JPH03202515A - Device for predicting demand of water delivery amount - Google Patents

Abstract

PURPOSE:To enable know-how and area originality of an operator to easily reflect a prediction result by a method wherein a meteorological correction factor is determined based on meteorological data and a holiday correction factor is determined, and the degree of a delay when the prediction day is a holiday is corrected and varied. CONSTITUTION:Water delivery amount data is inputted to a water distributing system 2 by a water distribution amount data collecting part 1 to prepare a water distribution data file 3. Meanwhile, a meteorological correction factor fuzzy inference part 7 performs fuzzy inference based on predicted weather and predicted maximum air temperature data to determine a meteorological correction factor. A holiday correction factor fuzzy inference part 10 determines a holiday correction factor from a relation between a predicted day and a holiday by referring to a calendar. Average water distribution amount data is corrected by means of a meteorological correction factor and a holiday correction factor, and a predicted water distribution amount is determined by a predicted water distribution amount computing part. Further, when the predicted day is a holiday, the time lag is corrected by a time lag correcting part 11, and an obtained predicted water distribution amount is displayed on a display part 12.

Description

DETAILED DESCRIPTION OF THE INVENTION A. Field of Industrial Application The present invention relates to a device for predicting demand for water distribution in a water supply distribution system.

B1 Summary of the Invention The present invention provides a device for predicting demand for water distribution on a predetermined time basis, which calculates a weather correction coefficient based on weather data and refers to a calendar to determine if the predicted date is a holiday or before or after a holiday. If the forecast day is a holiday, calculate the holiday correction coefficient, use the calculated weather correction coefficient and weather correction coefficient to correct the past average water distribution amount to calculate the predicted water distribution amount, and if the forecast day is a holiday, the forecast water distribution amount will be delayed in the morning. By modifying and changing the calculation rules for weather correction coefficients and holiday correction coefficients, and the degree of delay when the forecast date is a holiday, we will be able to improve the operator's know-how, the uniqueness of the region, etc. This makes it possible to easily reflect this in the prediction results.

C3 Prior Art In general, in a water distribution system, the amount of water to be distributed is predicted when performing water distribution control such as water level management in a distribution reservoir, water distribution pump control, and electric valve control.

This forecast includes hourly demand forecast and short-term demand forecast.

FIG. 16 shows conventional water distribution amount prediction processing.

It collects water distribution data from the water distribution system, and also collects data such as days of the week, holidays, weather, and temperature.

Then, data for a predetermined period in the past is analyzed, and prediction calculations are performed from data such as the day of the week of the prediction date. That is, undetermined coefficients of the prediction calculation formula are determined so that the error of the entire predicted value is minimized, and hourly demand prediction and daily demand prediction are performed based on the results.

The water distribution system was controlled based on the hourly demand and daily demand obtained in this way.

B1 Problems to be Solved by the Invention However, in the above-mentioned conventional techniques, predictions are made using various prediction calculation methods, so it is difficult to reflect the operator's intuition and know-how based on experience in the prediction results.

Furthermore, it is difficult to construct or maintain prediction rules and algorithms without special computer knowledge.

Furthermore, it was not possible to freely incorporate the unique characteristics of each region or region into the demand forecasting algorithm, making it difficult to package general-purpose software.

In view of these problems, it is an object of the present invention to provide a device that allows prediction rules etc. to be easily changed and that can predict the amount of water distributed taking into account the operator's knowledge and regional uniqueness. do.

81 Means for Solving the Problems In order to achieve the above object, the present invention provides a water distribution demand forecasting device for predicting water distribution demand in predetermined units of time, which includes the following means.

■ Water distribution data collection unit that collects water distribution data from the water distribution system.

■ Water distribution data storage unit that stores past water distribution data.

■ Weather data input section where weather data for the predicted day is entered manually. "Weather data" refers to data indicating weather conditions such as the weather and temperature of the day.

■ A weather correction coefficient calculation unit that performs calculations based on weather data to obtain a weather correction coefficient that indicates fluctuations in demand for water distribution due to weather conditions.

■ Calendar storage unit that stores calendars.

