths (1 / hundreds of Hz) to several Hz where the output fluctuation of the photovoltaic power generation facilities is severe.

  In addition, optimization of efficiency by predictive control and improvement of robustness can be achieved. Moreover, it is possible to incorporate optimal control of a plurality of distributed power sources, load control, control of photovoltaic power generation facilities, and the like.

In addition, it is possible to predict a rapid output fluctuation of the photovoltaic power generation facility, thereby enabling effective schedule control of a storage battery, a capacitor, and other distributed power sources. Also, the capacity for adjustment can be reduced.
(2) Moreover, according to invention of Claim 3, 4, 8, and 9, reserve power reserve and preparation are attained by knowing beforehand the fluctuation inclination and fluctuation width of photovoltaic power generation equipment.
(3) Further, according to the inventions according to claims 5 and 10, since the operation plan can be made in advance using the output predicted value of the photovoltaic power generation facility, the optimum calculation time can be secured, and highly efficient optimum control is achieved. It becomes possible to carry out.
(4) According to the inventions of claims 6 and 11, since the calculation is performed excluding the time constraint condition and the distributed power source on / off constraint condition, which are integer elements in the optimization calculation. High-speed calculation is possible.

The block diagram of the embodiment of this invention. The method of obtaining the photovoltaic power generation output predicted value according to the embodiment of the present invention is represented, (a) is a layout diagram of the prediction target PV and the surrounding PV, (b) is an output waveform diagram of the surrounding PV, and (c) is a prediction target. Output waveform diagram of PV, (d) is a frequency characteristic diagram after Fourier transform, (e) is a correlation diagram of frequency and phase difference. The PV output waveform diagram explaining the time window used by the means to calculate the time lag in the embodiment of the present invention. The PV output waveform diagram explaining operation | movement of the means to calculate the time shift | offset | difference in the embodiment of this invention. The characteristic view which shows transition of the output of the distributed power supply at the time of supply-and-demand adjustment control using the output predicted value of this invention. The control model using the output prediction value of this invention is represented, (a) is a waveform diagram which shows the output data of millisecond-several seconds, (b) is explanatory drawing which shows the control at the time of estimating the inclination of an output fluctuation, (C) is a waveform diagram showing output data for several seconds to several minutes, and (d) is an explanatory diagram showing control when the width of output fluctuation is predicted. Explanatory drawing which shows the mode of the supply-and-demand control of the microgrid by the example of embodiment of this invention. 1 is a configuration diagram of a cooperative control system for distributed power sources using an embodiment of the present invention. The PV output waveform figure which shows an example of PV short-term prediction evaluation by the example embodiment of this invention. Explanatory drawing showing the mode of the optimal calculation according to time at the time of the optimal driving | operation plan preparation by the example of embodiment of this invention. Explanatory drawing which shows an example of the output prediction method of the conventional solar power generation facility. The characteristic view which shows transition of the output of the distributed power supply at the time of supply-and-demand adjustment control by the conventional method. The block diagram of the conventional cooperative control system of a distributed power supply.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings, but the present invention is not limited to the following embodiments. FIG. 1 shows a configuration of a microgrid for supplying power to a load in an electric power system by a plurality of distributed power sources in an embodiment of the present invention, PV is a solar power generation facility provided at a plurality of locations, and 10 is a solar power plant. Illuminance sensors (photovoltaic power output measuring means of the present invention) each provided in the photovoltaic power generation facility PV or disposed at substantially the same position as the installation position of the photovoltaic power generation facility PV, 3 indicates a power storage device ing.

  The illuminance sensor 10 measures the solar irradiance of the photovoltaic power generation facility PV. In this embodiment, the illuminance sensor 10 handles an electrical signal photoelectrically converted by a photodiode or phototransistor as output data of the photovoltaic power generation facility PV. Therefore, in the following description, the output of the illuminance sensor 10 may be referred to as PV output. Note that the photovoltaic power generation output measuring means is not limited to the illuminance sensor 10 but may be a power detector that detects the power generation amount of the photovoltaic power generation facility PV itself.

  Among the plurality of solar power generation facilities PV, PV-0 represents one prediction target solar power generation facility for which power generation output is to be predicted (predicted), and PV-n represents the prediction target solar power generation facility PV-0. A plurality of surrounding solar power generation facilities provided around the.

  Therefore, 10 is an illuminance sensor provided in the prediction target solar power generation facility PV-0, and 10n is an illuminance sensor provided in a plurality of surrounding solar power generation facilities PV-n.

  20 captures each PV output of the illuminance sensors 10 and 10n in, for example, a unit time period, Fourier-transforms the PV outputs, and based on the phase difference for each frequency between the PV-0 output and the PV-n output after the Fourier transform. The correlation between the PV-0 output and the PV-n output is obtained by the phase-only correlation method, and the current output data of the surrounding solar power generation facility PV-n having the highest correlation with the PV power output PV-0 output to be predicted is obtained. And a photovoltaic power generation output predicted value calculation unit (solar power generation output predicted value calculation means) that is determined as an output predicted value of the prediction target solar power generation facility PV-0.

