JP6617476B2 - Demand power prediction apparatus, demand power prediction method, and computer program - Google Patents

Description

The present invention relates to a demand power prediction apparatus and method, and a computer program.
Specifically, the present invention relates to an improvement in a method for predicting demand power of power equipment including a load device as a power device to be managed.

As one of the functions of an energy management system (hereinafter referred to as “EMS”), for example, it is known to calculate an operation plan of electric power equipment included in electric power equipment by using linear programming. (See Patent Document 1).
The operation plan calculation process sets constraints such as device conditions modeled with variables for each time step of the power equipment to be managed, and aims to minimize costs under the set constraints. In order to satisfy the function, the solution (operation plan) of the variable of the electric power equipment for each time step is obtained.

In the calculation of the above operation plan, since the predicted value of the demand power of the power equipment is one of the input information, it is necessary to accurately calculate the predicted value of the demand power.
For example, techniques disclosed in Patent Documents 2 to 4 are known as conventional techniques for calculating a predicted value of such demand power.

In patent document 2, the period of the actual weather group similar to the weather of the predicted weather group including the forecast target date of the demand power and its immediate vicinity is extracted, and the weather data and demand for the period of the predicted weather group and the actual weather group are extracted. Create a demand forecast model using actual results.
By inputting the actual weather, the actual demand, or the predicted weather into this demand prediction model, the demand on the prediction target day is predicted, and the predicted demand is displayed.

In patent document 3, the maximum temperature of each day and the maximum demand amount and the minimum demand amount of the power of the day are extracted. If there are days with the same maximum temperature but different demand, the average value of the demand for each day is obtained. And the demand curve which connects the maximum demand amount and minimum demand amount of each temperature, and shows the electric power demand amount according to temperature is created.
Similarly, the minimum temperature of each day, the maximum demand amount and the minimum demand amount of power for the day are extracted, and a demand curve indicating the power demand amount by temperature is created. Demand prediction is performed by fitting the predicted temperature to this demand curve.

  In Patent Literature 4, the past day in which the three-day calendar division is the same as the prediction target day, the day in which the highest temperature / lowest temperature is closest to the highest temperature / lowest temperature forecast value of the prediction target day is predicted as described above. Obtained as the past date that is most similar to the target date (past similar date). And the performance data of this past similar day is acquired and demand prediction is performed.

JP 2015-35941 A Japanese Patent No. 5492848 JP, 2014-180187, A JP 2011-114944 A

In Patent Document 2, since the change in power demand is considered to be similar for periods with similar temperature changes, and there are few samples for forecasting singular days such as Saturdays, Sundays, and holidays, the correction value is added to the actual power demand value on weekdays. Is used as a past performance value.
However, the shape of the demand curve for weekdays and holidays is different, and the number of past samples obtained under conditions such as holidays during the same period in the past is small, so there is a high possibility that similar periods cannot be extracted well, which affects forecasting It will be.

In Patent Document 3, the relationship between the peak value of demand power and the temperature is calculated while the change in the power demand due to the influence of the temperature and the value of the power demand change due to human activities are mixed.
However, in order to add purely temperature correction to the demand power curve on similar days, it is necessary to calculate a coefficient considering only the effect of temperature. For example, even if the temperature is the same at dawn and night, there is a difference in the amount of equipment used in customer equipment, and the amount of power consumed by air conditioning is different. For this reason, it is conceivable that the influence of the temperature on the power demand during a time period when human activity is not active appears relatively larger than that during the activity time period.

In Patent Document 4, a past day in which the demand power has the same tendency is selected, and the similarity is determined based on the maximum / minimum temperature.
However, changes in temperature during the day are ignored. For example, since the change in the power demand when the temperature changes suddenly is ignored, there is a risk that the prediction error becomes large.

  In view of the above-described conventional problems, an object of the present invention is to provide a demand power prediction apparatus and the like that can accurately predict demand power of power equipment.

