JP5734325B2 - Power prediction apparatus, facility control system, power prediction method, and program - Google Patents

Description

  The present invention relates to a power prediction apparatus, a facility control system, a power prediction method, and a program for predicting the amount of power consumed by facility equipment.

  Conventionally, a method of predicting demand power at a specified time from the slope of increase in power consumption by time series data, a method of predicting from past performance data, etc., a method of predicting power from an operation plan in a factory, etc. The power consumption is predicted by various methods, and demand control is executed based on the prediction results. In such power consumption prediction, as a method for improving the prediction accuracy, a technique such as a model that predicts based on an error between prediction using a plurality of prediction methods and actual data is disclosed (for example, Patent Documents). 1).

JP 2001-14295 A

  However, in the conventional method, since the prediction accuracy is not sufficient, it is necessary to control with a sufficient margin between the power management value of the demand control and the power consumption, and the living environment is deteriorated more than necessary. There is a case.

  The present invention has been made under the above circumstances, and an object thereof is to provide a power prediction device, an equipment control system, a power prediction method, and a program capable of accurately predicting the amount of power consumed. To do.

In order to achieve the above object, a power prediction apparatus according to the present invention includes:
A power prediction device that predicts the amount of power consumed by equipment,
For each of a plurality of prediction means for calculating the amount of power expected to be consumed per unit time, with the time that put in a past unit time, the actual amount of power consumed up to that point and the prediction means A past predicted power amount acquisition unit that acquires a past predicted power amount calculated to be consumed in the past unit time, calculated using
Based on the actual amount of power consumed in the calculated issued said past predicted power amount and the previous SL past unit time, from among the plurality of prediction unit, the prediction unit selector for selecting one or a plurality of prediction means When,
A current predicted power amount acquisition unit that acquires a current predicted power amount that is predicted to be consumed in a unit time including the current time, using the actual power amount consumed up to the present time and the selected one or more prediction means. When,
It is characterized by providing.

  According to the present invention, it is possible to predict the amount of power consumed accurately.

It is a block diagram which shows schematic structure of the equipment control system which concerns on Embodiment 1. FIG. It is a block diagram which shows the hardware constitutions of the electric power prediction apparatus which concerns on Embodiment 1. FIG. It is a schematic block diagram which shows the function structure of the control part of the electric power prediction apparatus which concerns on Embodiment 1. FIG. It is a figure for demonstrating the prediction method of the electric energy by a prediction means. It is a figure for demonstrating the prediction method of the electric energy by a prediction means. It is a figure which shows an example of the data stored in prediction means evaluation information DB. It is a block diagram which shows the hardware constitutions of the equipment control apparatus which concerns on Embodiment 1. FIG. It is a schematic block diagram which shows the function structure of the control part of the equipment control apparatus which concerns on Embodiment 1. FIG. It is a flowchart which shows an example of the flow of an electric power prediction process. It is a flowchart which shows an example of the flow of a prediction means evaluation process. It is a flowchart which shows an example of the flow of a past prediction electric energy calculation process. It is a flowchart which shows an example of the flow of a prediction means selection process. It is a flowchart which shows an example of the flow of a prediction means selection process. It is a schematic block diagram which shows the function structure of the control part of the equipment control apparatus which concerns on Embodiment 2. FIG. It is a figure which shows an example of the data stored in apparatus power consumption reduction DB.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

(Embodiment 1)
FIG. 1 is a block diagram showing a schematic configuration of an equipment control system 1 according to Embodiment 1 of the present invention. As shown in FIG. 1, the facility control system 1 according to the present embodiment includes a power prediction device 100 and a facility control device 200.

  The power prediction apparatus 100 acquires the measurement data from the power measurement system 2 that measures the power consumed by the equipment 3 and is consumed by the equipment 3 per unit time including the current time based on the acquired measurement data. Predict the amount of power A detailed configuration of the power prediction apparatus 100 will be described later.

  In the present embodiment, the power prediction apparatus 100 acquires measurement data representing power consumption at a cycle of 1 minute from the power measurement system 2, and the power consumed by the equipment 3 per 30 minutes including the current time. Is to be predicted.

  The facility control device 200 controls the operation of the facility device 3 based on the amount of power predicted by the power prediction device 100 so that the amount of power consumed per unit time does not exceed a predetermined value. I do. The detailed configuration of the equipment control device 200 will be described later.

  Next, the configuration of the power prediction apparatus 100 will be described.

  FIG. 2 is a schematic block diagram illustrating a hardware configuration of the power prediction apparatus 100. As shown in FIG. 2, the power prediction apparatus 100 includes a control unit 110, a ROM (Read Only Memory) 120, a RAM (Random Access Memory) 130, a storage unit 140, an input unit 150, and an output unit 160. , An external interface (I / F) 170 and a timer unit 180.

  The control part 110 is comprised from CPU (Central Processing Unit), for example, and controls the whole electric power prediction apparatus 100. FIG. For example, the control unit 110 executes a power prediction process. Details of the power prediction process will be described later.

  The ROM 120 is a non-volatile memory that stores a program and the like for the control unit 110 to control the entire power prediction apparatus 100. For example, the ROM 120 stores a program for executing power prediction processing.

  The RAM 130 is a volatile memory for temporarily storing information generated by the control unit 110 and data necessary for generating the information.

  The storage unit 140 includes a storage device such as a hard disk drive. Specifically, the storage unit 140 stores a power information DB (database) 141, a parameter setting DB 142, and a prediction means evaluation information DB 143.

  The input unit 150 is connected to an input device 300 such as a mouse or a keyboard for the user to operate the power prediction apparatus 100, and receives various types of information input by the user.

  The output unit 160 is connected to an output device 400 such as a liquid crystal display. The output unit 160 outputs various information output by the control unit 110 to the output device 400.

  The external interface 170 is connected to a drive device (not shown) for reading from and writing to a removable disk such as a CD (Compact Disc). Further, it may be connected to a device for connecting to an external network (for example, the Internet, a LAN (Local Area Network), etc.) or a printer for printing information generated by the control unit 110. Furthermore, the external interface 170 is communicably connected to the power measurement system 2 and the facility control device 200.

