JPH08136032A - Load prediction correction device - Google Patents

Abstract

(57) [Summary] [Purpose] The present invention relates to a load prediction correction device for buildings and the like, and improves the accuracy of correction of the predicted load amount on all days. [Structure] A load is predicted by creating a regression equation by a regression analysis for each time set by the constant sampling interval setting means 7, and at the same time as the load predicted for the predicted outside temperature one sampling time ahead on the day. And the predicted load correction means 18 that corrects the predicted load by the outside air temperature correction means 17 that adds the difference between the predicted outside air temperature and the actually measured outside air temperature.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a load prediction correction device for predicting a load such as a heat load or a power load in a building or the like.

[0002]

2. Description of the Related Art Various methods have already been developed for load prediction correction methods for heat loads and electric power loads in buildings and the like.
There is a load prediction correction method as disclosed in Japanese Patent Publication No.

The conventional basic technique will be described below with reference to the drawings. In FIG. 6, 1 is an air conditioning load predicting unit that predicts the air conditioning load from past measured values of the air conditioning load obtained by a calorimeter, 2 is an average outside temperature calculating unit that calculates the average outside temperature of the day, and 3 is the past number. 7am to 12pm
The average outside temperature predicting means 4 for calculating the average outside temperature over a predetermined time such as time, and using this as the predicted value of the average outside temperature is a subtracting means, and the average outside temperature calculated by the average outside temperature calculating means 2. From the average outside temperature predicting means 3 is obtained.

Reference numeral 5 denotes load correction means for correcting the predicted air conditioning load amount calculated by the air conditioning load prediction means 1 based on the value calculated by the subtraction means 4.
Is composed.

[0005]

However, in the above-mentioned conventional example, the predicted average outside temperature calculated by the average outside temperature predicting means 3 for the entire day and the actual average outside temperature calculated by the average outside temperature calculating means 2. The predicted load amount is corrected based on the difference from the above, and since there is no correction value for each elapsed time in the middle, the result is that the accuracy of the all-day predicted load correction becomes poor.

SUMMARY OF THE INVENTION In view of the above problems, it is an object of the present invention to provide a load prediction correction apparatus which has a high accuracy in correcting the predicted load amount on all days.

[0007]

In order to solve the above problems, a load prediction correction apparatus according to the present invention comprises a constant sampling interval setting means for outputting a sampling signal at a predetermined frequency for each constant time, and a sampling interval for each sampling interval. An outside air temperature detecting means for detecting an actually measured outside air temperature, a load detecting means for detecting an actually measured load amount, a regression formula calculating means for calculating a regression formula from the actually measured outside air temperature and the actually measured load amount at the same sampling time,
The outside temperature predicting means for predicting the outside temperature at the same sampling time on the next day, the subtracting means for subtracting the predicted outside temperature from the measured outside temperature at the same sampling time on the next day, and 1 for this output
An adding means for adding the predicted outside temperature at the sampling time ahead,
This output is substituted into the regression equation calculated by the regression equation calculating means, and a predictive load correction means for calculating the load amount one sampling time ahead is provided.

Further, in place of the subtracting means and the adding means, the current predicted actual measured outside temperature subtracting means for subtracting the actually measured outside air temperature from the predicted outside air temperature at the same sampling time, and the current time for similarly subtracting the second corrected outside air temperature from the predicted outside air temperature Prediction second corrected outside temperature subtraction means, next predicted first corrected outside air temperature subtraction means for subtracting the first corrected outside air temperature from the predicted outside air one sampling time ahead, and second predicted outside air temperature is similarly subtracted from the predicted outside air temperature. Of the next predicted second corrected outside temperature subtracting means, the ratio of the output value of the currently predicted actually measured outside temperature subtracting means to the output value of the currently predicted second corrected outside temperature subtracting means, and the next predicted first corrected outside temperature subtracting means The second corrected outside air temperature calculating means for calculating the second corrected outside air temperature by providing the same ratio between the output value and the next predicted second corrected outside air temperature subtracting means is provided.

