Electricity demand forecasting device and electricity demand forecasting method

The power demand forecasting device and method enhance prediction accuracy by using corrected values based on past consumption and actual measurements to adjust predicted power demand, addressing the inaccuracies in existing methods.

JP7595476B2Active Publication Date: 2024-12-06MITSUBISHI HEAVY IND LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
JP2021016684
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2021-02-04
Publication Date
2024-12-06
Estimated Expiration
2041-02-04

AI Technical Summary

Technical Problem

Existing power demand forecasting methods, such as those described in Patent Document 1, fail to accurately reflect the trend of power demand on the day of a Demand Response (DR) event, leading to potential large errors in predicted electricity demand.

Method used

A power demand forecasting device and method that calculates a first and second predicted value based on past power consumption, uses an actual measured value to correct the first predicted value when the deviation exceeds a tolerance threshold, and applies a corrected adopted value to refine the prediction.

Benefits of technology

Improves the accuracy of power demand forecasting by minimizing deviations between predicted and actual values, ensuring precise predictions during DR events.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007595476000001
    Figure 0007595476000001
  • Figure 0007595476000002
    Figure 0007595476000002
  • Figure 0007595476000003
    Figure 0007595476000003
Patent Text Reader

Abstract

To make it possible to improve prediction accuracy of power demand in an arbitrary time zone.SOLUTION: A power demand prediction device for predicting power demand of a user in a first time zone comprises: a prediction value calculation unit that on the basis of consumption power consumed by the user before the first time zone, calculates a first prediction value for power demand in the first time zone and a second prediction value for power demand in a second time zone before the first time zone; an actual value acquisition unit that acquires an actual value of consumption power consumed in the second time zone by the user; and a prediction value correction unit that when the absolute value of a correction reference value obtained by subtracting the actual value from the second prediction value is larger than a predetermined allowable error threshold, subtracts a correction adoption value from the first prediction value. The absolute value of the correction adoption value is smaller than the absolute value of the correction reference value.SELECTED DRAWING: Figure 2
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present disclosure relates to an electric power demand forecasting device and an electric power demand forecasting method. [Background technology]

[0002] As a means for balancing the demand and supply of electricity, there is a demand response (hereinafter referred to as DR) that encourages consumers to adjust their electricity demand (power consumption). In general, an incentive is given to consumers who adjust their power consumption in accordance with a DR event. In order to confirm whether a consumer has adjusted their power consumption in accordance with a DR event, the actual power consumption during the DR time period and the predicted value of the consumer's power demand in the absence of a DR event are required.

[0003] Patent Document 1 discloses that, from among a number of days prior to the implementation date of a DR event, a day with similar conditions to the implementation date of the DR event (for example, a day with a small difference in temperature) is selected, and a predicted value (baseline) of power demand is determined based on the power consumption on the selected day. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 6128624 Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the technology described in Patent Document 1, the power demand on the day a DR event is held is predicted from the power consumption on days prior to the day the DR event is held, so the trend in power demand on the day the DR event is held is not reflected, and the predicted value of power demand may contain a large error.

[0006] The present disclosure has been made in consideration of the above-mentioned problems, and aims to provide a power demand forecasting device and a power demand forecasting method that can improve the accuracy of forecasting power demand in any time period. [Means for solving the problem]

[0007] In order to achieve the above-mentioned object, the power demand forecasting device of the present disclosure is a power demand forecasting device that forecasts the power demand of a consumer in a first time period, and includes a forecast value calculation unit that calculates a first forecast value of the power demand in the first time period and a second forecast value of the power demand in a second time period prior to the first time period based on the power consumption consumed by the consumer in the period prior to the first time period, an actual measurement value acquisition unit that acquires an actual measurement value of the power consumption consumed by the consumer in the second time period, and a forecast value correction unit that subtracts a correction adoption value from the first forecast value if the absolute value of a correction standard value obtained by subtracting the actual measurement value from the second forecast value is greater than a predetermined allowable error threshold, and the absolute value of the correction adoption value is smaller than the absolute value of the correction standard value.