  The timekeeping unit 180 measures the current date and time, and includes a timer, for example.

  Next, the functional configuration of the control unit 110 of the power prediction apparatus 100 will be described.

  FIG. 3 is a schematic block diagram illustrating a functional configuration of the control unit 110 of the power prediction apparatus 100 according to the first embodiment of the present invention. As illustrated in FIG. 3, the control unit 110 functions as a power information storage control unit 111, a power prediction unit 112, a prediction unit evaluation unit 113, and a current predicted power amount acquisition unit 114.

  The power information storage control unit 111 acquires and updates data stored in the power information DB 141.

  Here, the power information DB 141 stores power history data 141a, past prediction data 141b, and current prediction data 141c. The power information DB 141 may be configured from a database having a function of accumulating past real-time data and summing up as necessary.

  The power information storage control unit 111 acquires, for example, an integrated value of power consumption for one minute or total power consumption from the past from the power measurement system 2 in a cycle of one minute, and stores it in the power information DB 141 as power history data 141a. Store.

  In addition, the power information storage control unit 111 extracts the power history data of the latest past unit time from the power history data 141a and updates it as past prediction data 141b. For example, when the current time is 11: 3, the power information storage control unit 111 performs past prediction based on the power history data 141a between 10:30 and 11:00, which is the last 30 minutes. The use data 141b is updated. The power information storage control unit 111 does not need to update the past prediction data 141b when the facility control device 200 determines that the demand control has been performed in the unit time in the latest past. .

  Further, the power information storage control unit 111 extracts the power history data from the start time of the unit time including the current time to the current time from the power history data 141a, and updates it as the current prediction data 141c. For example, when the current time is 11: 3, the current prediction data 141c is updated with the power history data 141a for 3 minutes from 11:00 to 11: 3.

  Further, the power information storage control unit 111 stores the past prediction data 141b and the current prediction data 141c in the power information DB 141 in a data format necessary for prediction according to a plurality of prediction means described later.

  In addition, when it is necessary to obtain the power consumption of the equipment 3 from a plurality of measurement data, the power information storage control unit 111 performs an appropriate calculation from the plurality of acquired measurement data, so that the past prediction data 141b. And current prediction data 141c is generated.

  The power predicting unit 112 includes a past predicted power amount calculating unit 112a, a parameter setting unit 112b, and a current predicted power amount calculating unit 112c.

  For each of a plurality of prediction means for calculating a power amount predicted to be consumed in a unit time, the past prediction power amount calculation unit 112a and the past prediction data 141b for each of a plurality of time points in the past unit time. Using the prediction means, a past predicted power amount that is predicted to be consumed in the past unit time is calculated. The past predicted power amount calculation unit 112a acquires the actual power consumption amount at the end of the past unit time (hereinafter referred to as the actual power amount), and an error between the calculated past predicted power amount and the actual power amount. Is calculated.

  Here, the past predicted power amount represents the amount of power predicted to be consumed in the unit time calculated at a certain point in time in the past unit time. Specifically, the past prediction power amount calculation unit 112a uses a plurality of prediction units for each of the unit time start 10 minute point and the 15 minute point in the past unit time 30 minutes from which the past prediction data 141b is acquired. To calculate the past predicted electric energy. Further, the past predicted power amount calculation unit 112a acquires the actual power amount of the past unit time from which the past prediction data 141b is acquired from the power history data 141a of the power information DB 141. Then, the past predicted power amount calculation unit 112a calculates an error of the past predicted power amount from the acquired past predicted power amount and the actual power amount.

  In the present embodiment, when calculating the past predicted power amount using a plurality of prediction means, the past predicted power amount for the same past unit time is calculated for each of the plurality of prediction means as the past unit time. And However, the past predicted power amount in a past unit time that is different for each of the plurality of prediction means may be calculated. In this case, the past predicted power amount calculation unit 112a acquires the actual power amount in the unit time for each unit time for which the past predicted power amount has been calculated from the power history data 141a, and the past predicted power amount and the actual power amount. And the error is calculated.

  Next, a specific example of a plurality of prediction means will be described.

  As an example of the prediction means, there is a method of linear prediction using power history data within a unit time (hereinafter referred to as prediction means A). Hereinafter, a method for predicting the amount of power by the present predicting means will be described with reference to FIG. FIG. 4 is a diagram showing the time transition of the integrated power consumption consumed up to the lapse of 15 minutes in the unit time of 30 minutes. For example, when the power consumption amount at the end of the unit time is predicted at the time when 15 minutes have elapsed, how many minutes of data before the 15 minutes are averaged is set as a parameter (time T). Then, by calculating the average power for the set parameter (time T) and calculating the increase ΔP of the remaining 15 minutes from the average power, it is predicted that the unit time will be consumed for 30 minutes. A predicted electric energy can be obtained. As another example of the predicting means, there is a method using a moving average value (hereinafter referred to as predicting means B) instead of the average power in the example of the predicting means.

  In addition, as another example of the predicting means, a method of predicting by adding correction of a shift as an absolute value based on past data having high correlation, such as an increase tendency similar to past power history data ( Hereinafter, there is a prediction means C). Hereinafter, a method for predicting the amount of power by the present prediction means will be described with reference to FIG. FIG. 5 is a diagram illustrating a time transition of the integrated power consumption consumed up to the lapse of 15 minutes out of a unit time of 30 minutes and a plurality of comparison data. As the comparison data, an average value or a pattern of past representative data such as the same day of the week or time of the same day, or data on a specific date can be used. In the example shown in FIG. 5, the correlation between the predicted unit time and the three comparison data SD1 to SD3 on the same day of the week and the accumulated power consumed up to 15 minutes has been obtained by statistical processing, and the correlation Based on SD3, which is high data, the amount of increase from the 15-minute point is regarded as the same, or is assumed to increase at the same ratio, and the predicted power amount is calculated. Similarly, prediction may be made based on weekday data samples having similar weather conditions.

  Returning to FIG. 3, the parameter setting unit 112 b sets one or a plurality of parameter candidates included in each prediction unit when the past prediction power amount is calculated by the plurality of prediction units. The parameter setting unit 112b performs prediction control for finding the optimum solution of each prediction unit based on the comparison between the calculated past predicted power amount and the actual power amount while sequentially updating the parameter candidates. Then, the parameter setting unit 112b sets the parameter candidate with the highest accuracy as a parameter for calculating the past predicted power amount using the prediction unit.