[0009]

The load prediction correction apparatus of the present invention creates a regression equation by regression analysis at each time set by the constant sampling interval setting means to predict the load, and on the correction day, the same sampling time as the load prediction time. The predicted outside air temperature at the above is subtracted from the measured outside air temperature, and the value is added to the predicted outside air temperature one sampling time ahead, and the corrected outside air temperature is calculated by correcting the predicted outside air temperature at each sampling interval, and this value is regressed. Since the load amount is calculated by substituting it into the formula, it is possible to correct the load amount at regular sampling intervals.
That is, since it is not a batch correction of the predicted load amount on all the days, a high accuracy of the predicted load amount is always obtained.

Further, the load prediction correction device obtains the corrected outside temperature one sampling time ahead, assuming that the ratio of the predicted outside air temperature at the current time to the actually measured outside air temperature and the corrected outside air temperature is equal to that one sampling time ahead, Since the load amount is calculated by substituting this value into the regression equation, it is possible to correct the load amount at constant sampling intervals, and since the corrected outside air temperature at the current time is adopted as the correction factor, The overall prediction accuracy is further improved.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A load prediction correction apparatus according to an embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of a load prediction correction apparatus according to an embodiment of the present invention. Reference numeral 7 is a constant sampling interval setting means for outputting a sampling signal at a constant time interval, for example, every hour, and 8 is an outside air temperature at each time interval set by the constant sampling interval setting means 7 for the past several days, for example, 5 days. Outside temperature detecting means for detecting, 9 is load detecting means for detecting the load amount at the same sampling interval, and 10 is the past 5
The storage means stores the load and the outside temperature detected by the outside temperature detecting means 8 and the load detecting means 9 at each sampling interval of the day.

Reference numeral 11 denotes a regression formula calculating means for calculating a regression formula from the outside air temperature and load amount stored in the storage means 10 at the same sampling time for the past five days, and 12 predicts the outside air temperature at the same sampling time on the next day. It is a means to predict the outside temperature.

Reference numeral 13 denotes load calculating means for substituting the predicted outside air temperature predicted by the outside air temperature predicting means 12 into the regression formula calculated by the regression formula calculating means 11 for each sampling interval to calculate the load of the next day. The formula calculation means 11, the outside air temperature prediction means 12, and the load calculation means 13 operate at night when the indoor load can be ignored, and constitute a load prediction device 14.

Reference numeral 15 is an outside air temperature predicting means 1 at the same sampling time of the previous day from the outside air temperature detected by the outside air temperature detecting means 8.
Subtracting means for reducing the predicted outside air temperature predicted in step 2, 16 is addition means for adding the predicted outside air temperature one sampling time ahead to the output of the subtracting means 15, and 17 is subtraction means 15 and addition means 16
It is a means for correcting the outside air temperature.

Reference numeral 18 denotes 1 by substituting the outside temperature corrected by the outside temperature correcting means 17 into the regression equation calculated by the regression equation calculating means 11.
It is a predictive load correcting means for correcting the predictive load ahead of the sampling time, and includes an outside air temperature correcting means 17 and a predictive load correcting means 1.
The load compensating device 8 operates during the daytime the day after the load predicting device 14 operates. The load prediction correction device 20 is configured by the above components.

The operation of the load prediction correction apparatus 20 of the present embodiment configured as above will be described with reference to the flowchart of FIG.

STEP 1 of FIG. 2 is a routine for judging whether or not the air conditioning operation time (daytime) of 5 days has elapsed. If it has not elapsed, go to STEP2. If it has elapsed, S.
Move to TEP3.

In STEP 2, a constant sampling time (1
The routine for determining whether or not (time) has elapsed is performed by the constant sampling interval setting means 7, and if 1 hour has elapsed, the outside air temperature detecting means 8 detects the measured outside air temperature, and the load detecting means 9 detects the load amount. To do. These values are stored in the storage unit 10. The operation of STEP 2 is continuously performed unless the daytime of 5 days has passed.