[0008] In order to achieve the above-mentioned object, the power demand forecasting method of the present disclosure is a power demand forecasting method for forecasting power demand of a consumer in a first time period, and includes the steps of: calculating a first predicted value of power demand in the first time period and a second predicted value of power demand in a second time period prior to the first time period based on power consumption consumed by the consumer prior to the first time period; acquiring an actual measured value of power consumption consumed by the consumer in the second time period; and, if the absolute value of a correction standard value obtained by subtracting the actual measured value from the second predicted value is greater than a predetermined allowable error threshold, subtracting a correction adoption value from the first predicted value, wherein the absolute value of the correction adoption value is smaller than the absolute value of the correction reference value. Effect of the Invention

[0009] According to the power demand forecasting device and power demand forecasting method disclosed herein, it is possible to improve the accuracy of forecasting power demand in any time period. [Brief description of the drawings]

[0010] [Figure 1] 1 is a schematic diagram illustrating a configuration of a power demand prediction system according to an embodiment. [Diagram 2] 1 is a schematic functional block diagram of a power demand prediction device according to an embodiment; [Diagram 3] 13 is a graph showing a first predicted value, a second predicted value, and a corrected first predicted value. [Figure 4] 13 is a diagram showing a case where a first predicted value is corrected with a correction adopted value and a case where the first predicted value is corrected with a correction reference value. FIG. [Diagram 5] 1 is a flowchart of a power demand forecasting method according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, a power demand forecasting device and a power demand forecasting method according to an embodiment of the present disclosure will be described with reference to the drawings. The embodiment shows one aspect of the present disclosure, and does not limit the present disclosure. The embodiment can be arbitrarily modified within the scope of the technical idea of ​​the present disclosure.

[0012] <Electricity demand forecasting system> FIG. 1 is a schematic diagram illustrating the configuration of a power demand prediction system 100 according to an embodiment. The power demand prediction system 100 according to the present disclosure is a system for predicting power demand of a consumer 102 in a demand response time period (first time period). In the present disclosure, the first time period is not limited to the demand response time period. In some embodiments, the power demand prediction system 100 predicts power demand of the consumer 102 in any time period other than the demand response time period.

[0013] Demand response (hereinafter, DR) is a means for balancing the demand and supply of electricity, in which the consumer 102 adjusts the demand for electricity (power consumption). The power supply business operator predicts the power demand of the consumer 102 during the DR time period, and calculates the amount of power consumption to be adjusted by the consumer 102 and the incentive to be received by the consumer based on the predicted value. The power supply business operator is, for example, a power plant. The consumer 102 is, for example, a factory, a commercial facility, a home, etc.

[0014] As illustrated in Fig. 1, a power demand prediction system 100 according to an embodiment includes a power demand prediction device 1 that is connectable to the Internet 101, which is one of communication networks. The power demand prediction device 1 is connected to consumers 102 via the Internet 101. In other words, the power demand prediction device 1 and the consumers 102 are capable of exchanging information with each other via the Internet 101. Note that the communication network is not limited to the Internet 101 as long as the power demand prediction device 1 and the consumers 102 are capable of exchanging information with each other.

[0015] The power demand forecasting device 1 predicts power demand of the consumer 102 in a DR time period. Such a power demand forecasting device 1 is a computer such as an electronic control device, and includes a processor such as a CPU or GPU (not shown), memories such as a ROM and a RAM, and an I / O interface. The processor operates (calculates, etc.) in accordance with instructions of a program loaded into the memory, thereby realizing each of the functional units of the power demand forecasting device 1. Each of the functional units of the power demand forecasting device 1 according to one embodiment will be described with reference to FIG. 2.

[0016] <Configuration of the power demand forecasting device> Fig. 2 is a schematic functional block diagram of a power demand prediction device 1 according to an embodiment. Fig. 3 is a graph showing a first predicted value 50, a second predicted value 52, and a corrected first predicted value 54, with the horizontal axis representing time and the vertical axis representing power.

[0017] As shown in FIG. 2, the power demand forecasting device 1 includes a forecasted value calculation unit 2, an actual measurement value acquisition unit 4, and a forecasted value correction unit 6.