  The current predicted power amount calculation unit 112c is predicted to be consumed in the unit time including the current time using the current prediction data 141c and one or more prediction units selected by the prediction unit evaluation unit 113. The current predicted electric energy is calculated.

  The prediction means evaluation unit 113 evaluates a plurality of prediction means and selects an optimal prediction means. Specifically, the prediction means evaluation part 113 is comprised from the past prediction electric energy acquisition part 113a and the prediction means selection part 113b.

  For each of a plurality of prediction means, the past predicted power amount acquisition unit 113a determines the past predicted power amount from the power prediction unit 112 and the error between the past predicted power amount and the actual power amount for each of a plurality of time points in the past unit time. And is stored in the prediction means evaluation information DB 143.

  Here, data stored in the prediction means evaluation information DB 143 will be described. FIG. 6 shows an example of data stored in the prediction means evaluation information DB 143. The prediction means evaluation information DB 143 stores information for selecting one or more prediction means that are optimal for calculating the current predicted power amount among the plurality of prediction means. Specifically, as shown in FIG. 6, the prediction means evaluation information DB 143 includes, for each of a plurality of prediction means, a predicted value at 10 minutes, an error 1, a predicted value at 15 minutes, an error 2, and an average error. And the information indicating the error tendency are stored in association with each other.

  The 10 minute hour predicted value represents the past predicted power amount of the unit time at the 10 minute point of the past unit time, calculated using the corresponding prediction means.

  Error 1 is the difference between the predicted value at 10 minutes and the actual power amount in the past unit time, and is calculated as error 1 = (actual power amount−predicted value at 10 minutes) / actual power amount.

  The 15 minute hour predicted value represents the past predicted power amount of the unit time at the 15 minute point of the past unit time, calculated using the corresponding prediction means. Error 2 is the difference between the predicted value at 15 minutes and the actual power consumption in the past unit time, and is calculated as error 2 = (actual power amount−predicted value at 15 minutes) / actual power amount. .

  The average error is an average value of the corresponding error 1 and error 2.

  The error tendency is information indicating the tendency of the time transition of the error between the past predicted power amount calculated using the corresponding prediction means and the actual power amount. Specifically, the error decreases with time (error 1> error 2), that is, when the accuracy of the past predicted power amount increases, “High” is stored as the error tendency. If the error does not change over time (error 1 = error 2), that is, if the accuracy of the past predicted power amount does not change, “Stay” is stored as the error tendency. Further, when the error increases with the passage of time (error 1> error 2), that is, when the accuracy of the past predicted power amount decreases, “Down” is stored as the error tendency. When error 1 = error 2, the error tendency is “Stay”. However, when | (error 1−error 2) | <0.01, the error tendency may be “Stay”. A certain range may be given to the determination of “Stay”.

  Returning to FIG. 3, the prediction means selection unit 113b evaluates the prediction means based on the time transition of the error acquired for each of the plurality of prediction means by the past prediction power amount acquisition unit 113a, and calculates the current prediction power amount. Select the prediction means to use.

  Specifically, the prediction means selection unit 113b is configured to predict the error in which the error tends to converge with the passage of time based on the time transition of the error corresponding to the plurality of prediction means stored in the prediction means evaluation information DB 143. Is used to select the prediction means used to calculate the current predicted power amount. And the prediction means selection part 113b notifies the selected prediction means to the present prediction electric energy acquisition part 114. FIG.

  The current predicted power acquisition unit 114 acquires the current predicted power amount calculated by the power prediction unit 112 using the prediction unit selected by the prediction unit evaluation unit 113. Then, the current predicted power amount acquisition unit 114 outputs the acquired current predicted power amount and information indicating the error to the facility control device 200.

  Specifically, the current predicted power acquisition unit 114 uses the current prediction data 141c and the prediction unit selected by the prediction unit selection unit 113b in the prediction stage of 5 to 15 minutes within the current unit time. Then, the current predicted power amount is acquired from the power prediction unit 112, and the acquired current predicted power amount and the assumed error are output to the equipment control device 200. The current predicted power amount acquisition unit 114 may update, for example, every minute as a cycle for updating the current predicted power amount, or may be updated at a cycle according to each prediction unit.

  In addition, when a plurality of prediction means are selected by the prediction means selection section, the current prediction power amount acquisition section 114 gives an instruction from the prediction means selection section (average, priority with the larger prediction value error, prediction value error). Based on the smaller priority), the predicted power is obtained by processing a plurality of predicted values and is output to the equipment control apparatus 200.

  Next, the configuration of the equipment control device 200 will be described.

  FIG. 7 is a schematic block diagram illustrating a hardware configuration of the facility control apparatus 200. As shown in FIG. 7, the equipment control apparatus 200 includes a control unit 210, a ROM 220, a RAM 230, a storage unit 240, an input unit 250, an output unit 260, an external interface (I / F) 270, Part 280.

  The control part 210 is comprised from CPU (Central Processing Unit), for example, and controls the whole equipment control apparatus 200. FIG.

  The ROM 220 is a non-volatile memory that stores a program and the like for the control unit 210 to control the entire facility control apparatus 200.

  The RAM 230 is a volatile memory for temporarily storing information generated by the control unit 210 and data necessary for generating the information.

  The storage unit 240 includes a storage device such as a hard disk drive.

  The input unit 250 is connected to an input device 500 such as a mouse or a keyboard for the user to operate the equipment control device 200, and receives various types of information input by the user.

  The output unit 260 is connected to an output device 600 such as a liquid crystal display. The output unit 260 outputs various information output by the control unit 210 to the output device 600.

  The external interface 270 is connected to a drive device (not shown) or the like for reading from and writing to a removable disk such as a CD (Compact Disc). Further, it may be connected to a device for connecting to an external network (for example, the Internet, a LAN (Local Area Network), etc.) or a printer for printing information generated by the control unit 210. Furthermore, the external interface 270 is communicably connected to the power measurement system 2, the power prediction apparatus 100, and the equipment 3.