  The photovoltaic power generation output predicted value calculation unit 20 takes in the PV output data of the illuminance sensors 10 and 10n, stores the PV output data, stores the PV output data by Fourier transform processing, and calculates the correlation degree by the phase-only correlation method. The function to be performed and the surrounding photovoltaic power generation facility PV-n having the highest correlation with the target photovoltaic power generation facility PV-0 are identified, and the current PV output data of the identified PV-n is acquired from the illuminance sensor 10n. It has the function to do.

  Reference numeral 30 denotes a control unit (control means) that performs supply and demand control of the microgrid using the output predicted value determined by the photovoltaic power generation output predicted value calculation unit 20 as a control system parameter.

  The control unit 30 predicts the slope of the output fluctuation based on at least the data of the output predicted value determined by the photovoltaic power generation output predicted value calculation unit 20 in milliseconds to several seconds, and the predicted output fluctuation. A function for determining the start / stop of supply and demand adjustment control according to the slope and the output fluctuation range are predicted based on the data for several seconds to several minutes of the predicted output value, and the supply and demand is determined according to the predicted output fluctuation range. Based on the function to select the adjustment resource, the predicted output value, and the attribute that follows the output fluctuation, an optimal operation plan for each distributed power source in the microgrid is created by a mathematical programming method, and each distributed power source is operated. And a function of giving a stop command.

  The photovoltaic power generation output predicted value calculation unit 20 and the control unit 30 are configured by, for example, a computer, and hardware resources of a normal computer such as a ROM, a RAM, a CPU, an input device, an output device, a communication interface, a hard disk, a recording medium, and the like The drive device is provided.

  As a result of the cooperation between the hardware resource and the software resource (OS, application, etc.), the photovoltaic power generation output predicted value calculation unit 20 and the control unit 30 implement each processing function described above.

  Next, an example of the operation of the photovoltaic power generation output predicted value calculation unit 20 will be described with reference to FIG. Here, as shown to Fig.2 (a), one prediction object photovoltaic power generation equipment PV-0 and the some surrounding photovoltaic power generation equipment PV-1 to PV-7 shall be arrange | positioned around it. .

<Step S1>
The photovoltaic power generation output predicted value calculation unit 20 first outputs the PV output of the surrounding photovoltaic power generation facilities PV-1 to PV-7 for a set time, for example, several minutes to several tens of minutes before the own memory. Data (output waveform) is acquired (FIG. 2B).

<Step S2>
After the process of step S1 or simultaneously with the process of step S1, PV output data (output waveform) of the prediction target photovoltaic power generation facility PV-0 for the set time corresponding to the time after the acquisition time of step S1 from the memory. Is acquired (FIG. 2C).

<Step S3>
Each output waveform of the prediction target solar power generation facility PV-0 and the surrounding solar power generation facilities PV-1 to PV-7 is Fourier transformed to obtain frequency characteristics (FIG. 2D).

<Step S4>
A phase difference Δt of frequency characteristics after Fourier transform of one PV of the surrounding solar power generation facilities PV1 to PV7 and the prediction target solar power generation facility PV-0 is obtained for different frequencies, for example, ω1 to ω3 (Δt1 to Δt3). ), Based on the phase differences Δt1 to Δt3, the degree of similarity between them is calculated by the phase-only correlation method (the phase differences Δt1 to Δt3 are linear with respect to the frequencies (ω1 to ω3) as shown in FIG. 2 (e)). That is, the similarity to a straight line is large).

<Step S5>
The process of step S4 is performed about all the surrounding photovoltaic power generation equipment PV1-PV7.

<Step S6>
As a result of the similarity calculation in steps S4 and S5, an ambient photovoltaic power generation facility (any PV-n among PV1 to PV7) having the maximum similarity is determined, and the determined ambient solar power generation facility PV- The current PV output data (output waveform) of n is acquired from the illuminance sensor 10n as the predicted output value of the prediction target photovoltaic power generation facility PV-0.

  Here, examples of actual output data (output waveform) for each PV are shown in FIG. 3 and FIG. 4, and the surrounding photovoltaic power generation facility (PV-n) having the highest phase correlation (maximum similarity) is determined. An example of means for calculating the time lag until the current PV output data of the determined PV-n is acquired as the PV-0 output predicted value will be described.

  3 and 4 show the output waveforms (measurement results) of all PV (PV-0 to PV-7) in FIG. 2A, respectively. At the time of data acquisition in step S1, the surrounding solar power generation equipment A time window WINn shifted by a fixed time (a constant interval) is set for each of the measurement results of PV-1 to PV-7.