■ A holiday correction coefficient calculation unit that performs calculations with reference to this calendar, and calculates a holiday correction coefficient that indicates a change in demand for water distribution due to the predicted day, if it is a holiday, or before or after a holiday.

■ A water distribution forecasting unit that calculates the predicted water distribution amount by calculating the average water distribution amount from past water distribution data and correcting this average water distribution amount based on the weather correction coefficient and the holiday correction coefficient.

■ A delay correction unit that performs calculations with reference to the calendar and delays the predicted water distribution amount for a time period centered on the morning when the predicted day is a holiday.

Here, "delay" refers to correcting the water distribution amount for each hour so that the water distribution amount curve for a given time period shifts in the direction of delay.For example, for each hour, the water distribution amount for that hour and the previous It is possible to take a mode in which the intermediate value of the water distribution amount over time is taken as the corrected water distribution amount.

19 Effects In the water distribution demand forecasting device according to the present invention, the water distribution amount data from the water distribution system is collected by the water distribution amount data collection unit, and the water distribution amount data is calculated at an hourly rate over a predetermined period in the past. Saved in memory. Then, the past average water distribution amount is determined from this data, and the water distribution demand on the predicted day is predicted by applying a predetermined correction to this average water distribution amount.

To make a prediction, first, meteorological data such as expected weather and expected maximum temperature on the prediction day is input by the weather data human resource department.

Using this weather data, a weather correction coefficient calculating section performs calculations according to predetermined calculation rules. In other words, the weather correction coefficient is calculated by comprehensively considering the influence of expected weather and expected maximum temperature on demand for water distribution.

Also, based on the calendar stored in the calendar storage unit in advance, the holiday correction coefficient calculation unit determines whether the predicted date is a holiday or before or after a holiday, and if the predicted date falls under this, a predetermined calculation rule is applied. Calculate the holiday correction coefficient according to the following.

The water distribution amount prediction unit performs calculations based on the weather correction coefficient obtained in this way (or the weather correction coefficient and the holiday correction coefficient when a holiday correction coefficient is obtained), and corrects the past average water distribution amount. Calculate the predicted water distribution amount.

Furthermore, when the predicted day is a holiday, the delay correction unit performs calculations to compensate for the tendency for the water distribution amount curve in the morning to be delayed, and delays the predicted water distribution amount for the time period centered on the morning.

The above calculation calculates the predicted water distribution amount, so the degree to which the influence of weather and temperature is taken into account when calculating the weather correction coefficient, the degree to which the influence of holidays etc. is taken into account when calculating the holiday correction coefficient, and whether the forecast date is a holiday or not It is possible to easily change the degree of delay when

That is, it is possible to easily modify and change the handling of each factor such as weather data in predicting water distribution amount.

G Example Hereinafter, the present invention will be explained in detail using the drawings.

FIG. 1 shows a water distribution demand forecasting device according to an embodiment of the present invention.

The water distribution amount data collection unit 1 detects and collects water distribution amount data from the water distribution system 2 . The collected water distribution amount data is stored in the water distribution amount data file 3.

The predicted water distribution amount calculation unit 4 calculates the predicted water distribution ff1Q by correcting the average water distribution amount data Q, based on a weather correction coefficient of 1 and a holiday correction coefficient of 2, which will be described later.

The data input section 5 is for inputting the predicted weather and predicted maximum temperature on the predicted day. The predicted weather/predicted maximum temperature data file 6 stores data such as predicted weather for a predetermined period in the past.

The weather correction coefficient fuzzy inference unit 7 calculates the weather correction coefficient by performing fuzzy inference based on the predicted weather and predicted maximum temperature data. The weather correction coefficient of 1 is a value indicating the influence of meteorological conditions such as weather and temperature on the amount of water distribution.

The calendar file 8 stores a calendar for a predetermined period.

The calendar manual section 9 is used to set the calendar by inputting holidays such as Sundays and public holidays, and specific dates.

The holiday correction coefficient fuzzy inference unit 10 refers to the calendar and calculates a holiday correction coefficient based on the relationship between the predicted date and the holiday. Holiday correction coefficient K.

is a value indicating the influence of a holiday on the forecast date or the day before or after the forecast date on the water distribution amount.

The time delay correction unit 11 performs time delay correction when the predicted day is a holiday.