(1) An apparatus according to an aspect of the present invention is an apparatus for predicting demand power of power equipment, and includes two linear regression coefficients having different applicable temperature ranges representing a relationship between temperature and demand power, and A storage unit that stores the intercept for each predetermined time zone of each day of the week, and the coefficient and the intercept of the same time zone of the same day of the week as the predicted time after the current time in the temperature range including the predicted temperature of the predicted time A controller that reads out the coefficient and intercept to be used from the storage unit, applies the read coefficient and intercept to the predicted temperature of the predicted time, and calculates the demand power of the power facility at the predicted time.

(2) The computer program which concerns on 1 aspect of this invention is a computer program for making a computer perform the process which estimates the demand power of electric power installation, Comprising: In the said process, the relationship between temperature and demand power is included. The storage processing for storing the coefficients and intercepts of two linear regression equations representing different applicable temperature ranges in the storage unit for each predetermined time zone of each day of the week, and the same time zone of the same day of the week as the predicted time after the current time Of the coefficient and intercept, the coefficient and intercept used for the temperature range including the predicted temperature of the predicted time are read from the storage unit, and the read coefficient and intercept are applied to the predicted temperature of the predicted time, And a calculation process for calculating the power demand of the power facility at the predicted time.

(3) A method according to an aspect of the present invention is a method for predicting demand power of power equipment, and includes two linear regression coefficients having different applicable temperature ranges representing the relationship between temperature and demand power, and Storing the intercept in the storage unit for each predetermined time zone of each day of the week, and the temperature including the predicted temperature of the predicted time among the coefficient and the intercept of the same time zone on the same day of the week as the predicted time after the current time Reading the coefficient and intercept used for the range from the storage unit, applying the read coefficient and intercept to the predicted temperature of the predicted time, and calculating demand power of the power facility at the predicted time. .

  ADVANTAGE OF THE INVENTION According to this invention, the demand power of electric power equipment can be estimated accurately.

It is a block diagram which shows the structural example of the electric power system which concerns on embodiment of this invention. It is a block diagram which shows an example of a functional structure of an EMS server. It is explanatory drawing of a plan period and the step size (step space | interval) of a plan. It is a graph which shows the relationship between the temperature for one year in electric power facilities, and demand electric power. It is a graph which shows the relationship between the temperature of 1 hour of a specific day in power equipment, and demand power. It is explanatory drawing which shows an example of the regression equation for every hour of each day of the week. It is a graph showing the effect of the demand forecast based on temperature. It is a graph showing the relationship between the time zone on weekdays and the number of people in the factory.

<Outline of Embodiment of the Present Invention>
Hereinafter, an outline of embodiments of the present invention will be listed and described.
(1) An apparatus according to an embodiment of the present invention is an apparatus that predicts demand power of power equipment, and includes two linear regression coefficients having different applicable temperature ranges representing the relationship between temperature and demand power, and A storage unit that stores the intercept for each predetermined time zone of each day of the week, and the coefficient and the intercept of the same time zone of the same day of the week as the predicted time after the current time in the temperature range including the predicted temperature of the predicted time A controller that reads out the coefficient and intercept to be used from the storage unit, applies the read coefficient and intercept to the predicted temperature of the predicted time, and calculates the demand power of the power facility at the predicted time.

According to the demand power prediction apparatus of the present embodiment, the control unit uses the coefficient and intercept used for the temperature range including the predicted temperature of the predicted time, among the coefficient and intercept of the same time zone on the same day of the week as the predicted time after the current time. Read from the storage unit, apply the read coefficient and intercept to the predicted temperature of the predicted time, and calculate the demand power of the power facility at the predicted time.
For this reason, the power demand which fluctuates according to the change of human activity of the power equipment having an air conditioning equipment or the like as the load device can be accurately calculated.

(2) In the demand power prediction apparatus of this embodiment, it is preferable that the time length of the predetermined time zone is set to be equal to or less than a working time (for example, 8 hours) of a business entity that uses the power equipment.
The reason is that when the predetermined time zone exceeds the above time, demand power data in a state where the number of people in the station is greatly different will be mixed in the same time zone, and the accuracy of the coefficient and intercept may be deteriorated. Because there is sex.