[0018] The predicted value calculation unit 2 calculates a first predicted value 50 and a second predicted value 52. The first predicted value 50 is a predicted value of the power demand of the consumer 102 in the DR time period, and is calculated based on the power consumption (hereinafter referred to as past power consumption 51) consumed by the consumer 102 in the past before the set date on which the DR time period is set. The second predicted value 52 is a predicted value of the power demand of the consumer 102 in a predetermined time period (second time period) before the DR time period on the set date, and is calculated based on the past power consumption 51, similar to the first predicted value 50. In this way, the first predicted value 50 and the second predicted value 52 are calculated based on the past power consumption 51. Note that the past power consumption 51 is not limited to being earlier than the set date on which the DR time period is set, and may be earlier than the DR time period.

[0019] 2, the prediction value calculation unit 2 acquires the past power consumption 51 from outside the power demand prediction device 1, but the present disclosure is not limited to this embodiment. For example, the past power consumption 51 may be stored in advance in a memory of the power demand prediction device 1, and the prediction value calculation unit 2 may acquire the past power consumption 51 from the memory.

[0020] The first predicted value 50 and the second predicted value 52 will be described with reference to Fig. 3. In Fig. 3, both the first predicted value 50 and the second predicted value 52 are illustrated by solid lines. The portion of the solid line included in the time zone from time A1 to time A2 is the first predicted value 50, and the portion included in the time zone from time A3 to time A4 is the second predicted value 52. That is, in this embodiment, the DR time zone is the time zone from time A1 to time A2, and the predetermined time zone is the time zone from time A3 to time A4.

[0021] In some embodiments, the predetermined time period is set immediately before the DR time period, and for example, time A3 (start time of the predetermined time period) is set between 5 hours before time A1 (start time of demand response) and time A1. With such a configuration, the deviation of the predetermined time period from the DR time period is small, so that the difference in conditions that increase or decrease the power consumption of the consumer 102, such as temperature and weather, between the DR time period and the predetermined time period can be reduced.

[0022] Here, an example of a method for calculating the first predicted value 50 and the second predicted value 52 will be described. In the present disclosure, a method called "High X of Y" will be described. In this case, the past power consumption 51 is the power consumption of the consumer 102 for Y days immediately before the set date.

[0023] In one embodiment, the prediction value calculation unit 2 selects four days (X=4) with the highest power consumption among the five days (Y=5) immediately prior to the set date in one-hour units (for example, between midnight and 1:00), calculates the average power consumption from midnight to 1:00 for these four days, and sets this calculated average value as the prediction value (baseline) from midnight to 1:00 on the set date. In a similar manner, the average value is calculated for each time period including at least the DR time period and the specified time period. The calculation of the average value may be performed for each consumer 102, or may be performed for multiple consumers 102 at once.

[0024] The actual measurement value acquiring unit 4 acquires an actual measurement value 56b (56) of the power consumption consumed by the consumer 102 in a predetermined time period on a set date. For example, the actual measurement value acquiring unit 4 receives the actual measurement value 56 in real time from a meter that measures the power consumption of the consumer 102. In Fig. 3, the actual measurement value 56 is illustrated by a dashed dotted line. In the following, for the sake of explanation, the actual measurement value 56 in the DR time period is referred to as the actual measurement value 56a (56), and the actual measurement value 56 in the predetermined time period is referred to as the actual measurement value 56b (56).

[0025] When the absolute value of a correction reference value 57 obtained by subtracting the actual measurement value 56b from the second predicted value 52 is greater than a predetermined positive allowable error threshold 53, the predicted value correction unit 6 subtracts a correction adoption value 59 from the first predicted value 50 to correct the first predicted value 50. The absolute value of the correction adoption value 59 is smaller than the absolute value of the correction reference value 57. In FIG. 3, the corrected first predicted value 54 is shown by a thick line that is thicker than the solid lines indicating the first predicted value 50 and the second predicted value 52. On the other hand, in this embodiment, when the absolute value of the correction reference value 57 is smaller than the allowable error threshold 53, the predicted value correction unit 6 does not correct the first predicted value 50.