The display unit 12 displays the obtained predicted water distribution amount.

Next, the operation of this device will be explained.

The water distribution amount data collection unit 1 takes in water distribution amount data from the water distribution system 2 and creates a water distribution amount data file 3. The water distribution data Qt handled by this device is the water distribution amount per hour (time from t-1) in one day.

Figure 2 shows average water distribution data.

Water distribution amount data file 3 contains data from the past n days (n is 20 to 60
It has water distribution amount data for each hour at an arbitrary value of degree). From this data, average water distribution data can be obtained using the following formula. However, Q is the average water distribution amount at (t-1) ~.

(1) (t-1 to t-24) Next, the calculation of the predicted water distribution amount Q in the predicted water distribution amount calculation unit 4 will be explained.

The predicted water distribution amount Q, is determined by the weather correction coefficient of 1 and the holiday correction coefficient, as shown in the following formula, the average water distribution amount data Q,
By correcting , the predicted water distribution amount Q is obtained.

Q+-(++Kz to 1+) x Qt...(2)
Next, the operation of setting the weather correction coefficient to 1, which is performed by the weather correction coefficient fuzzy inference section 7, will be explained.

In determining the weather correction coefficient, first, 3 is determined as the weather coefficient.

Fluctuations in demand for water distribution are affected not only by the weather on the day, but also by the weather the day before or the day before. The weather coefficient of 3 is a value that evaluates the weather condition between several F3, and is determined from the numerical value of 4 indicating expected weather and the numerical value of 5 indicating weather change.

4 is input as a numerical value of O to 2.0 in the predicted weather.

The weather change can be calculated from the predicted weather 4 using the following formula. However, K, (0) is the expected weather on the forecast day, K
4(-1) is the expected weather on the day before the prediction date, K, (-2) is the expected weather on the day before the prediction date, and α is the coefficient.

K5=, (0)-[, (-t)・α0K, (-2)
・(1-α))...(3) Figure 3 shows the fuzzy rules used in the inference of weather coefficients. These fuzzy rules are nine types of rules in the form of ■F to THEN. An example of this is, ``If the predicted weather is rainy and the weather change is NB (small), the weather coefficient is NB (small).''

FIG. 4 shows the membership functions used in the inference of weather coefficients. Figure (a) shows predicted weather, Figure (b) shows weather changes, and Figure (C) shows membership functions of weather coefficients.

The weather coefficient of 3 can be determined by a method such as the maximum-minimum method using the above-mentioned fuzzy rule and membership function, using 4 for the expected weather and the weather change as condition parts.

Next, the weather correction coefficient fuzzy inference section 7 adds 3 to the weather coefficient.
Using this and the expected maximum temperature change Kll, a weather change coefficient of 7 is determined. The weather change coefficient of 7 is a value for evaluating changes in weather conditions and temperature.

The above predicted maximum temperature change 8 can be calculated using the following equation. However, K8 is the expected maximum temperature on the forecast day, K
8(n) is the average maximum temperature for the past 0 days.

K8=to, -to, (n) -(4
) Figure 5 shows the fuzzy rules used in the inference of weather change coefficients. Similar to FIG. 3, this figure also shows nine types of fuzzy rules.

Moreover, FIG. 6 shows membership functions used in inference of weather change coefficients. Figure (a) is the weather coefficient, Figure (b)
indicates the expected maximum temperature coefficient, and Figure (C) indicates the weather change coefficient.

The condition part is 3 for the weather coefficient and 6 for the expected maximum temperature change, and using the above fuzzy rule and membership function,
A weather change factor of 7 can be inferred.

Finally, the weather correction coefficient fuzzy inference section 7 calculates 7 for the weather change coefficient, the daytime coefficient, and the weather correction coefficient.

Fluctuations in water demand due to weather conditions mainly occur during the day and less frequently at night.

Figure 7 shows the membership function of the time coefficient, where the solid line in (a) is daytime, the -dotted chain line in (b) is nighttime, and the two-dotted line in (c) is
The dashed dotted line indicates the morning.

Further, FIG. 8 shows a value of 8 for the daytime coefficient.

Using this daytime coefficient, 1 can be calculated as the weather correction coefficient by the following formula.

