JP5014522B1 - Inflow amount prediction apparatus, inflow amount prediction method, and program - Google Patents

Abstract

An amount of water flowing into a water storage facility can be predicted in consideration of drought and water loss.
An inflow amount prediction apparatus 10 includes a model database 21 that stores a tank model, and a parameter database 24 that stores a set value of a parameter set in the tank model in association with a low water priority and a water discharge priority point. , An inflow amount prediction unit 113 that calculates a predicted inflow amount based on each of the tank models in which setting values for low water emphasis and outflow emphasis are set as parameters, and a predicted inflow amount output unit 117 that outputs the predicted inflow amount. .
[Selection] Figure 1

Description

  The present invention relates to an inflow amount prediction device, an inflow amount prediction method, and a program.

  For the purpose of water level management in river dam storage facilities and disaster prevention measures for rivers, the amount of water flowing into water storage facilities and the flow rate of water in rivers (hereinafter referred to as inflow) are being predicted. Yes. For example, in Patent Document 1, the trend of forecast error of forecast rainfall is analyzed, the forecast rainfall is corrected using the result, and the future flow is predicted using the current flow rate, the current rainfall, and the corrected forecast rainfall. ing.

JP 2006-92058 A

  In the operation of the water storage facility, if there is a request from the operator to prevent drought that the water of the water storage facility is depleted, if the inflow exceeds the water storage capacity of the water storage facility, it is discharged without generating electricity. There is also a desire to prevent water loss. However, in the conventional technology, prediction is performed so that the error between the predicted value of the inflow amount and the actual value is small (see Patent Document 1), and drought or water loss occurs even if the error as a whole is small. If the error is large at an easy time, a predicted value that does not meet the operator's intention may be calculated.

  The present invention has been made in view of such a background, and an inflow amount prediction device, an inflow amount prediction method, and an inflow amount prediction method capable of predicting the inflow amount of water into a water storage facility in consideration of drought and water loss, and The purpose is to provide a program.

The main invention of the present invention for solving the above problems is an apparatus for predicting the inflow amount of water into a water storage facility, and stores a model for calculating a predicted value of the inflow amount; A parameter storage unit that stores first and second setting values to be set for parameters included in the model, a past record value storage unit that stores past actual value of the inflow amount for each unit period, and the parameter The first coefficient is calculated such that the power of the first coefficient based on the first difference between the first predicted value calculated based on the model in which the first set value is set and the actual value is minimized. 1 is determined, the determined first set value is registered in the parameter storage unit, and the second prediction is calculated based on the model in which the second set value is set as the parameter. The difference between the actual value and the actual value An identification processing unit that determines the second setting value so that the power of the second value larger than the first coefficient is minimized, and registers the determined second setting value in the parameter storage unit When the calculates the first of said predicted value based on a first said model a set value set in the parameter of the second based on the model of the second set value is set to the parameter An inflow amount prediction unit that calculates the prediction value and a prediction value output unit that simultaneously outputs the first and second prediction values are provided.

  According to the inflow amount predicting apparatus of the present invention, the error increases as the predicted value of the inflow amount becomes larger than the predicted inflow amount that is predicted by setting the first set value in the parameter and the case where the first set value is set. It is possible to output the predicted inflow amount predicted by installing the second set value determined to be small. By using the second set value in which the error decreases as the predicted inflow amount increases, the prediction error can be reduced in a period in which the inflow amount such as rainy season increases. Therefore, it is only necessary to refer to one of the two predicted values depending on whether drought prevention is important in the water storage facility, and output a predicted value that is easy for the operator to use. Can do.

  In the inflow amount prediction apparatus of the present invention, the model is a tank model, and the parameter is multiplied by the difference between the tank water depth and the height of the outflow hole, and the water depth of the tank and the height of the outflow hole. A coefficient may be included.

In the inflow amount prediction device of the present invention, the first set value may be less than 1 and the second set value may be 1 or more.