(3) The computer program of this embodiment is related with the program for making a computer perform the process which the above-mentioned demand power prediction apparatus performs.
For this reason, the computer program of this embodiment has the same operation effect as the above-mentioned demand power prediction device.

(4) The demand power prediction method of the present embodiment relates to a demand power prediction method executed by the demand power prediction apparatus described above.
For this reason, the demand power prediction method of this embodiment has the same effect as the above-mentioned demand power prediction apparatus.

<Details of Embodiment of the Present Invention>
Hereinafter, details of embodiments of the present invention will be described with reference to the drawings. In addition, you may combine arbitrarily at least one part of embodiment described below.
[Overall system configuration]
FIG. 1 is a block diagram illustrating a configuration example of a power system according to an embodiment of the present invention.

As shown in FIG. 1, the power system of this embodiment includes an EMS server 1 and a power facility 2 that is a management target of the EMS server 1. The EMS server 1 manages the operating states of various power devices included in the power facility 2.
The EMS server 1 of this embodiment consists of a FEMS (Factory Energy Management System) server, for example. Therefore, the power facility 2 includes a power distribution network composed of the distribution lines 3 wired in the factory, and a load device 4, a power generation device 5, and a power storage device 6 connected to the distribution lines 3.

The load device 4 includes, for example, a non-adjustable load device such as a production machine in which power adjustment is not possible or possible even if power adjustment is not possible. The load device 4 may include an adjustable load device capable of adjusting power consumption, such as lighting and an air conditioner.
The load device 4 is connected to the distribution line 3 via a device capable of controlling and measuring power information, such as a smart tap (not shown) or a smart distribution board.

The power generation device 5 includes, for example, a power generation device that converts energy resulting from a chemical change into electrical energy, such as combustion energy such as gas or diesel oil, or a fuel cell. The power generator 5 is connected to the distribution line 3 via a one-way DC / AC converter (not shown).
The power storage device 6 includes, for example, at least one of a redox flow (RF) battery, a lithium ion battery, a molten salt battery, and a lead storage battery. The power storage device 6 is connected to the distribution line 3 via a bidirectional DC / AC converter (not shown).

  In the power facility 2 of the present embodiment, the distribution line 3 is connected to the power system (commercial power source) 7 via a measuring device such as a smart meter. For this reason, the power facility 2 can be connected to the power system 7.

The EMS server 1 is connected to various power devices of the power facility 2 via the communication line 8, and forms a wired LAN (Local Area Network) with the various power devices. The communication between the EMS server 1 and the power device may be wireless communication such as a wireless LAN.
The EMS server 1 can transmit a plurality of types of control commands E <b> 1 to E <b> 3 to communicable power devices included in the power facility 2. The EMS server 1 can receive current information S <b> 1 representing the operation status of the power equipment 2 from communicable power equipment included in the power equipment 2.

The control command E <b> 1 is a control command related to the control of the load device 4. For example, the EMS server 1 can turn on or off a smart tap to which the load device 4 is connected by a control command E1.
The EMS server 1 can also adjust the power consumption of a load capable of adjusting the power consumption included in the load device 4 by transmitting a control command E1 to the load device 4.

The control command E <b> 2 is a control command related to the control of the power generation device 5. For example, the EMS server 1 can turn on or off the DC / AC converter connected to the power generation device 5 according to the control command E2.
The EMS server 1 can also adjust the power generation amount of the power generation device 5 by transmitting a control command E2 to the power generation device 5 capable of adjusting the power generation amount.

Control command E <b> 3 is a control command related to the control of power storage device 6. For example, the EMS server 1 can turn on or off the DC / AC converter connected to the power storage device 6 by the control command E3.
The EMS server 1 can also adjust at least one of charging power and discharging power for the power storage device 6 connected to the distribution line 3 by a control command E3. This adjustment is performed, for example, by adjusting the duty ratio for the PWM circuit included in the DC / AC converter.