[0026] The allowable error threshold 53 will be described. The allowable error threshold 53 is a value that is determined in advance to determine whether or not the predicted value has sufficient accuracy with respect to the actual measured value 56. In some embodiments, when the consumer 102 is equipped with a power generation device capable of generating electricity, the allowable error threshold 53 is determined based on the power generation capacity of the power generation device. For example, the allowable error threshold 53 is a value (e.g., 200 kW) obtained by subtracting the lower limit error from the upper limit error of the power generation capacity of the power generation device during rated operation. With this configuration, the power generation capacity of the power generation device is determined in advance, so the allowable error threshold 53 can be easily determined.

[0027] 3, since the correction reference value 57 is equal to or greater than 0 (second predicted value 52>actual measured value 56b), the correction adoption value 59 is subtracted from the first predicted value 50 to calculate the corrected first predicted value 54. The correction adoption value 59 is a value obtained by subtracting a predetermined allowable error deviation suppression value 55 from the correction reference value 57. The allowable error deviation suppression value 55 is a value for suppressing the magnitude (absolute value) of the correction reference value 57. Although not shown, in this embodiment, when the correction reference value 57 is less than 0 (second predicted value 52<actual measured value 56b), the correction adoption value 59 is a value obtained by adding the allowable error deviation suppression value 55 to the correction reference value 57.

[0028] <Functions and Effects of the Power Demand Forecasting Device> The first predicted value 50 and the second predicted value 52 are calculated in the same manner based on the past power consumption 51. For this reason, if the second predicted value 52 (predicted value of the power demand of the consumer 102 in a predetermined time period) significantly differs from the actual measured value 56b (actual measured value of the power demand of the consumer 102 in a predetermined time period), the first predicted value 50 may also significantly differ from the actual measured value 56a of the power demand in the case where a DR event does not occur (actual measured value of the power demand of the consumer 102 in the DR time period).

[0029] In contrast, according to the present embodiment, when the absolute value of the correction reference value 57 is greater than the allowable error threshold value 53, the prediction value correction unit 6 corrects the first prediction value 50 by subtracting the correction adoption value 59 from the first prediction value 50. The absolute value of the correction adoption value 59 is smaller than the absolute value of the correction reference value 57. For this reason, it is possible to prevent the corrected first prediction value 54 from deviating significantly from the actual measurement value 56, similar to the second prediction value 52, and to improve the prediction accuracy of the power demand in the DR time period.

[0030] The operation and effect of the power demand forecasting device 1 will be described in more detail. FIG. 4 is a diagram showing a case where the first predicted value 50 is corrected with the correction adopted value 59 and a case where the first predicted value 50 is corrected with the correction reference value 57. According to this embodiment, when the absolute value of the correction reference value 57 (the deviation between the second predicted value 52 and the actual measured value 56b) is larger than the allowable error threshold value 53, the correction is not performed by subtracting the correction reference value 57 from the first predicted value 50, but by subtracting the correction adopted value 59 from the first predicted value 50. In other words, when the accuracy of the second predicted value 52 with respect to the actual measured value 56b is low, the accuracy of the first predicted value 50 with respect to the actual measured value 56a is also insufficient, and the first predicted value 50 is corrected. Here, as shown in FIG. 4, when the first predicted value 50 is corrected with the correction reference value 57, the absolute value of the correction reference value 57 is too large, so that the prediction accuracy of the first predicted value 61 corrected with the correction reference value 57 with respect to the actual measured value 56a may be low. In contrast, in this embodiment, instead of adopting the correction reference value 57 to correct the first predicted value 50, a correction adoption value 59 having an absolute value smaller than the correction reference value 57 is adopted, thereby making it possible to obtain a corrected first predicted value 54 with high prediction accuracy.

[0031] On the other hand, according to this embodiment, no correction is performed when the absolute value of the correction reference value 57 is smaller than the allowable error threshold value 53. In other words, when the prediction accuracy of the second predicted value 52 for the actual measurement value 56b is high, there is a high possibility that the prediction accuracy of the first predicted value 50 for the actual measurement value 56a is also sufficiently high. Therefore, it is possible to obtain the first predicted value 50 with high prediction accuracy.