  The time measuring unit 280 measures the current date and time, and includes, for example, a timer.

  Next, a functional configuration of the control unit 210 of the facility control apparatus 200 will be described.

  FIG. 8 is a schematic block diagram illustrating a functional configuration of the control unit 210 of the facility control apparatus 200 according to the first embodiment of the present invention. As illustrated in FIG. 8, the control unit 210 functions as a power management unit 211 and an equipment control execution unit 212.

  The power management unit 211 compares the current predicted power amount acquired from the power prediction apparatus 100 with a predetermined management value, determines whether or not demand control is necessary, and performs facility control when it is determined that demand control is necessary. The execution unit 212 is instructed to execute demand control.

  Specifically, the power management unit 211 determines whether or not the current predicted power amount exceeds a predetermined management value. If it is determined that the current predicted power amount exceeds the management value, the power management unit 211 reduces the amount of current predicted power amount that exceeds the management value (excess power amount) and the error information. Calculate the amount of power. Then, the power management unit 211 determines, for example, whether or not demand control needs to be performed when 10 minutes or 15 minutes have elapsed according to the elapsed time. The excess power amount and the reduced power amount from that time are notified.

  When receiving a demand control instruction from the power management unit 211, the facility control execution unit 212 controls the operation of each facility device so that the amount of power consumed by the facility device 3 is smaller than the management value.

  Specifically, the facility control execution unit 212 determines the facility device from the reduced power amount notified from the power management unit 211, the remaining time until the end of the unit time including the current time, the priority of the facility device 3, and the like. 3 determines an operation setting that saves power, and outputs control data for controlling the equipment 3. The control data is converted at the control interface of each equipment 3 and actual control is executed.

  Next, operation | movement of the equipment control system 1 in this embodiment is demonstrated.

  First, the flow of power prediction processing executed by the power prediction apparatus 100 will be described. FIG. 9 is a flowchart illustrating an example of the flow of power prediction processing executed by the power prediction apparatus 100. The power prediction process illustrated in FIG. 9 is started when, for example, the current time measured by the time measuring unit 180 is a predetermined time (for example, 0 minutes and 30 minutes every hour).

  First, the power information storage control unit 111 waits for 2 minutes (step S101).

  Next, the power information storage control unit 111 refers to, for example, a flag stored in the RAM 130 that indicates whether the demand control is being executed by the facility control apparatus 200, and the demand control is performed in the last 30 minutes. It is determined whether or not it has been (step S102). If the power information storage control unit 111 determines that demand control has been performed (step S102; Yes), after clearing the flag, the process proceeds to step S104.

  When the power information storage control unit 111 determines that demand control is not performed (step S102; No), the past prediction data 141b is converted into the power consumption amount for the last 30 minutes stored in the power history data 141a. Update (step S103).

  Next, the power information storage control unit 111 acquires data for prediction means (step S104). The data for prediction means is data used when, for example, the amount of electric power is predicted based on the tendency of the average value for one month on the same day and at the same time as in the prediction means C described above. Specifically, the power information storage control unit 111 extracts data for the prediction target unit time of 30 minutes, the same month, the same day, and the same time (30 minutes) from the power history data 141a, and holds the data as prediction means data. .

  Next, the prediction means evaluation part 113 performs a prediction means evaluation process (step S105).

  FIG. 10 shows a flowchart of the prediction means evaluation process executed by the prediction means evaluation unit 113.

  First, the past predicted power acquisition unit 113a sets 1 to the counter m (step S201).

  Next, the past predicted power acquisition unit 113a refers to the prediction means evaluation information DB 143 and identifies one prediction means for which the past predicted power has not been acquired (step S202).

  Next, the power prediction unit 112 performs a past predicted power amount calculation process using the prediction means (hereinafter referred to as prediction means X) identified in step S201 (step S203).

  FIG. 11 shows a flowchart of past predicted power amount calculation processing executed by the power prediction unit 112.

  First, the past predicted power amount calculation unit 112a sets 1 to the counter n (step S301).

  Next, the parameter setting unit 112b sets the parameter Y_n necessary for the calculation of the past predicted power amount for the prediction unit X (step S302).

  Next, the past predicted power amount calculation unit 112a uses the prediction unit X and the past prediction data 141b, and the past predicted power amount at the time of 10 minutes from the start of the past unit time when the past prediction data 141b is acquired. P10_n is calculated (step S303).

  Next, the past predicted power amount calculation unit 112a obtains the past unit time actual power amount from which the past prediction data 141b has been obtained from the power history data 141a, and the past predicted power amount P10_n calculated in step S302, An error 1_n from the actual power consumption is calculated (step S304).

  Next, the past predicted power amount calculation unit 112a uses the prediction unit X and the past prediction data 141b, and the past predicted power amount at the time of 15 minutes from the start of the past unit time when the past prediction data 141b is acquired. P15_n is calculated (step S305).

  Next, the past predicted power amount calculation unit 112a calculates an error 2_n between the past predicted power amount P15_n calculated in step S304 and the actual power amount (step S306).

  Next, the past predicted power amount calculation unit 112a determines whether or not the counter n is 1 (step S307). If the counter n is determined to be 1 (step S307; Yes), the parameter Y_n, 10 minutes has elapsed. The past predicted power amount P10_n at the time point, the error 1_n, the past predicted power amount P15_n at the time point after 15 minutes, and the error 2_n are respectively set to the parameter Y, the past predicted power amount P10 at the time point after 10 minutes, the error at the time point after 15 minutes. The past predicted power amount P15 and the error 2 are held (step S308).

  When it is determined that the counter n is not 1 (step S307; No), the past predicted power amount calculation unit 112a determines whether or not the error 2_n is smaller than the held error 2 (step S309). When it is determined that the error 2_n is not smaller than the error 2 (step S309; No), the past predicted power amount calculation unit 112a advances the process to step S311.

  When it is determined that the error 2_n is smaller than the error 2 (step S309; Yes), the past predicted power amount calculation unit 112a stores the parameter Y that is held, the past predicted power amount P10 when 10 minutes have elapsed, and the errors 1, 15 The past predicted power amount P15 and the error 2 when the minute has elapsed are updated to the parameter Y_n, the past predicted power amount P10_n and the error 1_n when the 10 minute point has elapsed, the past predicted power amount P15_n and the error 2_n when the 15 minute point has elapsed, respectively. (Step S310).