Another aspect of the present invention is a method for predicting the amount of water flowing into a water storage facility, the model for calculating the predicted value of the amount of inflow, and the record of the amount of inflow for each past unit period. A first value difference between the first predicted value and the actual value calculated based on the model in which a first setting value is set in a parameter included in the model. The first set value is determined so that the power of the first coefficient to be minimized is stored, the determined first set value is stored, and the second set value is set as the parameter, based on the model The second set value is determined so that the power of the second value larger than the first coefficient is minimized, with the difference between the second predicted value calculated in the above and the actual value as a base. and, determined by storing the second setting value, the said first set value Calculates a first of said predicted value based on the model set in the parameter, the second of said calculated predicted value based on said second set value to the model set in the parameter, the first And the second predicted value are output simultaneously .

Another aspect of the present invention is a program for predicting an inflow amount of water into a water storage facility, the model for calculating a predicted value of the inflow amount and the inflow amount for each past unit period. A first difference between the first predicted value calculated based on the model in which a first setting value is set in a parameter included in the model and the actual value is stored in a computer storing the actual value of The first setting value is determined so that the power of the first coefficient as a base is minimized, the determined first setting value is stored, and the second setting value is set as the parameter. The second set value so that the power of the second value larger than the first coefficient is minimized, with the difference between the second predicted value and the actual value calculated based on And storing the determined second setting value; Calculates a first of said predicted value based said first set value to the model set in the parameter, the second of the prediction based on the second set value to the model set in the parameter A step of calculating a value and a step of simultaneously outputting the first and second predicted values are executed.

  Other problems and solutions to be disclosed by the present application will be made clear by the embodiments of the invention and the drawings.

  According to the present invention, it is possible to predict the amount of water flowing into a water storage facility in consideration of drought and water loss.

It is a figure for demonstrating a 4-stage tank model. It is a figure which shows the hardware constitutions of the inflow amount prediction apparatus 10 of this embodiment. It is a figure which shows the software structure of the inflow amount prediction apparatus 10 of this embodiment. It is a figure which shows the structural example of the performance database. 3 is a diagram illustrating a configuration example of a condition database 23. FIG. It is a figure which shows the structural example of the parameter database. It is a figure which shows the flow of the identification process of a parameter. It is a figure which shows the flow of the prediction process of inflow amount. It is a figure which shows an example of Table 41 containing the actual value of precipitation, the predicted value of precipitation, the actual value of inflow, and the predicted outflow. It is a figure which shows an example of the graph 42 showing the actual value of precipitation, the actual value of inflow, and the predicted inflow.

  Hereinafter, the inflow amount prediction apparatus 10 according to an embodiment of the present invention will be described. The inflow amount prediction device 10 of the present embodiment predicts the inflow amount of water into a water storage facility such as a dam or the flow rate of water in a river (hereinafter referred to as the inflow amount).

  Specifically, the inflow amount prediction device 10 has a smaller error as a predicted value of an inflow amount calculated based on the model (hereinafter referred to as a predicted inflow amount) becomes a parameter of a model for predicting the inflow amount. Each of the predicted inflows based on the first set value determined to be the same and the second set value determined so that the error decreases as the predicted inflow increases. Are calculated and output simultaneously. As a result, the operator who wants to prevent drought that the water in the water storage facility is depleted uses the model in which the first setting value is set as a parameter, and conversely does not generate power when there is an inflow exceeding the water storage capacity of the water storage facility. An operator who wants to prevent the discharge of water discharged can calculate the predicted inflow using a model in which the second set value is set as a parameter, and can plan the water level of the water storage facility based on this.

  In the present embodiment, a four-stage tank model as shown in FIG. 1 is used as a model for predicting the inflow amount. In addition to the tank model, various models such as a multiple regression model, an AR model, and an ARMA model can be used. Further, the tank model is not limited to four stages, and may be a three-stage tank model.