STEP 3 is a routine for judging whether or not it is the non-air-conditioning operation time (night) on the 5th day. If it is the night of the fifth day, the process proceeds to STEP4. If not, that is, if it is the daytime of the sixth day, the process proceeds to STEP5.

In STEP 4, the regression equation calculating means 11 calculates a regression equation of the relationship between the two on the basis of the outside air temperature and the load amount of the storage means 10. Next, the outside temperature predicting means 12 calculates the average of the outside temperatures for the past five days and sets it as the predicted outside temperature at the same sampling time on the next day. Further, the load calculation means 13 substitutes the predicted outside temperature in the regression equation to predict the load of the next day.

In STEP 5, the subtracting means 15 in the outside air temperature correcting means 17 subtracts the predicted outside air temperature predicted by the outside air temperature predicting means 12 at the same sampling time of the previous day from the outside air temperature detected by the outside air temperature detecting means 8. In addition, the adding means 16 adds the predicted outside air temperature one sampling time ahead to the output of the subtracting means 15.

Further, in STEP 6, the predicted load correction means 1
At 8, the outside temperature corrected by the outside temperature correction unit 17 is substituted into the regression formula calculated by the regression formula calculation unit 11 to correct the predicted load one sampling time ahead.

For simplification of the description, the detection, prediction, and correction operation of the outside temperature and the load amount of the next day are described as a single-shot operation, but they are actually performed continuously. .

As described above, according to this embodiment, a model formula is created by regression analysis for each time set by the constant sampling interval setting means 7 to predict the load. The predicted outside temperature at the same sampling time is subtracted from the outside temperature, the value is added to the predicted outside temperature one sampling time ahead, and the predicted load is corrected by correcting the predicted outside temperature at each sampling interval. It becomes possible to correct the load at each sampling interval. That is, since it is not a batch correction of the predicted load amount on all the days, a high accuracy of the predicted load amount is always obtained.

Next, other embodiments will be described. However, since the embodiment is the same as the above embodiment except the outside air temperature correcting means 17 ', detailed description thereof will be omitted.

FIG. 3 is a block diagram of a load correction apparatus according to another embodiment of the present invention. Reference numeral 21 is a current predicted actual measured outside temperature subtracting means for subtracting the actually measured outside air temperature from the predicted outside air temperature at the same sampling time, and 22 is a current predicted second corrected outside air temperature subtracting means for subtracting the second corrected outside air temperature from the predicted outside air temperature at the same sampling time. , 23 is the next predicted first corrected outside temperature subtracting means for subtracting the first corrected outside temperature from the predicted outside temperature one sampling time ahead, and 24 is not predicted at the same sampling time as the next predicted first corrected outside temperature subtracting means 23. It is a next-time prediction second corrected outside air temperature subtracting means for subtracting the second corrected outside air temperature from the air temperature.

Reference numeral 25 denotes the ratio of the output value of the currently predicted actually measured outside air temperature subtracting means 21 to the output value of the currently predicted second corrected outside air temperature subtracting means 22 and the output value of the next predicted second corrected outside air temperature subtracting means 24. It is a second corrected outside air temperature calculation means for calculating the second corrected outside air temperature by making the ratios of the time prediction first corrected outside air temperature subtraction means 23 the same.

The operation of the load prediction correction apparatus of the other embodiment configured as described above will be described with reference to the flowchart of FIG. 4 and the operation explanatory diagram of FIG. The detailed description thereof will be omitted.

STEP 5 different from the previous embodiment is STEP 5
This is replaced with STEP 7 for description.

STEP 7-1 in STEP 7 is performed by the currently predicted actual measured outside temperature subtracting means 21, and subtracts the actually measured outside temperature B from the predicted outside temperature A at the current time in FIG.