  The parameter setting unit 112b determines whether or not parameters have been set for all of the preset parameter ranges for the prediction unit X (step S311). If the parameter setting unit 112b determines that no parameter is set in all parameter ranges (step S311; No), the process returns to step S302.

  When it is determined that parameters have been set in all parameter ranges (step S311; Yes), the parameter setting unit 112b uses the held parameter Y as a parameter used for calculating the predicted power amount in the prediction unit X. Store in the parameter setting DB 142 (step S312).

  As a result of the above processing, the data stored last becomes the most accurate data, and the data is output as a prediction result by the prediction means X to the past predicted power acquisition unit 113a. In addition, since the parameter setting when the data is calculated is an optimum parameter, the parameter setting is stored in the parameter setting DB 142 as a parameter of the prediction means X.

  Returning to FIG. 10, the past predicted power amount acquisition unit 113a obtains the past predicted power amount at the time when 10 minutes have elapsed, the error 1, the past predicted power amount at the time when 15 minutes have elapsed, acquired by the past predicted power amount calculation process in step S203, And error 2 are stored in the prediction means evaluation information DB 143 in association with the prediction means specified in step S202 (step S204).

  Next, the past predicted power acquisition unit 113a refers to the prediction means evaluation information DB 143 and determines whether or not the past prediction power has been acquired for all prediction means (step S205). When it is determined that the past predicted power amount has not been acquired for all prediction means (step S205; No), the past predicted power amount acquisition unit 113a returns the process to step S202, and the past predicted power amount for all prediction means. Until the amount is acquired, the processes in steps S202 to S205 are repeated.

  When it is determined that the past predicted power amount has been acquired for all the prediction means (step S205; Yes), the prediction means selection unit 113b, for each of the plurality of prediction means, The average error and the error tendency are acquired from the past predicted power amount at the time when the error 1 and 15 minutes have passed and the error 2 and stored in the prediction means evaluation information DB 143 (step S206).

  Next, the prediction means selection part 113b performs a prediction means selection process (step S207).

  12 and 13 show flowcharts of the prediction means selection process executed by the prediction means selection unit 113b.

  First, the prediction means selection unit 113b refers to the prediction means evaluation information DB 143 and determines whether or not there is one prediction means whose average error is within a predetermined value (for example, 3%) (step S401).

  When it is determined that there is one prediction means with an average error within a predetermined value (step S401; Yes), the prediction means selection unit 113b determines one prediction means with an average error within a predetermined value as the current predicted power amount. Is selected as a predicting means used for the calculation (step S402). And the prediction means selection part 113b advances a process to step S106 of FIG.

  That is, the prediction means selection unit 113b first extracts a prediction means with a small average error as a prediction means having a generally good prediction accuracy. At this time, if there is only one prediction means with a small average error, the prediction means is selected as a prediction means used for calculation of the current predicted power amount.

  When it is determined that there is not one prediction means with an average error within a predetermined value (step S401; No), the prediction means selection unit 113b refers to the prediction means evaluation information DB 143, and the prediction means with an average error within a predetermined value is detected. It is determined whether or not there are a plurality (step S403).

  When it is determined that there are a plurality of prediction means having an average error within a predetermined value (step S403; Yes), one prediction means with a corresponding error tendency of “High” or “Stay” is selected from the plurality of prediction means. It is determined whether or not (step S404).

  When it is determined that there is one prediction means whose error tendency is “High” or “Stay” (step S404; Yes), the prediction means selection unit 113b calculates the current prediction power amount. Is selected as the prediction means to be used (step S405). And the prediction means selection part 133b advances a process to step S106 of FIG.

  When it is determined that there is not one prediction means whose error tendency is “High” or “Stay” (step S404; No), the prediction means selection unit 113b has a prediction means whose error tendency is “High” or “Stay”. It is determined whether or not there is a plurality (step S406). When it is determined that there are not a plurality of prediction means whose error tendency is “High” or “Stay”, that is, there is no prediction means whose error tendency is “High” or “Stay” (Step S406; No), the prediction means selection The unit 113b advances the process to step S411 illustrated in FIG.

  When it is determined that there are a plurality of prediction means having an error tendency of “High” or “Stay” (step S406; Yes), the prediction means selection unit 113b selects the prediction means with the smallest error 2 as the current prediction power amount. It is determined as a prediction means used for calculation (step S407).

  Here, since the power consumption is a result of integration of time series data, the accuracy tends to improve with time. Therefore, when the error tendency is “Down”, there is a possibility that the prediction of the past predicted power amount at the time when 10 minutes have elapsed by the power prediction unit 112 is doubtful. Therefore, the prediction means whose error tendency is “Down” is excluded from the selection target in steps S404 and S406. In addition, when there are a plurality of prediction means having an error tendency of “High” or “Stay”, since a plurality of prediction means can perform highly accurate prediction, the time for simply starting demand control (here Then, the prediction means with the smallest error 2 which is an error at the time point of 15 minutes) is selected.

  Further, when it is determined that there are not a plurality of prediction means whose average error is within the predetermined value, that is, there is no prediction means whose average error is within the predetermined value (step S403; No), as shown in FIG. 113b determines whether or not there is one prediction means with error 2 equal to or less than a predetermined value (step S408).

  When it is determined that the number of prediction means for which the error 2 is equal to or less than the predetermined value is one (step S408; Yes), the prediction means selection unit 113b uses the one prediction means for calculating the current prediction power amount. (Step S409). And the prediction means selection part 113b advances a process to step S106 of FIG.

  When it is determined that there is not one prediction means with the error 2 equal to or less than the predetermined value (step S408; No), the prediction means selection unit 113b determines whether there are a plurality of prediction means with the error 2 equal to or less than the predetermined value ( Step S410).

  When it is determined that there are a plurality of prediction means with the error 2 equal to or less than the predetermined value (step S410; Yes), the prediction means selection unit 113b advances the process to step S404.