KI=7XK9...(5)
Next, the calculations performed by the holiday correction coefficient fuzzy inference section 10 will be explained.

In general, demand for water distribution tends to decrease during the daytime on holidays. To deal with this, when the predicted day is a holiday, a correction term -β is provided for -.

However, β is a coefficient appropriately set between 0 and 0.5.

In addition, when there are consecutive holidays, the demand for water distribution tends to decrease during the night (afternoon) of the preceding weekday. This tendency becomes more pronounced the longer the consecutive holidays are.

To deal with this, if the predicted date is the day before a consecutive holiday, we use a correction term based on the degree of consecutive holidays and the time coefficient. infer.

FIG. 9 shows the fuzzy rules used for this inference. This figure shows six rules.

Figure 10 shows the membership functions used for this inference. Figure (a) shows consecutive holidays (degree of continuity);
b) indicates a correction term. In the case of weekends (Saturday and Sunday),
It will be treated as 1.5 consecutive days off.

Additionally, on weekdays after consecutive holidays, demand for water in the morning tends to increase. To deal with this, when the predicted date is a weekday after consecutive holidays, a correction term K11 is inferred from the degree of consecutive holidays and the time coefficient.

Figure 111 shows the fuzzy rules used for this inference. This figure also shows six rules, similar to FIG. 8.

Figure 12 shows the membership functions used for this inference. Figure (a) shows consecutive holidays, and Figure (b) shows the correction term.

As shown in the following equation, by summing these correction terms, the holiday correction coefficient can be obtained.

1 to K2- to 1o+. − to 11・β・−(6
) thus to the weather correction factor, and to the holiday correction factor,
After determining, as described above, the predicted water distribution amount calculation unit 4 determines the predicted water distribution cover Q.

Here, the demand curve for water distribution tends to be delayed in the morning on holidays. To deal with this, the time delay correction unit 11
Accordingly, time delay correction is applied to the predicted water distribution amount Q.

FIG. 13 shows time delay correction. In this figure, the predicted water distribution amount (initial value) calculated by the predicted water distribution amount calculation unit 4 is shown by a broken line, and the corrected value is shown by a solid line.

This time delay correction is performed based on the following equation, targeting the time zone set in the morning (see FIG. 7 ( )). However, d is the delay time (can be set arbitrarily in the range O≦d≦120), and Q
, / is the corrected value of the predicted water distribution amount.

When d≦60 (6≦t≦11) When d>60, Q,' =Q5 Q,・Q・−Pa(60~d)+Q・0,・(120−t)・(7) (7 ≦t≦11) The predicted water distribution amount Q, (or Q,′
) is displayed on the display unit 12 and is further used for various controls of the water distribution system.

Note that if the actual weather or temperature is significantly different from the forecast, the predicted weather or predicted maximum temperature can be re-entered and recalculated. In this case, the operator will trigger the recalculation.

Next, a specific example of this embodiment will be explained.

FIG. 14 shows the hardware configuration of this specific example.

The main body of the device consists of an industrial computer 13 and its peripheral devices 14-17.

The industrial computer computer 13 performs the main control of this device. The fuzzy controller 14 performs fuzzy inference. The CRT 15 displays the predicted water distribution amount and various other displays.

The keyboard 16 and mouse 17 are used to input various data and instructions.

The process human output device 18 performs human output of water distribution amount performance data and other data, and is connected to the industrial computer 13 via a LAN (local area network).

In the above configuration, the apparatus shown in FIG. 1 can be realized by constructing predetermined software.

FIG. 15 shows the main procedure in this specific example.

As described in detail of the H3 invention, the water distribution amount prediction device according to the present invention The water distribution amount demand prediction device according to the present invention calculates the weather correction coefficient and the holiday correction coefficient from weather data and the calendar, and uses these coefficients to calculate the weather correction coefficient and the holiday correction coefficient. Correcting the past average water distribution amount, and
If necessary, the time delay in the morning is corrected and the predicted water distribution amount is calculated.

In order to obtain the predicted water distribution amount using the above calculation, the degree to which the influence of weather and temperature is taken into account in calculating the weather correction coefficient,
Easily modify and change the handling of each factor in predicting water distribution volume by changing the degree to which the influence of holidays, etc. is taken into account in calculating the holiday correction coefficient, the degree of delay when the forecast date is a holiday, etc. is possible.