  In the tank model, the upper tank is associated with the surface outflow, the middle tank is associated with the intermediate outflow, and the two lower tanks are associated with the base outflow (groundwater outflow). The outflow from the outflow hole on the side surface of each tank represents outflow to the river or the like, and the outflow from the infiltration hole on the bottom surface of the tank represents infiltration into the lower aquifer. Outflow from an outflow hole on the side of each tank during a unit period t (in this embodiment, it is 1 day, but can be any value such as 1 hour, 6 hours, or 1 week). The total amount of water used is the predicted value of inflow.

The tank model is represented by the following models (1) to (5). In the following description, the models (1) to (5) are referred to as tank models 1 to 5, respectively.

Here r _t is the precipitation of the date t, assumed given. Q _{i, t} is the outflow amount from the outflow hole of the i-th tank on the date t, and Q _t is the total inflow amount on the date t. H _{i, t} is the water depth of the i-th tank at date t, h _{i, t} is the height of the water stored in the i-th tank without flowing out at date t (reserved height), and E _t is Transpiration amount at date t, a _i is a coefficient for calculating the outflow amount from the i th tank, Z _i is the height of the outflow hole of the i th tank, p _{i, t} are the i th tank at date t And b _i is a coefficient for calculating the penetration amount p _{i, t} of the i th tank. In the case of H _{i, t} ≦ Z _i , Q _{i, t} is assumed to be 0 (zero), and in the case of the first tank, the Manning rule is assumed, and m = 5/3, and other tanks Is m = 1.

In the inflow amount prediction device 10 of the present embodiment, the initial water depth hi _{, 0} , water depth Hi _{, t} , storage height hi _{, t} , transpiration amount _Et , coefficient _ai , outflow, which are parameters included in the tank model. Change the set values of the hole height Z _i , the infiltration amount p _{i, t} , and the coefficient b _i , apply the parameter of the changed set value to the tank model, and calculate the predicted inflow amount given the actual precipitation amount Then, a difference from the actual value of the inflow amount is calculated, and a process (hereinafter referred to as an identification process) for determining an optimum set value according to the difference is performed. The identification process is performed by changing the weighting for the difference. What performed identification processing with weak weighting becomes the first setting value, and what performed identification processing with strong weighting becomes the second setting value.

In the present embodiment, a power value of 2A (corresponding to the first and second coefficients of the present invention) based on the difference between the predicted inflow amount and the actual inflow amount (predicted inflow amount−actual inflow amount) ^2A is The parameter setting value is determined so as to be minimized. In addition to | predicted inflow - the actual flow rate | also how to minimize the ^Å there. The coefficient A is increased when a point that emphasizes prevention of water loss (hereinafter referred to as “important point”) is increased (emphasis on flooding), and when importance is placed on prevention of drought (low water importance). ) Make it small. In the present embodiment, it is assumed that the priority point is any one of “super water discharge priority”, “water discharge priority”, “low water priority”, and “ultra low water priority”.

  FIG. 2 is a diagram illustrating a hardware configuration of the inflow amount prediction device 10 of the present embodiment. The inflow amount prediction device 10 includes a CPU 101, a memory 102, a storage device 103, an input device 105, and an output device 106. The storage device 103 stores various data and programs, for example, a hard disk drive, a flash memory, a solid state disk, a CD-ROM drive, or the like. The CPU 101 implements various functions by reading a program stored in the storage device 103 into the memory 102 and executing it. The input device 105 is, for example, a keyboard, a mouse, a touch panel, or a microphone that inputs data. The output device 106 is, for example, a display, a printer, or a speaker that outputs data.