  When it is determined that there are not a plurality of prediction means for which the error 2 is equal to or less than the predetermined value, that is, there is no prediction means for which the error 2 is equal to or less than the predetermined value (Step S410; No), the prediction means selection unit 113b has the counter m greater than 3. It is determined whether it is smaller (step S411).

  When it is determined that the counter m is not smaller than 3 (step S411; No), the prediction unit selection unit 113b determines a prediction unit with the smallest plus-side error 2 and a prediction unit with the smallest minus-side error 2. Then, it is selected as a prediction means used for calculation of the current predicted power amount (step S412). And the prediction means selection part 113b advances a process to step S106 of FIG.

  When it is determined that the counter m is smaller than 3 (step S411; Yes), the prediction means selection unit 113b resets the past prediction data 141b (step S413). Specifically, the prediction means selection unit 113b selects a previous unit time for the unit time for calculating the past predicted power amount, acquires the power consumption amount of the unit time from the power history data 141a, and performs the past prediction. It stores as data for data 141b. Moreover, the prediction means selection part 113b may reset the average data of the preset power consumption as the past prediction data 141b.

  Next, the prediction means selection unit 113b increments the counter m (step S414), and proceeds to step S202 in FIG.

  That is, when the power usage conditions are complex and lack regularity, the prediction accuracy does not increase. In this case, the prediction means is selected based on the error 2 when the demand control is executed, that is, when 15 minutes have elapsed. In the above flow chart, prediction means having an error 2 of a predetermined value (for example, 3%) or less are selected in the order of good system. If the desired accuracy cannot be obtained even by determining only the error 2, an older unit time is selected as the unit time for acquiring the past prediction data 141b, and the power prediction unit 112 again performs the past prediction in the unit time. Calculate the amount of power. If the error does not converge even if this process is repeated a plurality of times, the prediction means with the smallest error 2 is selected.

  Returning to FIG. 9, the current prediction power amount acquisition unit 114 uses the prediction unit selected by the prediction unit evaluation process in step S <b> 105 and the current prediction data 141 c by the power prediction unit 112. Is acquired (step S106).

  Specifically, the current prediction power amount acquisition unit 114 instructs the current prediction power amount calculation unit 112c of the power prediction unit 112 to calculate the current prediction power amount by the prediction unit selected by the prediction unit selection unit 113b. To do. The current predicted power amount calculation unit 112c calculates the current predicted power amount based on the selected prediction means, the parameter stored in the parameter setting DB 142, and the current prediction data 141c, and obtains the current predicted power amount. Output to the unit 114. Then, the current predicted power amount acquisition unit 114 outputs the acquired current predicted power amount to the equipment control device 200.

  When a plurality of prediction means are selected by the prediction means evaluation unit 113, the current prediction power amount acquisition unit 114 acquires the current prediction power amount based on a predetermined condition set in advance. As the predetermined condition, for example, the current predicted power amount is obtained from the average value of the current predicted power amounts calculated using a plurality of prediction means, or the current prediction is performed using a prediction means having a large or small error. The condition is that the amount of electric power is acquired.

  Next, the current predicted power acquisition unit 114 determines whether there is a request for output of the current predicted power from the facility control device 200 (step S107). If it determines with the output request | requirement of the present prediction electric energy having been received (step S107; Yes), the present prediction electric energy acquisition part 114 will output the present prediction electric energy and prediction error to the equipment control apparatus 200 (step S108).

  When it is determined that there is no output request for the current predicted power amount (step S107; No), the current predicted power amount acquisition unit 114 determines whether the power history data 141a has been updated (step S109). When determining that the power history data 141a has been updated, the current predicted power acquisition unit 114 returns the process to step S106.

  When it is determined that the power history data 141a has not been updated (step S109; No), the current predicted power acquisition unit 114 determines whether demand control is being executed by the equipment control device 200 (step S110). When it is determined that the demand control is being executed by the equipment control device 200 (step S110; Yes), the current predicted power acquisition unit 114 sets a flag indicating whether the demand control is being executed by the equipment control device 200, A value indicating that demand control is being executed is set (step S111).

  If it determines with demand control not being performed by the equipment control apparatus 200 (step S110; No), the present prediction electric energy acquisition part 114 will determine whether 30 minutes passed from the start of an electric power prediction process (step). S112). If it is determined that 30 minutes have not elapsed since the start of the power prediction process, the current predicted power acquisition unit 114 returns the process to step S106. If it is determined that 30 minutes have elapsed since the start of the power prediction process, the power prediction process ends.

  As described above, the equipment control system 1 according to the present embodiment selects a prediction method that is automatically optimized for the power consumption characteristics of the equipment 3 from a plurality of prediction means, and uses the selected prediction means. Therefore, it is possible to predict the power consumption with higher accuracy than the conventional prediction method.

  In addition, since the prediction error can be grasped, when the prediction error is small, the equipment control device 200 can adjust the capacity of the equipment 3 by the gentle demand control at an early stage, thereby minimizing environmental deterioration. You can stay.

  Moreover, when demand control is performed, since the data used for prediction are not updated, the problem that unexpected electric power amount becomes small does not generate | occur | produce.

(Embodiment 2)
Next, Embodiment 2 of the present invention will be described.
FIG. 14 is a schematic block diagram illustrating a functional configuration of the control unit 210 of the facility control apparatus 200 according to the second embodiment of the present invention. As illustrated in FIG. 14, the control unit 210 of the facility control apparatus 200 according to the second embodiment further includes a device control information acquisition unit 213, a power consumption information acquisition unit 214, a device power consumption analysis unit 215, and a correction information generation unit 216. Function as.

  The device control information acquisition unit 213 acquires control information indicating the control state of the facility device 3 or control state change information from the facility control execution unit 212. Here, the control state change information is, for example, information indicating a change in the set temperature when the equipment 3 is an air conditioner.

  The power consumption information acquisition unit 214 sequentially acquires the power consumption amount for each facility device 3 or each group of the facility devices 3 as time series data.