The EMS server 1 receives current information S1 including the connection status (on / off) of various converters and smart taps of the power equipment 2 and the operation status and power value of each device 4 to 6 for a predetermined time (for example, 1 Every second).
The current information S1 acquired by the EMS server 1 includes the value of the remaining amount of charge of the power storage device 6 (hereinafter also referred to as “SOC (State Of Charge)”) or the current information necessary for calculating the SOC. .

The current SOC value can be calculated by any one of a table reference method, a current integration method, and a combination thereof.
The table reference method is a method for obtaining the SOC value corresponding to the open circuit voltage estimated from the terminal voltage of the battery cell from the reference table stored in advance. The current integration method is a method of calculating the SOC value by integrating the current flowing through the battery cell every minute time.

The current SOC value may be autonomously calculated by the power storage device 6 and notified to the EMS server 1 or may be calculated by the EMS server 1.
In the former case, the power storage device 6 transmits the SOC value calculated by itself to the EMS server 1 as current information S1, and the communication unit 13 (see FIG. 1) of the EMS server 1 receives the transmitted SOC value. Therefore, in this case, the communication unit 13 of the EMS server 1 becomes a current SOC value acquisition unit.

In the latter case, the power storage device 6 transmits the current voltage value and current value of the battery cell to the EMS server 1 as current information S1, and the control unit 11 (see FIG. 1) of the EMS server 1 receives the received voltage value and The SOC value may be calculated based on the current value.
Therefore, in this case, the control unit 11 of the EMS server 1 itself becomes a current SOC value acquisition unit.

[Configuration of EMS server]
As shown in FIG. 1, the EMS server 1 is configured by a computer device including a control unit 11, a storage unit 12, and a communication unit 13.
The control unit 11 includes an information processing apparatus including a CPU (Central Processing Unit). The storage unit 12 includes a memory including a RAM (Random Access Memory) and a large-capacity storage unit such as an HDD (Hard Disk Drive).

Although not shown in FIG. 1, the EMS server 1 manages an input device including a mouse and a keyboard for an operation input by an administrator of the power equipment 2 and image data output from the control unit 11. A display device such as a liquid crystal display to be presented to a person is connected.
The communication unit 13 can communicate with various power devices included in the power facility 2 by a wired LAN, a wireless LAN, or other communication methods.

The control unit 11 reads out and executes the computer program stored in the storage unit 12, thereby performing communication control on the communication unit 13, input / output control on the input device and the display device, energy management of the power equipment 2 to be managed, and the like. Various controls are performed.
The communication unit 13 transmits control commands E1 to E3 to each communicable power device included in the power facility 2 based on communication control by the control unit 11, and presents current information S1 indicating the operation status of the power facility 2 to each of the current information S1. Received from the power device and transferred to the control unit 11.

The control unit 11 of the EMS server 1 can execute “plan control” and “dynamic control” as part of energy management for the power equipment 2.
The plan control is a control for calculating a future operation plan of the power facility 2 over a relatively long period (for example, 24 hours) and operating the power equipment according to the calculated operation plan. The dynamic control is control for operating a predetermined power device in order to cope with a disturbance such as an increase in demand that occurs during execution of an operation plan.

When the energy management for the power facility 2 is performed, the control unit 11 loads the load device 4, the power generation device 5, and the power storage device included in the power facility 2 so that the power supply and demand at the receiving point Q (see FIG. 1) of the power system 7 is balanced. 6 is controlled.
The reason is that, for example, the received power (instantaneous value) fluctuates greatly, and if the 30-minute average value of the received power exceeds a predetermined target power (≦ contract power), there is a high possibility that a penalty for the power company will occur. Because.

For this reason, the control part 11 is calculating the present demand power in the electric power equipment 2 based on the collected various present information S1.
The current power demand in the power facility 2 can be calculated by summing the current received power (for example, the measured value of the smart meter), the current generated power by the power generation device 5 and the current discharge power by the power storage device 6. it can. The current demand power in the power equipment 2 can also be calculated by summing the current power consumption of the load device 4.