Then, the phase limited comparison is performed between the time window WIN ₀ set for the measurement result (output waveform) of the prediction target photovoltaic power generation facility PV-0 acquired in step S2 and all the time windows WINn (see above). Steps S3 to S5 are executed).

Next, in step S6, as a result of the phase-only comparison, a time window having the highest correlation, for example, WIN _3x of PV- ₃ and WIN _{0x + T} of PV-0 is extracted, and times t8 and t11 at which the waveform becomes maximum. Pay attention to. Then, the difference between the times t11 and t8 at the time of each maximum value is calculated and set as a difference ΔT between two predicted times (however, the minimum time is used when the maximum time is at the bottom of the time window). . And let the waveform after (DELTA) T of the surrounding solar power generation facility PV-3 be a prediction waveform (output prediction value) of the prediction object solar power generation facility PV-0.

  Note that, for example, the technique described in Patent Document 1 is used for pattern matching processing of the Fourier transform and the phase-only correlation method in steps S1 to S6.

  The control unit 30 performs supply and demand control of the microgrid using the PV output predicted value (output waveform) predicted and determined by the photovoltaic power generation output predicted value calculation unit 20 as a control system parameter. An example of the control will be described with reference to FIG.

  FIG. 5 shows the transition (solid line) of the output of the photovoltaic power generation that fluctuates in the same manner as FIG. 11 as the clouds cover the sun, and the output of the distributed power source (DER) operated by the control unit 30 of the present embodiment. The transition (broken line) is shown.

  (1) When a sudden change is predicted in advance at time t1, DER starts a preparatory exercise corresponding to the change. For example, the reciprocating engine generator performs a preceding operation in the direction of change, and the storage battery performs a charge / discharge operation (charging in the example of FIG. 5) in the direction of change.

  (2) Since the fluctuation gradient and fluctuation width of the output can be known based on the PV output predicted value (output waveform) as shown in FIG. 6 to be described later, an operation for reducing the fluctuation is performed within the prediction time frame. For example, reciprocating engine generators operate within inertia, and storage batteries maintain charge / discharge capacity.

  (3) The connection operation is thus facilitated.

  Next, an example of how to predict the fluctuation slope and fluctuation width of the PV output will be described with reference to FIG. FIG. 6 shows some of the functions processed by the control unit 30, and (a) and (b) are PV output predicted values obtained by the photovoltaic power generation output predicted value calculation unit 20 in milliseconds to several seconds. (C), (d) shows the PV output predicted value obtained by the photovoltaic power generation output predicted value calculation unit 20 for several seconds to several minutes of data. The case where the range of output fluctuation is predicted based on this is shown.

  From the data of milliseconds to several seconds in FIG. 6 (a), the slope of fluctuation as shown by the thick solid line in FIG. 6 (b) is predicted, and start / stop of supply and demand adjustment control is determined according to the predicted slope. , Load cutoff, DER control.

  Note that the five broken lines in FIG. 6B show an example of the possibility of the inclination of the output fluctuation when the inclination prediction as in the present invention is not used. I don't know if the trend will change, and I can't judge whether to start or stop the supply and demand adjustment control.

  The phase correlation coefficient obtained from the prediction results obtained by the processes of FIGS. 6A and 6B can be used as the probability of prediction. If the scale of change (inclination size and duration) can be predicted, the following DER / load control is possible.

  (1) Start operation to ease the inclination in advance.

  Adjust the output according to the accuracy of the prediction (definitely increase the relaxation output).

(2) When it is judged that the scale of fluctuation will be larger than the fluctuation mitigation reserve of the operating DER → Adjust by starting the stopped DER or stopping the operating DER → If it is judged that adjustment is difficult by DER Adjust by load shut-off (Since these controls with starting and stopping take time, the time margin given by prediction as in the present invention is important)
Adjust the reaction time according to the accuracy of the prediction (the more sensitive you are, the more sensitive you are).

(3) When it is difficult to adjust with a single DER fluctuation scale → Mitigating fluctuations with multiple DERs → Schedule the role time according to the DER reaction performance and provide a dead zone (Scheduling takes time) The margin of time given by prediction as in the present invention is important).

  An output fluctuation range as shown by a solid line in FIG. 6D is predicted from the data of several seconds to several minutes in FIG. 6C, and a supply and demand adjustment resource is selected and executed according to the predicted output fluctuation range. .

  That is, load cutoff / der control preparation, load cutoff, DER control, cutoff load switching, DER switching (number control), and the like are performed.

  By predicting the output fluctuation range as shown in FIGS. 6C and 6D and controlling it accordingly, the optimum supply and demand adjustment resource is selected and executed, and the load control schedule can be created and executed.

  That is, the phase correlation coefficient obtained from the prediction results obtained by the processes of FIGS. 6C and 6D can be used as the probability of prediction. If the scale of change (magnitude and duration) can be predicted, the following DER / load control is possible.