  The device power consumption analysis unit 215 controls the facility device 3 based on the control state of the facility device 3 acquired by the device control information acquisition unit 213 and the power consumption amount acquired by the power consumption information acquisition unit 214. Analyzes how much the power consumption changes when. Specifically, when the control state of the equipment device 3 changes, the device power consumption analysis unit 215 elapses after the difference in power consumption per minute and the control state change as time passes. The average value of the difference values for the calculated time is calculated, and the calculated difference value and average value are stored in the device power consumption reduction DB 241 in association with the equipment device 3 whose control state has changed. However, if there is already accumulated data indicating a change in the control state of the same pattern and a change in power consumption at that time, the device power consumption analysis unit 215 determines that the addition of data is unnecessary and does not perform the additional processing. .

  Here, FIG. 15 shows an example of data stored in the device power consumption reduction DB 241. The device power consumption reduction DB 241 illustrated in FIG. 15 stores information representing facility device information, change information, a reduction value per minute, and an average value for elapsed time.

  The equipment information represents the equipment 3 whose control state has changed. Specifically, in the device power consumption reduction DB 241 illustrated in FIG. 15, an “air conditioning device group” that is a set of a plurality of air conditioning devices is stored as facility device information. Moreover, change information represents the change of the control state of the equipment 3 represented by equipment information. Specifically, in the device power consumption reduction DB 241 illustrated in FIG. 15, the change information represents that the control state has changed so that the set temperature is changed from 22 ° C. to 20 ° C.

  The time represents the time when the control state of the equipment device 3 has changed and the time every minute thereafter. Specifically, the time “10:00” represents the time when the control state of the equipment 3 has changed, and for each minute after that time, the reduction value per minute and the average value for the elapsed time are the equipment. Calculated by the power consumption analysis unit 215 and stored in the device power consumption reduction DB 241.

  The reduction value per minute is the amount of power consumed in one minute from one minute before the corresponding time to the corresponding time, and the amount of power consumed in one minute from the corresponding time to one minute after the corresponding time. And represents the reduced power consumption. For example, when the device power consumption analysis unit 215 determines that the control state of the facility device 3 has changed at “10:00”, the power consumption amount from “10:00” to “10:01” and “10: A reduction value “1” between the power consumption amounts from “01” to “10:02” is calculated, and stored as a reduction value per minute corresponding to the time “10:01”.

  The average value for the elapsed time represents a value obtained by averaging, for 30 minutes, the difference in power consumption before and after the change of the control state for the time that has elapsed since the control state of the equipment device 3 has changed. For example, the difference in power consumption from “10:00” to “10:03” in which the control state has changed is one minute corresponding to “10:01”, “10:02”, and “10:03”. The total of the reduction values for each, that is, 2.5 (kw). Therefore, the average value for the elapsed time corresponding to “10:03” is calculated as 2.5 / 30 = 0.033 (kw).

  The device power consumption analysis unit 215 calculates a reduction value for each minute and an average value for the elapsed time for a predetermined time (for example, 15 minutes) after the control state of the equipment device 3 changes, thereby reducing device power consumption. Store in the DB 241. Then, when the facility control execution unit 212 receives a demand control instruction from the power management unit 211 and controls the operation of the facility device 3, the facility control execution unit 212 saves the facility device 3 based on the data stored in the device power consumption reduction DB 241. Determine the operation setting to be used as electric power. For example, if the facility control execution unit 212 determines that a reduction in power consumption of about 0.083 kw is necessary 3 minutes before 30 minutes of unit time has elapsed, refer to the device power consumption reduction DB 241. The operation of the equipment 3 is controlled so that the control state is changed similarly to the change of the control state at the time “10:00”.

  The correction information generation unit 216 generates correction information used for correcting the electric energy calculated by the prediction unit, based on the control state of the equipment 3 and the data stored in the standard power consumption DB 242. The standard power consumption DB 242 stores in advance the control state of the equipment 3 and the standard power consumption of the equipment 3 in the control state. The correction information generation unit 216 refers to the standard power consumption DB 242, and uses the standard power consumption amount corresponding to the control state of the equipment 3 acquired by the device control information acquisition unit 213 and the power consumption information acquisition unit 214. The difference with the power consumption of the acquired equipment 3 is calculated. This difference is output to the power prediction unit 112 of the power prediction apparatus 100 as correction information used for correcting the power amount calculated by the prediction unit when the power consumption amount is predicted based on the control state of the equipment device 3. The The power prediction unit 112 captures the correction information as correction data such as a parameter of the prediction unit. Thereby, it is possible to predict the power consumption with high accuracy.

  With the above configuration, by calculating the device reduction power based on actual data, the demand power can be suppressed within the management value by a more reliable and optimal means. Therefore, it is possible to construct a system that can suppress environmental deterioration to a minimum and does not exceed demand power. Further, the accuracy of the prediction can be improved by feeding back the correction information based on the actual measurement data to the prediction means. In addition, by accumulating actual measurement data, it is possible to further improve the accuracy of both power prediction and demand control execution as the operating time elapses.

  As mentioned above, although embodiment of this invention was described, this invention is not limited by embodiment.

  For example, in the first and second embodiments, when estimating the power consumption of air conditioning equipment, lighting equipment, and OA equipment installed in a general office building as the equipment equipment 3, the air conditioning equipment, lighting equipment, and OA equipment respectively By using the optimal prediction method for each individual, air conditioning power parameters are optimized for air conditioning power, lighting power parameters are optimized for lighting power, etc., and overall prediction accuracy is improved. You may plan.

  Moreover, in Embodiment 1 and 2, although the aspect comprised from an apparatus different from the electric power prediction apparatus 100 and the equipment control apparatus 200 was demonstrated, the structure of the equipment control system 1 is not restricted to this. For example, the power prediction apparatus 100 may include the function of the facility control apparatus 200.

  In addition, the equipment control system 1 according to the present invention can be realized by using a normal computer system without depending on a dedicated system. For example, a program for executing the above operation is stored in a computer-readable recording medium (CD-ROM, MO, etc.) and distributed to a computer connected to a network, and the program is distributed to the computer system. You may comprise the equipment control system 1 which performs the above-mentioned process by installing.