[Configuration of operation plan calculation device]
FIG. 2 is a block diagram illustrating an example of a functional configuration of the control unit 11 of the EMS server 1.
As shown in FIG. 2, the control unit 11 of the EMS server 1 includes a power generation prediction unit 31, a demand prediction unit 32, a plan calculation unit 33, and a plan execution unit 34 as functional parts realized by execution of the computer program. ing.

When a natural energy power generation device such as a solar power generation device is included in the power equipment 2, the power generation prediction unit 31 calculates a predicted value of the power generation amount of the power generation device based on weather information and the like. The power generation prediction unit 31 records the calculated predicted value in the power generation prediction database DB1 included in the storage unit 12.
In the case of the power equipment 2 that does not include the natural energy power generation device, the power generation prediction unit 31 may be omitted or the power generation prediction unit 31 may not function.

Based on the past actual power demand values (time-series data of measured values of the smart meter) and weather information, etc., the demand prediction unit 32 uses, for example, a statistical method to calculate the future demand power in the power equipment 2. Calculate the predicted value.
The demand prediction unit 32 calculates a predicted value of demand power for a predetermined prediction period (for example, 24 hours) from the present time every unit time (for example, 30 minutes), and the calculated predicted value is included in the storage unit 12. It records in database DB2 for demand prediction.

The storage unit 12 includes three databases DB1, DB2, and DB3. The database DB1 is a database for predicted power generation values. The power generation predicting unit 31 stores the calculated predicted power generation amount in the database DB1 in time series in association with the date data.
The database DB2 is a database for predicted power demand values. The demand prediction unit 32 stores the calculated predicted value of demand power in the database DB2 in time series in association with the date data.

In the database DB3, coefficients a and intercepts b of a plurality of linear regression equations or other parameters that are independent for each specified time zone are recorded. The coefficient a and the intercept b are coefficients and intercepts that are defined for each time zone when the relationship between the temperature and the power demand is linearly approximated. For example, for each hour of each day of the week from Sunday to Saturday, It is predetermined as a coefficient value to be applied. The time zone that defines the coefficient a and the intercept b may be a weekday, a holiday, etc., instead of a day of the week. Further, the time length of one period in the specified time zone is not limited to one hour, and other time lengths may be set.
The demand prediction unit 32 reads the coefficient a and the intercept b of the time zone corresponding to the forecast time tp from the database DB3, applies the coefficient a and the intercept b to the forecast temperature T (tp) at the forecast time tp, and performs the forecast Demand power P (tp) of the power equipment at time tp is calculated.

As the predicted temperature T (tp) at the predicted time tp, for example, a value obtained from the predicted temperature included in the weather information announced by the Japan Meteorological Agency at a predetermined announcement time can be used, and the measured temperature at the latest time can be used. Alternatively, it may be used. Specific examples of the coefficient a and the intercept b for each hour of each day will be described later.

  The plan calculation unit 33 uses the power generation amount and the predicted value of the demand power recorded in the databases DB1 and DB2 and the status information (state variable) of the power equipment included in the current information S1, and the power included in the power equipment 2 For at least one target device of the devices 4 to 6, the calculation of the operation plan in the predetermined planning period Tp is repeatedly executed every predetermined re-planning period.

FIG. 3 is an explanatory diagram of the plan period Tp and the step size (step interval) Tc of the plan.
As shown in FIG. 3, the plan calculation unit 33 sets a plan period Tp having a predetermined time length (for example, 24 hours) from the present to the future to a step size of the plan having a predetermined time length (hereinafter also referred to as “step interval”) Tc. Each time step t is divided into t (t = 1 to N: N = Tp / Tc).
And the plan calculation part 33 makes the state variable (variable showing ON / OFF of the said apparatus, connection, connection cancellation | release, electric energy etc.) of the electric power equipment contained in the electric power installation 2 of the N-dimensional time step t (time variable). Expands to a relational expression.

Thereafter, the plan calculation unit 33 obtains a variable value of each time step t at which the predetermined objective function is minimum (or maximum), thereby at least one of the devices 4 to 6 of the power equipment 2. An operation plan for each time step t is calculated.
That is, the operation plan represents the operation state of the target device for each time step t included in the predetermined plan period Tp. For example, mathematical programming represented by linear programming can be used as an algorithm when performing the planning calculation by the plan calculation unit 33.