  (1) Start the operation to reduce the amplitude in advance.

  Adjust the output according to the accuracy of the prediction (definitely increase the relaxation output).

(2) When it is judged that the scale of fluctuation is larger than the reserve capacity for DER during operation → Adjustment is made by starting DER while stopped or when DER is stopped during operation → Loading when it is difficult to adjust by DER Adjust by shut-off (Since these controls with starting and stopping take time, the margin of time given by prediction as in the present invention is important)
Adjust the reaction time according to the accuracy of the prediction (the more sensitive you are, the more sensitive you are).

  (3) A plurality of DER optimum schedules are created for changes in time windows (output fluctuations).

  Here, an example of a microgrid system that realizes high-speed optimal control (assuming a control cycle of several seconds to several minutes) by applying the present invention will be described with reference to FIG. In FIG. 7, a PV 51, a storage battery 52, and a reciprocating engine generator 53 are provided as distributed power sources, and an important load 54, a load 55 that can be cut off, and a load 56 that cannot be cut off are provided as loads.

  61 is the function of the photovoltaic power generation output predicted value calculation unit 20 in FIG. 1 for obtaining the predicted output value of the solar power generation facility, and the control unit 30 described in FIG. 6 predicts the fluctuation gradient and fluctuation width of the PV output. This is a short-term prediction unit having a function and a function of predicting a load amount.

  Reference numeral 62 denotes a high-speed measurement control unit that performs energy optimal distribution calculation based on the data acquired from each of the distributed power sources 51 to 53, the loads 54 to 56, and the short-term prediction unit 61, and performs centralized control of each distributed power source. .

  Next, an example of the operation performed by the high-speed measurement control unit 62 of the system of FIG. 7 will be described.

  (1) Collecting the distributed power source in operation, the energy output / usage of the load, the adjustment capacity, and the energy intensity.

  (2) The necessary total energy amount and the amount of adjustment are calculated from the short-term energy prediction obtained by the short-term prediction unit 61.

  (3) Organize requirements such as storage battery status, importance of PV output (how much load will be carried), presence / absence of load that can be cut off, and period of load cut-off.

  (4) Perform energy optimal allocation calculation (possible to implement by simple linear programming → high speed calculation possible).

  (5) Issue an adjustment command to each energy resource.

  Next, an embodiment in which the present invention is applied to a cooperative control system for distributed power sources is shown in FIGS. 8, the same parts as those in FIG. 13 are denoted by the same reference numerals.

  8 differs from FIG. 13 in that the short-term peripheral PV history data DB (database) 71, a tens of seconds PV prediction unit 72, a tens of seconds PV slope prediction unit 73, a tens of minutes PV prediction unit 74, a tens of minutes PV prediction curve creation unit 75, a determination unit 76, a short-term optimum operation plan creation unit 77 and a prediction evaluation / adjustment unit 78 are provided, and the other parts are configured in the same manner as FIG. .

  In the system of FIG. 8, the day control and the time control are performed in the same manner as in FIG. That is, on the day before the control, the microgrid system side (the control unit) obtains the load / PV predicted value based on the historical data of the past several years of the photovoltaic power generation output, and the 24-hour load / PV from the predicted value. Get a curve. Then, an optimal operation plan is calculated based on the 24-hour load / PV curve, and 24-hour distributed power operation data (operation schedule) is created (daily control).

  On the distributed power controller side, on the day of control, the 24-hour distributed power operation data (schedule) is received, and each distributed power supply is individually managed according to the optimum operation plan while monitoring its own voltage, current, and frequency. Operate optimally. Similarly to FIG. 13, output control based on simple inclination prediction is performed.

  When the predicted value obtained on the previous day deviates from the threshold value, a subsequent load / PV predicted value is obtained on the microgrid system side, a 24-hour load / PV curve is obtained from the predicted value, and the optimum based on the curve is obtained. The operation plan is recalculated, and 24-hour distributed power operation data (operation schedule) is created again (time control).

  The distributed power controller side receives the re-created 24-hour distributed power operation data (schedule) and individually monitors each distributed power source according to the optimum operation plan while monitoring its own voltage, current, and frequency. Operate optimally.

  The short-term peripheral PV history data DB 71 stores PV output data acquired from the illuminance sensor 10n on the peripheral photovoltaic power generation facility PV-n side in FIG.

  The tens of seconds PV prediction unit 72 is based on the PV output data acquired from the illuminance sensor 10 on the prediction target photovoltaic power generation facility PV-0 side in FIG. 1 and the PV output data in the short-term peripheral PV history data DB 71. The output predicted value is determined by the function of the first photovoltaic power generation output predicted value calculation unit 20 to obtain PV output predicted data (data for several tens of seconds) as shown in FIG.