Therefore, the operator's intuition and know-how can be easily reflected in the prediction results, and furthermore, there is an advantage that predictions can be made that incorporate regional uniqueness of the target water distribution system.

Moreover, even without special knowledge of computers, prediction rules can be constructed simply by determining the weight of each factor in calculating the predicted water distribution amount, and maintenance is also easy.

[Brief explanation of drawings]

Figure 1 is a block diagram showing a water distribution demand forecasting device according to an embodiment of the present invention, Figure 2 is an explanatory diagram showing average water distribution data, and Figures 3 and 4 are used for inference of weather coefficients. Fig. 5 is an explanatory diagram showing fuzzy rules and membership functions used in inference of weather change coefficients; Fig. 7 is an explanatory diagram showing membership functions of time coefficients. Figure 8 is an explanatory diagram showing the value of the daytime coefficient, Figures 9 and 10 are explanatory diagrams showing the fuzzy rule and the menpantsubu function used in the inference of the pre-holiday correction, and Figures 11 and 12 are
FIG. 13 is an explanatory diagram showing fuzzy rules and membership functions used for inference of post-holiday correction, FIG. 13 is an explanatory diagram showing time delay correction, and FIG. 14 is a block diagram showing the hardware configuration of one specific example. FIG. 15 is a flowchart showing the main procedure in the specific example of FIG. 14, and FIG. 16 is an explanatory diagram showing conventional water distribution amount prediction processing. 1... Water distribution amount data collection unit, 3 Water distribution amount data file, 4... Predicted water distribution amount calculation unit, 5... Data input unit, 6
Forecasted weather/expected maximum temperature Cheetah file, 7. Weather correction coefficient fuzzy inference section, 8. Calendar file,
10...Holiday correction coefficient fuzzy inference section, 11 Time delay correction section. Figure 2 Average water distribution time - Figure 3 Fuzzy rules for weather coefficient inference Membership functions for weather coefficient inference (a) (b) (C) Figure 7 Membership number of time coefficients Value of daytime coefficient Figure 5 Fuzzy rules for weather change coefficient inference Fuzzy rules for weather change coefficient inference (a) (b) (c) Figure 9 Fuzzy rules for pre-holiday correction Figure 10 Membership functions for pre-holiday correction (a) (b) 11 Figure 15 Fuzzy rules for post-holiday correction Membership functions for post-holiday correction (a) (b) Figure 15 Main steps of the specific example Figure 13 Time - Figure 14 Hardware configuration of the specific example Predictive treatment water distribution water distribution pump Pump procedure correction quantity (method) 7. Contents of the correction May 8, 1990 (1) In the specification, page 26, line 1, ``Figure 5'' was replaced with ``Figure 5''. 5 and 6”. (2) Figure 6 of the drawings is corrected as shown in the attached sheet. 1999 Patent Application No. 344782 and above 2゜Name of invention Water distribution demand forecasting device 4゜Agent 〒104 5゜Date of amendment order

Claims (1)

[Claims] (1) A device that predicts demand for water distribution amount on a predetermined time basis, comprising a water distribution amount data collection unit that collects water distribution amount data from a water distribution system, a water distribution amount data storage unit that stores past water distribution amount data, and a prediction unit. A weather data input section into which the day's weather data is input; a weather correction calculation section which performs calculations based on the input weather data to obtain a weather correction coefficient that indicates fluctuations in demand for water distribution due to weather conditions; a calendar storage unit that stores the calendar; a holiday correction calculation unit that performs calculations with reference to the calendar and calculates a holiday correction coefficient that indicates a change in demand for water distribution when the predicted date is a holiday or before or after a holiday; A water distribution amount forecasting unit that calculates a predicted water distribution amount by calculating an average water distribution amount from past water distribution amount data and correcting this average water distribution amount based on a weather correction coefficient and a holiday correction coefficient; A demand forecasting device for water distribution amount, comprising: a delay correction unit that performs calculations and delays the predicted water distribution amount for a time period centered on the morning when the predicted day is a holiday.