  The plan calculation unit 33 according to the present embodiment has a processing capability to repeatedly calculate an operation plan every predetermined “re-planning period” (for example, 30 minutes) that is significantly shorter than the planning period Tp. In this case, the plan calculation unit 33 outputs the calculated operation plan to the plan execution unit 34 for each replanning period.

When the operation plan is input from the plan calculation unit 33, the plan execution unit 34 generates control commands E <b> 1 to E <b> 3 based on the operation plan, and sends the control commands E <b> 1 to E <b> 3 of the power equipment 2 via the communication unit 13. Send to each power device.
Specifically, when the plan execution unit 34 acquires a control command E1 for connecting or disconnecting the load device 4 or adjusting power consumption, the plan execution unit 34 sends the control command E2 to a smart tap or power consumption in the power facility 2. To the load device 4 that is the adjustment target.

When the plan execution unit 34 acquires the control command E2 for connecting or disconnecting the power generation device 5 or adjusting the power generation amount, the plan execution unit 34 adjusts the DC / AC converter for the power generation device 5 or the power generation amount. Transmit to possible power generator 5.
When the plan execution unit 34 acquires the control command E3 for connecting or disconnecting the power storage device 6 or adjusting the charge / discharge power, the plan execution unit 34 sends the control command E3 to the DC / AC converter of the power storage device 6 or the power storage device 6. Send to.

[Principle of power demand forecasting method]
The load device 4 of the power facility 2 in the factory includes a load device whose demand power is considered to fluctuate depending on the air temperature, such as an air conditioner or a heating facility, in addition to a constant power demand of a production machine or the like.
Therefore, the inventor of the present application investigated whether or not there is a correlation between the air temperature and the power demand at the company's own factory.

FIG. 4 is a graph showing the relationship between the temperature and power demand for one year in the power equipment 2. That is, FIG. 4 is created by plotting the data for one year of the actual values of the temperature and power demand in the applicant's own factory on a single graph.
As shown in FIG. 4, when the total data is the actual values of all temperatures and demand power for one year, no significant correlation was found between the temperature and demand power.

Therefore, it has been found that demand power is affected by human activities in addition to temperature, and it is difficult to correlate with demand power considering only the influence of temperature independently.
Therefore, the present inventor assumed a time zone for each day of the week as a parameter that affects the actual activity of human beings. In other words, the temperature and power demand data for one year are divided into hourly data for each day of the week, and whether or not there is a correlation between the hourly temperature of each day and the power demand. investigated.

FIG. 5 is a graph showing the relationship between the temperature of one hour on a specific day of the week and the demand power in the power facility 2. That is, FIG. 5 shows one hour of data from 10:00 to 2:00 on Monday among the data for one year of the actual values of temperature and demand power at the applicant's own factory. Is plotted in
As shown in FIG. 5, when the data was narrowed down to the actual value of the temperature and demand power for one hour on Monday, a significant correlation was recognized between the temperature and demand power.

Specifically, in 1 hour from 1:00 to 2:00 on Monday, the demand power P (t) is about a first straight line L1 applicable in a temperature range of about 20 ° C. or less and about 20 ° C. It has been found that it can be substantially represented by two straight lines that can be applied in the above temperature range and the second straight line L2.
It has also been found that the demand power P (t) can be expressed by a linear relationship between two straight lines L1 and L2 in the case of hourly data on other days of the week. That is, different linearity was seen in the relationship between temperature and demand power, regardless of the season.

According to the above example, past temperature and actual power demand values are classified into hourly data for each day of the week, and coefficients of straight lines L1 and L2 established between the hourly temperature and demand power for each day of the week. If a and the intercept b are stored in the storage unit 12 in advance, the predicted value of the demand power after the present time can be accurately calculated by applying the coefficient a and the intercept b to the predicted temperature of the same time on the same day of the week. It becomes like this.