  The tens of seconds PV slope prediction unit 73 predicts the slope of output fluctuation from the tens of seconds PV output prediction data obtained by the tens of seconds PV prediction unit 72 (for example, as shown by the thick solid line in FIG. 6B). To predict the slope) (min-second control).

  The tens of PV prediction unit 74 is based on the PV output data acquired from the illuminance sensor 10 on the prediction target photovoltaic power generation facility PV-0 side in FIG. 1 and the PV output data in the short-term peripheral PV history data DB 71. The output predicted value is determined by the function of the first photovoltaic power generation output predicted value calculation unit 20 to obtain PV output predicted data (data for several tens of minutes) as shown in FIG.

  The tens of minutes PV prediction curve creation unit 75 predicts the output fluctuation range from the tens of minutes PV output prediction data obtained by the tens of minutes PV prediction unit 74 and creates a prediction curve (for example, FIG. 6D). ) Create a curve like a solid line) (hour to minute control).

  The determination unit 76 checks how the PV predicted value obtained by the ten-second PV slope prediction unit 73 and the tens of minutes PV prediction curve creation unit 75 affects the current optimum driving solution. To do.

  For example, when the determination unit 76 determines that the maximum change in the statistical short-term load prediction predicted by the function of the short-term prediction unit 61 in FIG. In addition, the short-term optimal operation plan creation unit 77 performs the optimal operation distribution calculation for ensuring the surplus capacity.

  In addition, when the determination unit 76 determines that the system cannot be maintained with the currently operating device configuration with respect to the maximum increase width and the maximum decrease width value of the short-term prediction generated by the tens of minutes PV prediction curve generation unit 75, The optimum operation plan creation unit 77 executes the optimum operation distribution calculation including the stopped equipment.

  When it is necessary to start / stop the currently operating device configuration, a threshold value for starting / stopping is determined and prepared.

  In addition, when the determination unit 76 determines that there is a device that is operating inefficiently, the short-term optimal operation plan creation unit 77 performs optimal operation distribution calculation with the device configuration in operation.

  Moreover, the autonomous behavior according to a role is taken by making into the target value the calculated value of the optimal driving | operation distribution calculation calculated in the short-term optimal driving | operation plan preparation part 77 with respect to the present value of load and PV electric power generation amount.

  The prediction evaluation / adjustment unit 78 evaluates and adjusts short-term predictions such as the tens of seconds PV slope prediction unit 73 and the tens of minutes PV prediction curve creation unit 75, and is configured as follows.

  FIG. 9 shows the PV short-term predicted value (PV output predicted value) obtained by the photovoltaic power generation output predicted value calculation unit 20 and the actual PV output value of the photovoltaic power generation facility PV-0 to be predicted, and these predicted values. Based on the actual measurement values, the following prediction evaluation method is used.

・ Standard deviation ・ Mean square error Maximum deviation, minimum deviation distance (output value deviation)
The difference between the predicted value and the actual measurement value at the same time is calculated for all times. Maximum value = Maximum output value deviation (Δe−Pmax)
Minimum value = Minimum output value deviation (Δe-Pmin)
(Time deviation)
The prediction curve is superimposed on the past actual measurement curve and shifted in the future / past direction, and the time when the earliest crossing occurs. Maximum value = maximum time deviation (Δe−tmax)
Minimum value = Minimum time deviation (Δe−tmin)
Next, an example of automatic correction using the PV short-term prediction evaluation performed by the prediction evaluation / adjustment unit 78 will be described.

・ Possibility of use of PV short-term forecast value (discovery of judgment threshold)
1. The correlation between the allowable error (upper and lower thresholds of the difference between the predicted value and the actually measured value, the threshold of the gradient difference, the maximum time divergence, the maximum output value divergence) and the phase correlation coefficient and other prediction parameters is checked.

2. When there is a correlation, it is possible to check the threshold with the phase correlation coefficient, eliminating the need for extra calculations.
-Correct PV short-term forecast values if possible (output shift)
1. The prediction curve is shifted in the output direction according to the actual measurement value at the same time, and then used (time shift).
1. The prediction curve is superimposed on the past actual measurement curve and shifted in the future / past direction, and the prediction curve is shifted in the time direction by the time of the earliest crossing, and then used.

・ Use of the degree of coincidence between the predicted value and the actual measurement value (discovery of the variation point origin element; the starting point for pattern matching by the phase-only correlation method)
1. The start of the time window described with reference to FIGS. 3 and 4 is a large factor for specifying the predicted time (for example, time t11 in FIG. 4). The time window having the maximum correlation coefficient is searched for while shifting the time window WINn in FIG. 3. The one having the higher time correlation of other elements with respect to the start time of the searched time window is searched. There are things like power flow, voltage, frequency, reactive power, inverter behavior, DER behavior, weather, temperature, humidity, barometric pressure (if found, use that element and sift time window WINn in advance The number of pattern matching samples by the phase-only correlation method can be reduced in advance).