[0032] According to this embodiment, when the correction reference value 57 is equal to or greater than 0, the correction adoption value 59 is a value obtained by subtracting the allowable error deviation suppression value 55 from the correction reference value 57, and when the correction reference value 57 is less than 0, the correction adoption value 59 is a value obtained by adding the allowable error deviation suppression value 55 to the correction reference value 57. Therefore, the prediction accuracy of the first predicted value 50 can be improved both in the case where the second predicted value 52 is smaller than the actual measured value 56b and in the case where the second predicted value 52 is larger than the actual measured value 56b.

[0033] In some embodiments, the allowable error deviation suppression value 55 is equal to or greater than the allowable error threshold 53 and equal to or less than twice the allowable error threshold. Since the correction of the first predicted value 50 is performed by subtracting the correction adoption value 59 (=correction reference value 57-allowable error deviation suppression value 55) from the first predicted value 50, the corrected first predicted value 54 can be made closer to the actual measurement value 56a by setting the allowable error deviation suppression value 55 to equal to or greater than the allowable error threshold 53. On the other hand, if the allowable error deviation suppression value 55 becomes too large, the corrected first predicted value 54 moves away from the actual measurement value 56a. For this reason, by setting the allowable error deviation suppression value 55 to equal to or less than twice the allowable error threshold 53, the prediction accuracy of the corrected first predicted value 54 can be ensured to be at least a certain level. In some embodiments, the allowable error deviation suppression value 55 and the allowable error threshold 53 are the same value. With this configuration, the prediction accuracy of the corrected first predicted value 54 can be ensured to be at least a certain level.

[0034] <Electricity demand forecast method> 5 is a flowchart of a power demand forecasting method according to an embodiment. As illustrated in FIG 5, the power demand forecasting method includes a calculation step S1, an acquisition step S2, and a correction step S3.

[0035] The calculation step S1 calculates a first predicted value 50 and a second predicted value 52 based on past power consumption 51. The acquisition step S2 acquires an actual measured value 56b of power consumption consumed by the consumer 102 during a predetermined time period on a set date. The correction step S3 subtracts a corrected adopted value 59 from the first predicted value 50 when the absolute value of a correction reference value 57 obtained by subtracting the actual measured value 56b from the second predicted value 52 is greater than an allowable error threshold value 53. In this power demand prediction method, the absolute value of the corrected adopted value 59 is smaller than the absolute value of the correction reference value 57.

[0036] According to this electricity demand forecasting method, it is possible to prevent the corrected first forecast value 54 from deviating significantly from the actual measured value 56, as with the second forecast value 52, and to improve the forecast accuracy of electricity demand during the DR time period.

[0037] The contents described in each of the above embodiments can be understood, for example, as follows.

[0038] [1] A power demand forecasting device (1) according to the present disclosure is a power demand forecasting device that forecasts power demand of a consumer (102) in a first time slot, and includes: a forecasted value calculation unit (2) that calculates a first forecasted value (50) of power demand in a first time slot and a second forecasted value (52) of power demand in a second time slot prior to the first time slot based on power consumption (51) consumed by the consumer in the past prior to the first time slot; an actual measured value acquisition unit (4) that acquires an actual measured value (56b) of power consumption consumed by the consumer in the second time slot; and a forecasted value correction unit (6) that subtracts a corrected adopted value (59) from the first forecasted value when an absolute value of a correction reference value (57) obtained by subtracting the actual measured value from the second forecasted value is greater than a predetermined allowable error threshold value (53), and the absolute value of the corrected adopted value is smaller than the absolute value of the correction reference value.

[0039] The first predicted value of power demand and the second predicted value of power demand are calculated in the same manner based on the power consumption consumed by the consumer before the first time slot. Therefore, if the second predicted value of power demand is significantly different from the actual measured value of power demand, the first predicted value of power demand may also be significantly different from the actual measured value of power demand. In response to this, according to the configuration described in [1] above, when the absolute value of the correction reference value is greater than the allowable error threshold, the prediction value correction unit subtracts the correction adoption value from the first predicted value. The absolute value of this correction adoption value is smaller than the absolute value of the correction reference value. Therefore, it is possible to prevent the first predicted value after correction from deviating significantly like the second predicted value, and to improve the prediction accuracy of the power demand in an arbitrary time slot (first time slot).