[Specific examples of forecasting method of power demand]
FIG. 6 is an explanatory diagram showing an example of the regression equation for each hour of each day of the week.
In this case, as shown in FIG. 6, the database DB3 of the storage unit 12 includes a time table Tb in which each day of the week from Sunday to Saturday is divided into a total of 168 every hour. In each item of the time table Tb, the values of the coefficient a and the intercept b of the linear regression equation used in the time zone of each day of the week are stored in advance.

For example, in the item of 0:00 on Monday, a coefficient a1 and an intercept b1 applied to a temperature range of 20 ° C. or lower, and a coefficient a2 and an intercept b2 applied to a temperature range of 20 ° C. or higher are stored.
Similarly, the item of 22:00 on Friday stores a coefficient a3 and an intercept b3 applied to a temperature range of 20 ° C. or lower, and a coefficient a4 and an intercept b4 applied to a temperature range of 20 ° C. or higher. ing.

The demand prediction unit 32 (see FIG. 2) of the control unit 11 stores the coefficient a and the intercept b of the same day of the week and the same time zone after the current time when calculating the predicted value of the demand power of the power equipment 2. 12 is read out.
Moreover, the demand prediction part 32 calculates the demand power of the electric power equipment 2 in the said prediction time, applying the read coefficient a and intercept b to the prediction temperature in prediction time.

For example, when the time zone of the forecast date is 0:00 on Monday, the control unit 11 calculates the demand power by applying one of the following linear formats to the forecast temperature T (t) at the forecast time. .
P1 (t) = a1 × T (t) + b1 When the predicted temperature T (t) is 20 ° or less P2 (t) = a2 × T (t) + b2… The predicted temperature T (t) is 20 ° or more in the case of

When the time zone of the predicted date is 22:00 on Friday, the control unit 11 calculates the demand power by applying one of the following linear formats to the predicted temperature T (t) at the predicted time. .
P3 (t) = a3 × T (t) + b3 When the predicted temperature T (t) is 20 ° or less P4 (t) = a4 × T (t) + b4… The predicted temperature T (t) is 20 ° or more in the case of

FIG. 7 is a graph showing the effect of the demand prediction of this embodiment.
In FIG. 7, the horizontal axis is the time zone for one day, and the vertical axis is the amount of difference from the actual power. The thin line graph is the difference amount (absolute average error) when the temperature is not considered, and the thick line graph is the difference amount (absolute average error) when the temperature is considered. As shown in FIG. 7, prediction is made by dividing the year-round temperature and power demand data into predetermined time zones (for example, 1 hour) of each day of the week and considering the influence of the temperature for each time zone. The error of the value can be reduced.

[Effect of EMS server]
As described above, according to the EMS server 1 of the present embodiment, the control unit 11 reads from the storage unit 12 the coefficients a and intercepts b of two linear regression equations in the same time zone on the same day of the week as the predicted time after the current time. Then, by applying the read coefficient a and intercept b to the predicted temperature of the predicted time, the demand power of the power equipment 2 at the predicted time is calculated.
For this reason, for example, the power demand which fluctuates according to human activities, such as the power equipment 2 having the air conditioning equipment as the load device 4, can be accurately calculated.

Further, according to the EMS server 1 of the present embodiment, the coefficients a and the intercepts b classified and stored for each predetermined time zone of each day of the week are used. Compared to the method of setting the day, there are effects that it is easy to secure the number of samples to be data, and that there is an effect that a correlation between the temperature and power for demand is obtained by avoiding the influence of the season.

[Method of determining a predetermined time zone]
FIG. 8 is a graph showing the relationship between the time zone on weekdays and the number of people in the factory.
As shown in Fig. 8, the number of people in buildings such as factories including power facilities 2 usually changes suddenly before and after regular working hours except in special cases such as nighttime operation. Since demand fluctuates synchronously with changes in the number of people, electricity demand is also expected to change suddenly before and after regular working hours. For example, in the example of FIG. 8, the power demand sharply increases in the morning time from 7:00 to 9:00, and the power demand sharply decreases in the evening time from 17:00 to 19:00.