・ Selection of illuminance sensor 1. Statistics of the illuminance sensor selected by the phase only correlation method (the illuminance sensor 10n of the surrounding PV-n having the highest correlation with the output of the prediction target PV-0) are taken. The illuminance sensor 10n that is hardly selected is excluded from the candidates (this can reduce the calculation time).

  Next, another example of automatic correction using the PV short-term prediction evaluation performed by the prediction evaluation / adjustment unit 78 will be described.

-Use of mismatch between PV short-term predicted value and actual measured value Specify the cause of the mismatch (obtain correlation with other factors) and use it to automatically improve prediction accuracy. By changing the parameters, it is useful for subsequent control decisions.

(Discovery of circulation fluctuations)
1. 1. For the same measurement point, calculate the phase correlation of the waveform with respect to the arbitrary time window WINn shown in FIG. 2. Measure the intensity of periodicity by spectral analysis using Fourier transform for the change in correlation coefficient. If there is a strong periodicity, it is considered that there is a circulation fluctuation, and that period is set as the width of the time window WINn for the time being. Conversely, when periodicity is not seen, it can be regarded as an irregular motion, and for the time being, the prediction value correction from the history data (for example, the short-term peripheral PV history data DB 71 in FIG. 8) is canceled.

(Discovering an appropriate time window width)
1. The width of the time window WINn is expected to greatly affect the prediction accuracy. 1. Randomly change in an arbitrary ratio with respect to a predetermined maximum value of the time window width, and periodically perform an operation of obtaining a correlation with the degree of inconsistency. Even after stabilization, it is changed from time to time to check the correlation with the degree of inconsistency, and if necessary, it is corrected.

(Confirmation probability)
1. 1. Calculate the degree of inconsistency from the value of the prediction error so that the probability can be judged numerically. Find the other elements with high time correlation. Candidate elements include the following: tidal current, voltage, frequency, reactive power, inverter behavior, DER behavior, weather, temperature, humidity, barometric pressure (if found, use that element in advance The number of samples for pattern matching by the phase-only correlation method can be reduced in advance.
3. Useful for optimal supply and demand control including uncertainties (optimal calculation priority when accuracy is high, continuity priority when accuracy is low).

  As described above, the inclination prediction of the output fluctuation performed by the microgrid system of FIG. 8 (the respective units 71 to 78) has an advantage that the prediction accuracy is high because of the time spread.

  Using this advantage, the microgrid system side can perform integrated predictive control, which further increases the synergistic effect of the integrated predictive control on the microgrid system side and the individual control on the distributed power supply side. Highly accurate optimum control can be performed.

  Further, according to the system of FIG. 8, the load can be used as a stabilization device, which leads to cost reduction and improvement of reliability, and is a key to promote the spread of distributed power sources.

  The short-term optimum operation plan created by the short-term optimum operation plan creation unit 77 of the microgrid system in FIG. 8 realizes the optimum distribution of electric power energy. Next, an example of the optimum operation distribution calculation will be described with reference to FIG. It explains together.

  The embodiment of FIG. 10 executes an optimal calculation method for each time based on linear programming, and the constraint condition expression used in the 24-hour optimal operation plan of the microgrid system is converted into an optimal operation for each time of the optimal calculation engine 80 for each time. It is used as a generalization constraint conditional expression.

  In this case, integer element constraints such as time constraints and device (distributed power supply) on / off constraints are excluded.

  First, short-term prediction of solar power generation (for example, prediction data of the short-term prediction unit 61 in FIG. 7) is input (illustration (1)). At this time, a value obtained by subtracting the PV short-term prediction from the 24-hour load prediction is used as the current short-term prediction value and input for optimization calculation by the hourly optimal calculation engine 80 ((2) in the figure).

  Next, the optimum calculation engine for each time 80 divides the short-term prediction curve created by the tens of minutes PV prediction curve creation unit 75 at an arbitrary time, and obtains an optimum solution in each time zone. The output value and load value of the current distributed power source are set as initial values.

  At each time, the overall optimal calculation is performed in accordance with the overall optimal objective function (the current objective function excluding the time aggregation calculation: economic / environmental selection is possible). Depending on the approximation method of the economic efficiency curve, linear programming (LP), quadratic programming (QP), n-order programming, and sequential quadratic programming (SQP) are properly used (recommended: sequential quadratic programming) (illustration (3 )).

  Next, the process of (3) in the figure is repeated for the time (illustration (4)).

  Take out the operation schedule for each distributed power source and notify each distributed power source, and each distributed power source monitors the voltage, current, and frequency of its own end based on the received schedule, and optimizes it individually according to its role Operation is performed ((5) in the figure).

  An example of the objective function and the constraint condition expression in the optimal calculation engine 80 for each time is as follows.