  FIG. 3 is a diagram illustrating a software configuration of the inflow amount prediction device 10 of the present embodiment. The inflow amount prediction device 10 includes a specification acquisition unit 111, an identification processing unit 112, an inflow amount prediction unit 113, an actual value acquisition unit 114, a condition acquisition unit 115, a precipitation acquisition unit 116, a predicted inflow amount output unit 117, a model database. 21, a performance database 22, a condition database 23, and a parameter database 24. The specification acquisition unit 111, the identification processing unit 112, the inflow amount prediction unit 113, the actual value acquisition unit 114, the condition acquisition unit 115, the precipitation amount acquisition unit 116, and the predicted inflow amount output unit 117 are the inflow amount prediction device 10. The CPU 101 is implemented by reading the program stored in the storage device 103 into the memory 102 and executing it. Further, the model database 21, the performance database 22, the condition database 23, and the parameter database 24 are realized as part of a storage area provided by the memory 102 and the storage device 103.

  The model database 21 stores the tank models 1 to 5 described above.

  FIG. 4 is a diagram illustrating a configuration example of the performance database 22. As shown in FIG. 4, the performance database 22 stores the actual values of precipitation, transpiration, and inflow for each date. In addition, when precipitation, inflow, and transpiration are provided by, for example, the Japan Meteorological Agency or a meteorological company, the data may be acquired and registered in the results database 22, or various types of weather data, rivers, You may make it calculate based on the measured value etc. which were measured in the water storage facility.

  The condition database 23 stores conditions used for parameter setting value identification processing. FIG. 5 is a diagram illustrating a configuration example of the condition database 23. In the example of FIG. 5, the condition database 23 stores a parameter identification condition 231 and a parameter constraint condition 232.

  The parameter identification condition 231 includes the value of the coefficient A in association with each importance point. In the present embodiment, “2” is set to the largest value for the coefficient A at the priority point for “super-watering emphasis”, and “2” is set to the second largest value for the coefficient A at the point emphasizing “watering emphasis”. In the importance point of “emphasis on low water”, the second smallest “0.25” is set in the coefficient A, and in the importance point of “ultra low water importance”, the coefficient A has the smallest “0.1”. "Is set. In addition, it is not restricted to these values, for the emphasis point on emphasizing water discharge, a value larger than 1 is set to the coefficient A so as to increase according to the degree of emphasis on prevention of water loss, and the emphasis point on emphasis on low water The coefficient A may be set to a value smaller than 1 so as to decrease according to the degree of importance on drought prevention.

The parameter constraint condition 232 includes the above-described parameters (water depth H _{i, t} , storage height h _{i, t} , transpiration amount E _t , coefficient a _i , outflow hole height Z _i , permeation amount p _{i, t} , coefficient b _i. ) Includes the lower limit value and the upper limit value.

  FIG. 6 is a diagram illustrating a configuration example of the parameter database 24. As shown in the figure, the parameter database 24 stores the setting values of each parameter determined by the parameter identification process for each important point.

  The specification acquisition unit 111 acquires specifications necessary for calculation of identification processing and inflow amount prediction processing. In the present embodiment, the specification acquisition unit 111 acquires the water collection area of the tank and the start date and end date of the period to be calculated for the identification process and the inflow amount prediction process. These specifications may be received by the specification acquisition unit 111 from the user, or each specification may be stored in the memory 102 or the storage device 103. The specification acquiring unit 111 may use the oldest date stored in the result database 22 as the start date. The specification acquisition unit 111 may use the latest date stored in the result database 22 as the end date for the identification process.

The identification processing unit 112 performs identification processing for a specified target period (hereinafter referred to as an identification target period). As described above, the identification processing unit 112 includes the water depth H _{i, t} , the storage height h _{i, t} , the transpiration amount E _t , the coefficient a _i , and the outflow hole height Z _i that are parameters included in the tank models 1 to 5. The predicted inflow amount is calculated by a simulation that changes the set values set in the permeation amount p _{i, t} and the coefficient b _i , and the optimum parameter is determined according to the difference between the predicted inflow amount and the actual inflow amount. Incidentally, transpiration E _t may be calculated from the precipitation without the conversion by simulation. Details of the identification process will be described later. The identification processing unit 112 performs the above simulation for each important point, and registers the determined parameter setting value in the parameter database 24 in association with the important point.