When the yearly temperature and actual power demand data is divided into predetermined time zones on each day of the week, when the time granularity is too large compared to working hours (for example, one day) It is considered that the power demand data in a state where the number of people in the site is greatly different are mixed in the same time zone, and the accuracy of the coefficient a and the intercept b deteriorates.
Considering this, the maximum granularity Tmax of the time zone for dividing the data is the time from the completion of the rapid increase in the number of people in the station to the start of the rapid decrease (for example, the working hours of the entity using the power equipment 2) = About 8 hours) is considered appropriate.

On the other hand, the minimum granularity Tmin in the time zone for dividing the data does not need to be particularly limited. However, if the time zone for classifying the data is made smaller than necessary, the number of data included in the same time zone becomes too small, and the accuracy of the prediction model may deteriorate.
For this reason, although it is also related to the number of data for one year as a parameter, the minimum granularity of the time zone for dividing the data is the number of data at which significant coefficient a and intercept b are expressed at least in the same time zone. It is necessary to set the time length that can be included.

[Other variations]
The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

  In the above-described embodiment, the case where the power equipment 2 that is the management target of the EMS server 1 includes the load device 4, the power generation device 5, and the power storage device 6 is illustrated, but the power equipment that the EMS server 1 manages power supply and demand. The minimum configuration of 2 is not limited as long as it includes at least the load device 4 and the power storage device 6 on the assumption that it is connected to the power system 7. Good.

  In the above-described embodiment, the FEMS having the EMS server 1 and the power facility 2 is exemplified. However, HEMS (Home Energy Management System), BEMS (Building Energy Management System), MEMS (Mansion Energy Management System), etc. The EMS server 1 which performs the prediction of the demand power of this embodiment can be employ | adopted.

DESCRIPTION OF SYMBOLS 1 EMS server 2 Electric power equipment 3 Distribution line 4 Load apparatus 5 Power generation apparatus (auxiliary power supply apparatus)
6 Power storage device (auxiliary power supply device)
7 Power system (commercial power)
8 communication line 11 control unit 12 storage unit 13 communication unit 31 power generation prediction unit 32 demand prediction unit 33 plan calculation unit 34 plan execution unit

Claims (4)

A device for predicting the power demand of power equipment,
A storage unit that stores coefficients and intercepts of two linear regression equations having different applicable temperature ranges representing the relationship between temperature and demand power for each predetermined time zone of each day;
Among the coefficients and the intercept of the same time zone of the same day as the estimated time moment after, reading out the coefficient and the intercept is used to the temperature range that includes a predicted temperature of the predicted time from the storage unit, the coefficient and the intercept was read And a control unit that calculates the demand power of the power facility at the predicted time by applying the value to the predicted temperature of the predicted time.   The demand power prediction apparatus according to claim 1, wherein the time length of the predetermined time zone is set to be equal to or less than a working time of a business entity that uses the power equipment. A computer program for causing a computer to execute a process for predicting power demand of power equipment, wherein the process includes:
A storage process for storing a coefficient and an intercept of two linear regression equations having different applicable temperature ranges representing the relationship between the temperature and demand power in a storage unit for each predetermined time zone of each day;
Among the coefficients and the intercept of the same time zone of the same day as the estimated time moment after, reading out the coefficient and the intercept is used to the temperature range that includes a predicted temperature of the predicted time from the storage unit, the coefficient and the intercept was read And a calculation process for calculating the demand power of the electric power facility at the predicted time by applying to the predicted temperature of the predicted time. A method for predicting the power demand of power equipment,
Storing the coefficients and intercepts of two linear regression equations having different applicable temperature ranges representing the relationship between temperature and demand power in the storage unit for each predetermined time zone of each day;
Among the coefficients and the intercept of the same time zone of the same day as the estimated time moment after, reading out the coefficient and the intercept is used to the temperature range that includes a predicted temperature of the predicted time from the storage unit, the coefficient and the intercept was read Applying the power to the predicted temperature of the predicted time to calculate the demand power of the power facility at the predicted time.