[0040] [2] In some embodiments, in the configuration described in [1] above, the first time period is a demand response time period, and the predicted value calculation unit calculates a first predicted value of power demand during the first time period on a set date and a second predicted value of power demand during the second time period on the set date based on the power consumption consumed by the consumer prior to a set date on which the first time period is set, and the actual value acquisition unit acquires an actual value of the power consumption consumed by the consumer during the second time period on the set date.

[0041] According to the configuration described in [2] above, it is possible to improve the accuracy of forecasting power demand during a demand response period.

[0042] [3] In some embodiments, in the configuration described in [2] above, the correction adoption value is a value obtained by subtracting a predetermined allowable error deviation suppression value (55) from the correction reference value when the correction reference value is equal to or greater than 0, and is a value obtained by adding the allowable error deviation suppression value to the correction reference value when the correction reference value is less than 0.

[0043] According to the configuration described in [3] above, the accuracy of forecasting power demand during a DR time period can be improved in both cases where the second predicted value is smaller than the actual measured value and where the second predicted value is larger than the actual measured value.

[0044] [4] In some embodiments, in the configuration described in [3] above, when the prediction target includes a power generation device capable of generating electricity, the allowable error threshold is determined based on the power generation capacity of the power generation device.

[0045] According to the configuration described in [4] above, since the power generation capacity of the power generation device is determined in advance, the allowable error threshold can be easily determined.

[0046] [5] In some embodiments, in the configuration described in [3] or [4] above, the tolerance deviation suppression value is greater than or equal to the tolerance threshold and less than or equal to twice the tolerance threshold.

[0047] According to the configuration described in [5] above, the correction of the first predicted value is performed by subtracting the correction adoption value (= correction reference value - allowable error deviation suppression value) from the first predicted value, so by setting the allowable error deviation suppression value to be equal to or greater than the allowable error threshold, the prediction accuracy of the corrected first predicted value can be improved. On the other hand, if the allowable error deviation suppression value becomes too large, the corrected first predicted value will deviate from the actual measured value. For this reason, by setting the allowable error deviation suppression value to be equal to or less than twice the allowable error threshold, the prediction accuracy of the corrected first predicted value can be ensured to be at least a certain level.

[0048] [6] In some embodiments, in the configuration according to any one of [3] to [5] above, the tolerance deviation suppression value and the tolerance threshold value are the same value.

[0049] According to the configuration described in [6] above, it is possible to ensure that the prediction accuracy of the corrected first predicted value is at least a certain level.

[0050] [7] In some embodiments, in the configuration according to any one of [1] to [6] above, the prediction value correction unit does not correct the first prediction value when an absolute value of the correction reference value is smaller than the allowable error threshold value.

[0051] When the prediction accuracy of the second predicted value is high, it is highly likely that the prediction accuracy of the first predicted value is also sufficiently high. According to the configuration described in [7] above, when the absolute value of the correction reference value (second predicted value - actual value) is smaller than the allowable error threshold, the first predicted value is not corrected, so that a first predicted value with high prediction accuracy can be obtained.

[0052] [8] The power demand forecasting method disclosed herein is a power demand forecasting method for forecasting a power demand of a consumer in a first time period, comprising the steps of: calculating (S1) a first predicted value of power demand in the first time period and a second predicted value of power demand in a second time period prior to the first time period based on power consumption consumed by the consumer prior to the first time period; acquiring (S2) an actual measured value of power consumption consumed by the consumer in the second time period; and, if the absolute value of a correction reference value obtained by subtracting the actual measured value from the second predicted value is greater than a predetermined allowable error threshold, subtracting (S3) a correction adoption value from the first predicted value, wherein the absolute value of the correction adoption value is smaller than the absolute value of the correction reference value.