  The actual value acquisition unit 114 acquires actual values such as inflow, precipitation, and transpiration during the identification target period. In the present embodiment, the actual value acquisition unit 114 acquires these actual values from the actual database 22, but for example, acquires actual values from computers operated by the Japan Meteorological Agency, weather companies, weather stations, and the like. May be.

  The condition acquisition unit 115 acquires various conditions used for the identification process. In the present embodiment, the condition acquisition unit 115 acquires the parameter identification condition 231 and the parameter constraint condition 232 from the condition database 23. For example, the condition acquisition unit 115 receives the coefficient A, the lower limit value and the upper limit value of the parameter from the user. May be.

  The inflow amount prediction unit 113 performs inflow amount prediction processing for a designated target period (hereinafter referred to as a prediction target period). The inflow amount prediction unit 113 performs inflow amount prediction processing for at least two important points.

  The precipitation amount acquisition unit 116 acquires the precipitation amount in the prediction target period. When the precipitation corresponding to the date included in the prediction target period is registered in the performance database 22, the precipitation acquisition unit 116 reads the actual rainfall value from the performance database 22, and the other dates. For, accept the input of predicted precipitation. The precipitation amount acquisition unit 116 can acquire predicted precipitation amount provided as a weather forecast from a computer operated by, for example, the Japan Meteorological Agency or a weather company.

  The predicted inflow amount output unit 117 outputs the predicted inflow amount calculated by the inflow amount prediction unit 113 for each importance point.

  Hereinafter, the process performed in the inflow amount prediction apparatus 10 will be described.

FIG. 7 is a diagram showing a flow of parameter setting value identification processing.
The specification acquisition unit 111 acquires specifications necessary for the identification process (S301). As described above, the specification acquisition unit 111 acquires the water collection area of the tank and the start date and end date of the identification target period. The specification acquisition unit 111 also reads tank models 1 to 5 from the model database 21.

  The actual value acquisition unit 114 reads the actual values of the inflow, precipitation, and transpiration amount stored in the actual database 22 (S302), and the condition acquisition unit 115 stores the parameter identification condition 231 stored in the condition database 23. Then, the parameter constraint condition 232 is read (S303).

The identification processing unit 112 performs the following processing on the four important points of “super water outflow emphasis”, “high water outflow emphasis”, “low water emphasis”, and “ultra low water emphasis”. That is, the identification processing unit 112 includes the water depth H _{i, t} , the storage height h _{i, t} , the transpiration amount E _t , the coefficient a _i , the outflow hole height Z _i , and the infiltration amount, which are parameters included in the tank models 1 to 5. The values of p _{i, t} and coefficient b _i are changed in the range from the lower limit value to the upper limit value included in the parameter constraint condition 232, and the changed values are set in the tank models 1 to 5 and read from the results database 22 the precipitation of the actual value of precipitation r _i, continue to calculate the predicted flow rate Q _t is given as _t, and the bottom and the predicted flow rate Q _t, the difference between the actual inflow of date t, emphasis point The set value of each parameter is set so that the sum Σ (predicted inflow amount−actual value) ^2A of the power of the value obtained by doubling the coefficient A corresponding to 2 from the start date to the end date of the identification target period is minimized. Determine (S305). Incidentally, without changing the transpiration E _t, it may be the actual value of the transpiration amount read from the result database 22 to set the tank model 1 to 5 as a transpiration rate E _t.

  The identification processing unit 112 registers the determined setting value of each parameter in the parameter database 24 in association with the importance point (S306).

  As described above, for each priority point, the optimum parameter setting value corresponding to the priority point is registered in the parameter database 24.

FIG. 8 is a diagram illustrating a flow of inflow amount prediction processing.
The specification acquisition unit 111 acquires specifications necessary for the inflow amount prediction process (S321). As described above, the specification acquisition unit 111 acquires the water collection area of the tank and the start date and end date of the prediction target period. The specification acquisition unit 111 also reads tank models 1 to 5 from the model database 21.