<Integrated EMS Optimization Objective Function by Time>
CO ₂ [kg−CO ₂ ] → minimum CO ₂ [kg−CO ₂ ] = purchased power CO ₂ [kg−CO ₂ ] + electric heat source CO ₂ [kg−CO ₂ ]
Purchased electric power CO ₂ [kg−CO ₂ ] = purchased electric energy [kWh] × purchased electric power CO ₂ emission basic unit [kg−CO ₂ / kWh]
Electric heat source CO ₂ [kg−CO ₂ ] = city gas consumption [Nm ³ ] × city gas CO ₂ emission basic unit [kg−CO ₂ / Nm ³ ].

<Integrated EMS time-specific optimization constraint formula>
Supply power [kW] = Power demand [kW]
Supply power [kW] = Supply power 1 [kW] + Supply power 2 [kW] + Supply power 3 [kW] + Supply power 4 [kW]
Power demand [kW] = Power demand 1 [kW] + Power demand 2 [kW] + Power demand 3 [kW] + Power demand 4 [kW]
Supply power 1 [kW] = purchased power 1 [kW] + heat source power output 1 [kW] + storage battery discharge output 1 [kW]
Electricity source power output 1 [kW] = city gas consumption 1 [Nm ³ ] × power generation coefficient 1 [kWh / Nm ³ ]
Electric power demand 1 [kW] = electric power load 1 [kW] + storage battery charging electric power 1 [kW] + air conditioning electric power 1 [kW] + hot water supply electric power 1 [kW] ……
Supply power 2 [kW] = Power demand 2 [kW]
Supply power 2 [kW] = purchased power 2 [kW] + heat source power output 2 [kW] + storage battery discharge output 2 [kW]
Electric power source power output 2 [kW] = city gas consumption 2 [Nm ³ ] × power generation coefficient 2 [kWh / Nm ³ ]
Electric power demand 2 [kW] = electric power load 2 [kW] + storage battery charging electric power 2 [kW] + air conditioning electric power 2 [kW] + hot water supply electric power 2 [kW].

  As described above, in the present embodiment, the output of the surrounding solar power generation facility PV-n having the highest correlation with the output data of the prediction target solar power generation facility PV-0 in a predetermined time is the current time zone. Is also assumed to be the same as the output of PV-0, so this is determined as the output predicted value of PV-0. As a result, a highly accurate output predicted value can be easily obtained.

  Since the supply and demand control of the microgrid is performed based on this highly accurate predicted output value, it is possible to perform a preparation operation or a fluctuation mitigation operation corresponding to the output fluctuation.

  In addition, in the said Example, although the surrounding photovoltaic power generation equipment PV-n was provided with two or more, one PV-n may be sufficient. In this case, for example, when the similarity between the outputs of PV-0 and PV-n obtained by the method of FIG. 2 is equal to or greater than a set threshold value, the PV-n output data is used as the PV-0 output predicted value. The predicted output value is determined by a method such as

1,51 ... PV
2 ... Monogeneration 3, 52 ... Storage battery 4 ... Cogeneration 5 ... Heat storage device 6 ... Electric power load 7 ... Thermal load 10, 10n ... Illuminance sensor 20 ... Photovoltaic power generation output predicted value calculation unit 30 ... Control unit 53 ... Reciprocating generator 54 ... Important load 55 ... Load that can be shut off 56 ... Load that cannot be shut off 61 ... Short-term prediction unit 62 ... High-speed measurement control unit 71 ... Short-term peripheral PV history data DB
72 ... Several tens of seconds PV prediction unit 73 ... Several tens of seconds PV inclination prediction unit 74 ... Several tens of PV prediction unit 75 ... Several tens of PV prediction curve creation unit 76 ... Decision unit 77 ... Short-term optimum operation plan creation unit 78 ... Prediction Evaluation / adjustment unit 80 ... Optimum calculation engine by time

Claims (11)