  Further, the method for providing the program to the computer is arbitrary. For example, the program may be uploaded to a bulletin board (BBS) on a communication line and distributed to a computer via the communication line. The program may be transmitted by a modulated wave obtained by modulating a carrier wave with a signal representing the program, and a device that receives the modulated wave may demodulate the modulated wave to restore the program. Then, the computer activates this program and executes it in the same manner as other applications under the control of the OS. Thereby, a computer functions as the equipment control system 1 which performs the above-mentioned process.

DESCRIPTION OF SYMBOLS 1 Equipment control system, 100 Power prediction apparatus, 110 Control part, 111 Power information storage control part, 112 Power prediction part, 112a Past prediction electric energy calculation part, 112b Parameter setting part, 112c Current prediction electric energy calculation part, 113 Prediction means Evaluation unit, 113a Past predicted power acquisition unit, 113b Prediction means selection unit, 114 Current predicted power acquisition unit, 120 ROM, 130 RAM, 140 storage unit, 141 Power information DB, 141a Power history data, 141b Past prediction data , 141c Current prediction data, 142 Parameter setting DB, 143 Prediction means evaluation information DB, 150 input unit, 160 output unit, 170 external I / F, 180 timing unit, 200 equipment control device, 210 control unit, 211 power management unit 212 equipment control execution unit, 213 equipment control Information acquisition unit, 214 Power consumption information acquisition unit, 215 Device power consumption analysis unit, 216 Correction information generation unit, 220 ROM, 230 RAM, 240 storage unit, 241 Device power consumption reduction DB, 242 Standard power consumption DB, 250 input unit , 260 output unit, 270 external I / F, 280 timing unit, 300 input device, 400 output device, 500 input device, 600 output device, 2 power measurement system, 3 equipment

Claims (9)

A power prediction device that predicts the amount of power consumed by equipment,
For each of a plurality of prediction means for calculating the amount of power expected to be consumed per unit time, with the time that put in a past unit time, the actual amount of power consumed up to that point and the prediction means A past predicted power amount acquisition unit that acquires a past predicted power amount calculated to be consumed in the past unit time, calculated using
Based on the actual amount of power consumed in the calculated issued said past predicted power amount and the previous SL past unit time, from among the plurality of prediction unit, the prediction unit selector for selecting one or a plurality of prediction means When,
A current predicted power amount acquisition unit that acquires a current predicted power amount that is predicted to be consumed in a unit time including the current time, using the actual power amount consumed up to the present time and the selected one or more prediction means. When,
An electric power prediction apparatus comprising: The prediction means selection unit acquires, for each prediction means, an error between the calculated past predicted power amount and the actual power amount consumed in the past unit time, and the time transition of the error converges. Select one or more predictors that have been determined to be prone,
The power prediction apparatus according to claim 1. When the past predicted power amount is calculated by the plurality of prediction means, the parameters of the prediction means are sequentially updated, and the highest accuracy is obtained based on the comparison between the calculated past predicted power amount and the actual power amount. A parameter setting unit for setting a parameter as a parameter for calculating the past predicted power amount,
The power prediction apparatus according to claim 1 or 2, wherein At least one prediction means among the plurality of prediction means includes a time transition pattern of the electric energy acquired based on the actual electric energy consumed in the past unit time satisfying a predetermined condition, and consumption by a predetermined time point. Calculating the amount of power that is predicted to be consumed in a unit time including the predetermined time point based on a comparison with the time transition of the actual power amount
The power prediction apparatus according to claim 1, wherein the power prediction apparatus is a power prediction apparatus. By storing time-series data of the previous unit time as past data in the power data storage unit and overwriting and updating each time the unit time elapses, the past data of the most recent demand time can be retained in units of one minute. A power information storage controller capable of further,
The power prediction apparatus according to claim 1, wherein the power prediction apparatus is a power prediction apparatus. A power prediction device according to any one of claims 1 to 5,
A power management unit that compares the current predicted power amount with a predetermined management value, determines whether or not it exceeds, and calculates a reduced power amount based on the excess power amount when exceeding,
An equipment control execution unit that controls the operation of the equipment;
From the facility control execution unit, a device control information acquisition unit that acquires control information representing the control state of the facility device,
When the control state of the equipment device represented by the control information is changed, the change information indicating the change in the control state of the equipment device is associated with the amount of power reduced by the change in the control state of the equipment device. A device power consumption analysis unit to be stored in the device power consumption reduction storage unit,
With
The facility control execution unit refers to the device power consumption reduction storage unit so that the power consumption of the facility device in a unit time including the current time is reduced by at least the amount of reduced power. Control the operation of the
An equipment control system characterized by that. A standard power consumption storage unit that stores the standard power consumption of the equipment;
Based on the difference between the standard power consumption amount stored in the standard power consumption storage unit and the actual power amount, the power amount predicted to be consumed per unit time, calculated by the plurality of prediction means. A correction information generation unit that generates correction information used for the correction of
The equipment control system according to claim 6. A power prediction method for predicting the amount of power consumed by equipment,
For each of a plurality of prediction means for calculating the amount of power expected to be consumed per unit time, with the time that put in a past unit time, the actual amount of power consumed up to that point and the prediction means A past predicted power amount obtaining step for obtaining a past predicted power amount calculated to be consumed in the past unit time, calculated using
And calculated issued the last predicted power amount, on the basis on the actual amount of power consumed in the past unit time, the prediction unit selection step from selecting one or more prediction means among the plurality of prediction means When,
A current predicted power acquisition step of acquiring a current predicted power predicted to be consumed in a unit time including the current time using the actual power consumed until the current time and the selected one or more prediction means. When,
A power prediction method characterized by comprising: A computer that predicts the amount of power consumed by equipment
For each of a plurality of prediction means for calculating the amount of power expected to be consumed per unit time, with the time that put in a past unit time, the actual amount of power consumed up to that point and the prediction means A past predicted power acquisition means for acquiring a past predicted power calculated to be consumed in the past unit time, calculated using
And calculated issued the last predicted power amount, on the basis on the actual amount of power consumed in the past unit time, from among the plurality of prediction unit, the prediction unit selecting means for selecting one or more prediction means ,
Current predicted power amount acquisition means for acquiring a current predicted power amount predicted to be consumed in a unit time including the current time using the actual power amount consumed up to the present time and the selected one or more prediction means. ,
A program characterized by functioning as

Images