  The actual value acquisition unit 114 reads the actual values of the inflow, precipitation, and transpiration amount corresponding to the date included in the prediction target period from the result database 22 (S322). The precipitation amount acquisition unit 116 acquires the predicted value of precipitation amount for dates that are not registered in the results database 22 (S323).

The inflow amount prediction unit 113 receives input of two or more important points (S324). In the following description, as an example, it is assumed that two emphasis points “emphasis on flooding” and “emphasis on low water” are designated. The inflow amount predicting unit 113 reads the set value of the parameter corresponding to the important point from each of the received important points from the parameter database 24 (S325), and sets the read set value to each parameter of the tank models 1 to 5. , the tank model 1 to 5 set to calculate the predicted outflow Q _t actual values or predicted values of precipitation from within the start and end dates of the prediction period given as precipitation r _t (S326).

FIG. 9 shows the actual value of precipitation and the actual value of inflow read by the actual value acquisition unit 114, the predicted value of precipitation acquired by the precipitation acquisition unit 116, and the predicted outflow Q calculated by the inflow prediction unit 113. _This is an example in which _t is in the format shown in Table 41.

  The predicted inflow amount output unit 117 outputs the predicted inflow amount calculated by the inflow amount prediction unit 113 for each importance point (S327). For example, the predicted inflow output unit 117 may output the table 41 shown in FIG. 9, or may output a graph based on the table 41. FIG. 10 is a diagram illustrating an example of the graph 42 that represents the actual value of precipitation, the actual value of inflow, and the predicted inflow of “emphasis on flooding” and “emphasis on low water”.

  Thus, in the inflow amount prediction device 10 of the present embodiment, the inflow amount predicted by the tank model optimized for each of a plurality of important points can be displayed at the same time. , You can refer to the predicted inflow amount that matches the priority point that you focus on.

  Moreover, in the inflow amount prediction apparatus 10 of this embodiment, the value of the coefficient A is 1 or more in the case of emphasizing water discharge (including emphasis on super-water discharge), and emphasis on low water (including emphasis on ultra-low water). In this case, since the value of the coefficient A is less than 1, if the value of predicted inflow and actual inflow is large when the emphasis is on water discharge, the error is evaluated larger than the case of emphasizing low water, and the error is minimized. The parameter setting values are determined so that Therefore, in the case where emphasis is placed on water discharge, the parameter setting values can be determined so that the error at the time when the inflow is large is smaller than in the case where emphasis is on low water. Therefore, in the case of emphasizing water discharge, the prediction accuracy of the inflow amount at the time when the inflow amount is large can be improved as compared with the case of emphasizing low water.

  In the present embodiment, the inflow prediction device 10 is a single computer. However, the present invention is not limited to this, and the inflow amount prediction device 10 may be configured as a system including a plurality of computers. For example, each database of the model database 21 to the parameter database 24 may be managed by a database server and accessed from the inflow amount prediction device 10. Further, for example, the identification process and the inflow amount prediction process may be executed by different computers. In this case, one computer includes a specification acquisition unit 111, an identification processing unit 112, a performance value acquisition unit 114, and a condition acquisition unit 115, and the other computer includes a specification acquisition unit 111, an inflow amount prediction unit. 113, the precipitation amount acquisition part 116, and the prediction inflow amount output part 117 can be provided.

  Moreover, in the inflow amount prediction apparatus 10 of the present embodiment, the values of all parameters are changed, but the identification process may be performed by changing only the values of some parameters.

  Further, in the present embodiment, “super flooding emphasis” and “outflow emphasis” are described as examples of emphasis points on emphasizing prevention of water loss, but only “outflow emphasis” may be used. Moreover, you may provide three or more important points according to the degree which attaches importance to prevention of water loss. In this case, the coefficient A is set to a large value according to the degree of emphasizing prevention of water loss.

  Similarly, in the present embodiment, “ultra low water emphasis” and “low water emphasis” are described as examples of importance points emphasizing prevention of drought, but only “low water emphasis” may be used. Three or more importance points may be provided depending on the degree of importance on drought prevention. In this case, the coefficient A is set to a small value according to the degree of importance on prevention of drought.