A microgrid supply and demand control system comprising at least a plurality of photovoltaic power generation facilities,
A solar power generation facility to be predicted, which is a target of solar power generation output prediction, and a surrounding solar power facility installed around the solar power generation facility to be predicted Solar power output measuring means for measuring the output of the power generation facility;
The output of the prediction target solar power generation facility and the output of the surrounding solar power generation facility measured by the solar power generation output measuring means a predetermined time ago are each Fourier transformed, and the output of the prediction target solar power generation facility after the Fourier transform Based on the phase difference of the output of the surrounding photovoltaic power generation equipment for each frequency, the degree of correlation between the output of the forecasting photovoltaic power generation equipment and the output of the surrounding photovoltaic power generation equipment is obtained by the phase-only correlation method, and the forecasting photovoltaic power generation equipment A solar power output predicted value calculating means for determining the current output of the surrounding solar power generation facility having the highest correlation with the output of
Control means for performing supply and demand control of the microgrid using the output predicted value determined by the photovoltaic power generation output predicted value calculating means as a control system parameter;
Microgrid supply and demand control system characterized by comprising: A plurality of the surrounding solar power generation facilities are arranged,
The photovoltaic power generation output predicted value calculating means is
Obtain solar power output data for the past set time of the prediction target solar power generation facility and the surrounding solar power generation facility measured by the solar power generation output measuring means,
The acquired photovoltaic power generation output data is Fourier transformed, and based on the phase difference for each frequency of the photovoltaic power generation output data after the Fourier transformation, the similarity with the output data of the prediction target photovoltaic power generation equipment Calculate the output data of the surrounding solar power generation facilities,
The present photovoltaic power generation output data of the surrounding photovoltaic power generation facility having the highest calculated similarity is acquired from the photovoltaic power generation output measuring means as an output predicted value of the prediction target photovoltaic power generation facility. The supply and demand control system for a microgrid according to claim 1.   The control means predicts the slope of the output fluctuation based on the data of millisecond to several seconds of the predicted output value determined by the predicted photovoltaic power generation output value calculation means, and according to the predicted slope of the output fluctuation. The supply / demand control system for a microgrid according to claim 1, wherein start / stop of supply / demand adjustment control is determined.   The control means predicts the width of output fluctuation based on the data of several seconds to several minutes of the predicted output value determined by the predicted photovoltaic power generation output value calculation means, and the supply and demand according to the predicted width of output fluctuation. The supply / demand control system for a microgrid according to any one of claims 1 to 3, wherein an adjustment resource is selected.   The control means determines an optimum operation plan for each distributed power source in the microgrid by a mathematical programming method based on the output predicted value determined by the photovoltaic power generation output predicted value calculation means and the attribute following the output fluctuation. The supply / demand control system for a microgrid according to any one of claims 1 to 4, wherein the supply / demand control system according to any one of claims 1 to 4 is configured to give an operation and stop command to each distributed power source.   The optimal operation plan is a linear programming method for each predetermined time zone using an optimization constraint equation and an objective function excluding time constraint conditions and distributed power source on / off constraint conditions divided for each predetermined time zone. The microgrid supply and demand control system according to claim 5, wherein the microgrid supply and demand control system is created by performing optimization calculation according to the above. A supply and demand control method for a microgrid equipped with at least a plurality of photovoltaic power generation facilities,
The solar power provided in each of the prediction target solar power generation equipment that is the target of the prediction of the solar power output in the power system and the surrounding solar power generation equipment provided around the prediction target solar power generation equipment Photovoltaic power output measuring means measures the output of each photovoltaic power generation facility, solar power generation output measuring step,
Photovoltaic power generation output predicted value calculation means for obtaining solar power generation output data for the past set times of the prediction target solar power generation equipment and surrounding solar power generation equipment measured by the solar power generation output measurement means Power generation output data acquisition step;
The photovoltaic power generation output predicted value calculation means Fourier-transforms the acquired photovoltaic power generation output data, and based on the phase difference for each frequency of the photovoltaic power generation output data after the Fourier transformation, the prediction target photovoltaic power generation facility A similarity calculation step for calculating the output data of all the surrounding photovoltaic power generation facilities,
The photovoltaic power generation output predicted value calculation means obtains the current photovoltaic power generation output data of the surrounding photovoltaic power generation facility having the highest calculated similarity from the photovoltaic power generation output measurement means, and predicts the acquired data A predicted value determining step for determining the output predicted value of the target photovoltaic power generation facility;
A control step for controlling supply and demand of the microgrid using the output predicted value determined by the photovoltaic power generation output predicted value calculating unit as a parameter of a control system;
A supply and demand control method for a microgrid characterized by comprising:   The control step predicts the slope of the output fluctuation based on the data of the output prediction value determined in the prediction value determination step for milliseconds to several seconds, and the supply and demand adjustment control is performed according to the predicted slope of the output fluctuation. The supply / demand control method for a microgrid according to claim 7, wherein start / stop is determined.   The control step predicts a range of output fluctuation based on data of a few seconds to several minutes of the predicted output value determined by the predicted value determination step, and selects a supply and demand adjustment resource according to the predicted output fluctuation range The method for controlling supply and demand of a microgrid according to claim 7 or 8, characterized in that:   The control step creates an optimal operation plan for each distributed power source in the microgrid by a mathematical programming method based on the predicted output value determined in the predicted value determination step and the attribute that follows the output fluctuation, The supply / demand control method for a microgrid according to any one of claims 7 to 9, wherein a command for operation and stop is given to the distributed power source.   The optimal operation plan is a linear programming method for each predetermined time zone using an optimization constraint equation and an objective function excluding time constraint conditions and distributed power source on / off constraint conditions divided for each predetermined time zone. The method for controlling supply and demand of a microgrid according to claim 10, wherein the supply and demand control method is performed by performing optimization calculation according to the above.

Images