Step 7-2 is performed by the currently predicted second corrected outside air temperature subtracting means 22 and subtracts the second corrected outside air temperature C which is the final corrected outside air temperature at the current time from the predicted outside air temperature A at the current time. .

Step 7-3 is performed by the next predicted first corrected outside temperature subtracting means 23, and the current predicted actual measured outside temperature subtracting means 2 is calculated from the predicted outside temperature D one sampling time ahead of the current time.
The output value of 1 is subtracted, and this calculated value is used as the first corrected outside air temperature.

In STEP 7-4, the second predicted second corrected outside air temperature subtracting means 24 subtracts the temporarily set second corrected outside air temperature from the predicted outside air temperature D one sampling time ahead of the current time.

STEP 7-5 is performed by the second corrected outside air temperature calculating means 25, and outputs the output value of STEP 7-1 and STEP 7-5.
-2 output value ratio, STEP7-4 output value and STE
By making the ratio of the output values of P7-3 the same, STEP
The second corrected outside temperature temporarily set in 7-4 is obtained.

As described above, according to the present embodiment, the load prediction correction apparatus determines that the ratio of the predicted outside temperature at the current time to the measured outside temperature and the corrected outside temperature is equal to that of one sampling time ahead. Since the load amount is calculated by calculating the corrected outside temperature at the sampling time ahead and substituting this value into the regression equation, it becomes possible to correct the load amount at fixed sampling intervals and the corrected outside temperature at the current time. Since is used as a correction factor, the overall prediction accuracy is further improved.

[0037]

As is apparent from the embodiments described above, the present invention provides a constant sampling interval setting means for outputting a sampling signal at a predetermined frequency for each constant time, and an outside temperature detecting means for detecting the outside air temperature at each sampling interval. An air temperature detecting means, a load detecting means for similarly detecting a load amount, a regression formula calculating means for calculating a regression equation from the outside air temperature and the load for the past several days, and an outside air temperature for predicting the outside air temperature at the same sampling time on the next day. Load predicting means comprising predicting means and load calculating means for substituting the outside temperature predicted by the outside temperature predicting means into the regression equation calculated by the regression equation calculating means at each sampling interval to calculate the load of the next day. And subtraction means for subtracting the predicted outside temperature predicted by the outside temperature prediction means at the same sampling time of the previous day from the outside temperature, and the predicted outside temperature one sampling time ahead is added to the output of the subtraction means. And an estimated temperature correction means for correcting the estimated load one sampling time ahead by substituting the outside temperature corrected by the outside temperature correction means into the regression equation calculated by the regression equation calculation means. Since the load correction means that operates during the daytime on the next day after the operation of the load prediction means is provided, the load can be corrected at regular sampling intervals. That is, since it is not a batch correction of the predicted load amount on all the days, a high accuracy of the predicted load amount is always obtained.

Also, the current predicted actual measured outside temperature subtracting means for subtracting the measured outside temperature from the predicted outside temperature at the same sampling time and the current predicted second corrected outside for subtracting the second corrected outside temperature from the predicted outside temperature at the same sampling time. Temperature subtraction means and 1
Next-time prediction first corrected outside air temperature subtraction means for subtracting the first corrected outside air temperature from the predicted outside air temperature ahead of the sampling time, and next-time prediction second corrected outside air temperature subtraction means for similarly subtracting the second corrected outside air temperature from the predicted outside air temperature. , The ratio of the output value of the current predicted actual measured outside temperature subtracting means and the output value of the current predicted second corrected outside temperature subtracting means, the output value of the next predicted second corrected outside temperature subtracting means and the next predicted first
Since the second corrected outside air temperature calculation means for calculating the second corrected outside air temperature by providing the same ratio of the corrected outside air temperature subtraction means is provided, it becomes possible to correct the load amount at constant sampling intervals. Moreover, since the corrected outside temperature at the current time is adopted as the correction factor, the overall prediction accuracy is further improved.

[Brief description of drawings]

FIG. 1 is a block diagram of a load prediction correction apparatus according to an embodiment of the present invention.