[0053] According to the method described in [8] above, when the absolute value of the correction reference value obtained by subtracting the actual measurement value from the second predicted value is greater than the allowable error threshold, the correction adoption value is subtracted from the first predicted value. The absolute value of this correction adoption value is smaller than the absolute value of the correction reference value. This prevents the first predicted value from deviating significantly, as with the second predicted value, and improves the prediction accuracy of the power demand in a given time period (first time period). [Explanation of symbols]

[0054] 1. Electricity demand forecasting device 2. Prediction value calculation section 4. Actual measurement acquisition section 6. Prediction value correction section 50 First predicted value 51 Past power consumption 52 Second predicted value 53 Tolerance Threshold 54 Corrected first predicted value 55 Permissible error deviation suppression value 56 Actual value 56a DR time zone actual measurement value 56b Actual measurements during a given period 57 Correction standard value 59 Corrected Value 100 Electricity Demand Forecasting System 101 Internet 102 Consumer S1 Calculation Step S2 Acquisition Step S3 correction step

Claims

1. An electric power demand forecasting device that forecasts electric power demand of a consumer in a first time slot, a predicted value calculation unit that calculates a first predicted value of power demand in the first time slot and a second predicted value of power demand in a second time slot prior to the first time slot, based on power consumption consumed by the consumer in the past prior to the first time slot; an actual measurement value acquisition unit that acquires an actual measurement value of the power consumption consumed by the consumer during the second time period; a predicted value correcting unit that subtracts a correction adoption value from the first predicted value when an absolute value of a correction reference value obtained by subtracting the actual measurement value from the second predicted value is greater than a predetermined allowable error threshold value, the absolute value of the correction adopted value is smaller than the absolute value of the correction reference value; Electricity demand forecasting device.

2. the first time period is a demand response time period, the predicted value calculation unit calculates a first predicted value of power demand during the first time slot on a set date and a second predicted value of power demand during the second time slot on the set date based on power consumption consumed by the consumer in the past before a set date on which the first time slot is set; the actual measurement value acquisition unit acquires an actual measurement value of the power consumption consumed by the consumer during the second time slot on the set date; The power demand forecasting device according to claim 1 .

3. The correction adoption value is If the correction reference value is equal to or greater than 0, the correction reference value is a value obtained by subtracting a predetermined tolerance deviation suppression value from the correction reference value, When the correction reference value is less than 0, the correction reference value is added to the tolerance deviation suppression value. The power demand forecasting device according to claim 1 or 2.

4. When the consumer is equipped with a power generation device capable of generating power, the allowable error threshold is determined based on the power generation capacity of the power generation device. The power demand forecasting device according to claim 3 .

5. the tolerance deviation suppression value is equal to or greater than the tolerance threshold value and equal to or less than twice the tolerance threshold value; The power demand forecasting device according to claim 3 or 4.

6. The tolerance deviation suppression value and the tolerance threshold value are the same value. The power demand forecasting device according to any one of claims 3 to 5.

7. the predicted value corrector does not correct the first predicted value when the absolute value of the correction reference value is smaller than the allowable error threshold value; The power demand forecasting device according to claim 1 .

8. 1. A power demand forecasting method for forecasting power demand of a consumer in a first time slot, comprising: calculating a first predicted value of power demand during the first time slot and a second predicted value of power demand during a second time slot prior to the first time slot, based on power consumption consumed by the consumer prior to the first time slot; acquiring an actual measurement value of power consumption consumed by the consumer during the second time period; and when an absolute value of a correction reference value obtained by subtracting the actual measurement value from the second predicted value is greater than a predetermined allowable error threshold, subtracting a correction adoption value from the first predicted value, the absolute value of the correction adopted value is smaller than the absolute value of the correction reference value; Electricity demand forecasting methods.

Citation Information

Patent Citations

  • Slope-stabilizing construction

    JP1986028624A

  • Power demand-and-supply adjustment requirement amount determination device and method therefor

    JP2016135040A

  • Monitoring control device

    JP2017200297A

  • Power storage device control method, power storage device control apparatus, and power storage system

    JP2018057092A

  • Prediction system and prediction method

    JP2018092445A