  Although the present embodiment has been described above, the above embodiment is intended to facilitate understanding of the present invention and is not intended to limit the present invention. The present invention can be changed and improved without departing from the gist thereof, and the present invention includes equivalents thereof.

10 Inflow prediction device 101 CPU
DESCRIPTION OF SYMBOLS 102 Memory 103 Memory | storage device 105 Input device 106 Output device 111 Specification acquisition part 112 Identification processing part 113 Inflow amount prediction part 114 Actual value acquisition part 115 Condition acquisition part 116 Precipitation amount acquisition part 117 Predicted inflow amount output part 21 Model database 22 Actual Database 23 Condition database 231 Parameter identification condition 232 Parameter constraint condition 24 Parameter database

Claims (5)

A device for predicting the amount of water flowing into a water storage facility,
A model storage unit for storing a model for calculating a predicted value of the inflow amount;
A parameter storage unit for storing first and second setting values to be set for parameters included in the model;
An actual value storage unit that stores the actual value of the inflow amount for each past unit period;
The power of the first coefficient based on the first difference between the first predicted value calculated based on the model in which the first set value is set as the parameter and the actual value is minimized. The first setting value is determined as described above, the determined first setting value is registered in the parameter storage unit, and the first setting value is calculated based on the model in which the second setting value is set as the parameter. The second set value is determined so that the power of the second value larger than the first coefficient is the minimum, based on the difference between the second predicted value and the actual value, and the determined An identification processing unit for registering a second set value in the parameter storage unit;
Calculates a first of said predicted value based said first set value to the model set in the parameter, the second of the prediction based on the second set value to the model set in the parameter An inflow predictor for calculating a value ;
A predicted value output unit that simultaneously outputs the first and second predicted values;
An inflow amount prediction apparatus comprising: The inflow amount prediction device according to claim 1,
The model is a tank model;
The parameters include the depth of the tank, the height of the outflow hole, and a coefficient that multiplies the difference between the depth of the tank and the height of the outflow hole.
An inflow amount predicting device characterized by the above. The inflow amount prediction device according to claim 1 or 2,
The first set value is less than 1 and the second set value is 1 or more;
An inflow amount predicting device characterized by the above. A method for predicting the amount of water flowing into a water storage facility,
A computer that stores a model for calculating the predicted value of the inflow amount and the actual value of the inflow amount for each past unit period ,
A power of a first coefficient based on a first difference between the first predicted value and the actual value calculated based on the model in which a first setting value is set in a parameter included in the model is The first setting value is determined to be the minimum, the determined first setting value is stored, and the second setting value calculated based on the model in which the second setting value is set as the parameter The second set value is determined such that the power of the second value larger than the first coefficient is minimized, with the difference between the predicted value and the actual value as a base, and the determined second Memorize the set value,
Calculates a first of said predicted value based said first set value to the model set in the parameter, the second of the prediction based on the second set value to the model set in the parameter Calculate the value ,
Outputting the first and second predicted values simultaneously ;
An inflow rate prediction method characterized by A program for predicting the amount of water flowing into a water storage facility,
In a computer that stores a model for calculating the predicted value of the inflow rate and the actual value of the inflow rate for each past unit period ,
A power of a first coefficient based on a first difference between the first predicted value and the actual value calculated based on the model in which a first setting value is set in a parameter included in the model is The first setting value is determined to be the minimum, the determined first setting value is stored, and the second setting value calculated based on the model in which the second setting value is set as the parameter The second set value is determined such that the power of the second value larger than the first coefficient is minimized, with the difference between the predicted value and the actual value as a base, and the determined second Storing a set value; and
Calculates a first of said predicted value based said first set value to the model set in the parameter, the second of the prediction based on the second set value to the model set in the parameter Calculating a value ;
Outputting the first and second predicted values simultaneously ;
A program for running

Images