FIG. 2 is an operation flowchart of the load prediction correction apparatus in the embodiment.

FIG. 3 is a block diagram of a load correction device according to another embodiment of the present invention.

FIG. 4 is an operation flowchart of the load prediction correction apparatus in the embodiment.

FIG. 5 is a characteristic diagram for explaining the operation of the load prediction correction apparatus in the same embodiment.

FIG. 6 is a block diagram of a conventional load prediction correction device.

[Explanation of symbols]

 7 constant sampling interval setting means 8 outside temperature detecting means 9 load detecting means 11 regression formula calculating means 12 outside temperature predicting means 14 load predicting device 15 subtracting means 16 adding means 17, 17 'outside air temperature correcting means 18 predicted load correcting means 19 load Correction device 20 Load prediction correction device 21 Current prediction actual measurement outside temperature subtraction means 22 Current prediction second corrected outside temperature subtraction means 23 Next prediction first corrected outside temperature subtraction means 24 Next prediction second corrected outside air temperature subtraction means 25 Second Corrected outside temperature calculation means

Claims (2)

[Claims] 1. A load predicting apparatus in a building or the like, comprising: a constant sampling interval setting means for outputting a sampling signal at a predetermined frequency for each constant time; and an outside air temperature detecting means for detecting a measured outside air temperature at each sampling interval. Similarly, a regression equation is calculated from the load detecting means for detecting the measured load amount, the measured outside air temperature detected by the outside air temperature detecting means for the past several days at the same sampling time, and the measured load amount detected by the load detecting means. Regression formula calculation means,
A load predicting device that operates at night when the indoor load can be ignored and that consists of an outside air temperature predicting unit that predicts the outside air temperature at the same sampling time on the next day, and at the same sampling time from the measured outside air temperature detected by the outside air temperature detecting unit. An outside air temperature correction means including subtraction means for reducing the predicted outside air temperature predicted by the outside air temperature prediction means, an addition means for adding the predicted outside air temperature one sampling time ahead to the output of the subtraction means, and the regression equation calculation means Load correction that operates in the daytime on the next day when the load predicting means operates And a load prediction correction device. 2. A load predicting apparatus in a building or the like, comprising: a constant sampling interval setting means for outputting a sampling signal at a fixed frequency for each constant time; and an outside air temperature detecting means for detecting an actually measured outside air temperature at each sampling interval. Similarly, a regression equation is calculated from the load detecting means for detecting the measured load amount, the measured outside air temperature detected by the outside air temperature detecting means for the past several days at the same sampling time, and the measured load amount detected by the load detecting means. Regression formula calculation means,
A load prediction device that consists of an outside temperature prediction means that predicts the outside temperature at the same sampling time on the next day and operates at night when the indoor load can be ignored, and a current prediction measurement that subtracts the measured outside temperature from the predicted outside temperature at the same sampling time. Outside temperature subtraction means, current prediction second subtraction means that also subtracts the second corrected outside temperature from the predicted outside temperature, and next prediction first subtraction means that subtracts the first corrected outside temperature from the predicted outside temperature one sampling time ahead Corrected outside air temperature subtraction means, next predicted second corrected outside air temperature subtraction means for subtracting the second corrected outside air temperature from predicted outside air temperature, output value of the current predicted actual measured outside air temperature subtraction means and the current predicted second corrected outside air temperature. By making the ratio of the output value of the temperature subtraction means, the output value of the next-time prediction second corrected outside air temperature subtraction means and the ratio of the next-time predicted first corrected outside air temperature subtraction means the same, the second corrected outside air temperature can be obtained. Calculate The outside air temperature correction means including the second corrected outside air temperature calculation means, and the corrected outside air temperature corrected by the outside air temperature correction means are substituted into the regression equation calculated by the regression equation calculation means and the load amount one sampling time ahead And a load correction device that operates in the daytime on the next day when the load prediction device operates.