JP2001167078A - Flow predicting method for dam or river - Google Patents

Abstract

PROBLEM TO BE SOLVED: To grasp or predict the increase/decrease tendency of a flow with a simple method and to construct a predictive model exactly reflected with basin characteristics such as flow-down time. SOLUTION: Concerning the flow predicting method for predicting the flow of a dam or river from a precipitation in an upper basin while using the predictive model constructed by a computer, the graph of flow result values for a prescribed period in the past is displayed on a display screen with that of present time as a reference, cumulative precipitations in the upper basin for this prescribed period are displayed while being overlapped on the flow result values, and the increase/decrease tendency of the flow is predicted from the correlative relation of the cumulative precipitations and the flow result values.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for automatically predicting the flow rate of a dam or a river on a computer in a system control station, a power supply command station, a dam management station, a hydroelectric power station, or the like.

[0002]

2. Description of the Related Art Prediction of the inflow of a dam and the flow rate of a river upstream of the dam (hereinafter, both flows are simply referred to as flow rate as needed) are important for safety and economy of dam operation.
Estimation of the discharge several hours ahead is necessary to properly determine the discharge of the dam during rainfall, and can help ensure the safety of the basin. In addition, prediction from several hours ahead to several days ahead, especially the next day's flow forecast, can make an appropriate power generation plan and contribute to effective use of hydro energy and improvement of economic efficiency.

Conventionally, this kind of flow rate prediction is often performed according to the experience and intuitive knowledge of a skilled operator. For this reason, various methods using a tank model, a storage function method, a neural network, and the like have been proposed as examples of automating the prediction operation up to several hours ahead, and the prediction accuracy has been improving year by year. By the way, when creating a dam flow prediction model, it is essential to analyze the characteristics of rivers and basins, such as topography and geology, in other words, it is indispensable to analyze how the rainfall in the upstream basin is reflected in the flow. . In particular,
The flow-down time, which is the time required for the rainfall in the upstream area to be reflected in the flow rate of the dam, etc. (the time until the flow rate increases after rainfall) is an important factor. The difference between the peak time and the peak time of the flow rate was determined, and this was defined as the flow time.

In general, a plurality of rain gauges are used for estimating a flow rate. Since the information of the rain gauge has a large variation, it is common to average the measured values of a plurality of rain gauges and use them. At this time, when there are a plurality of rivers upstream, a method of calculating and using the average rainfall for each river as described later (FIG. 12 described later), and a method of using the basin average rainfall summarizing all the basins are described below. (FIG. 13 described later)
And is known.

[0005]

By the way, operators who predict the flow rate of dams need enormous expertise and many years of experience, but in recent years the number of skilled operators having this knowledge has been steadily decreasing. ing. On the other hand, flow prediction is the foundation of dam operation, and improvement of prediction accuracy and automation are keenly desired. In recent years, prediction accuracy has been significantly improved by new technologies such as neural networks, and even when using existing tank models, it has become possible to perform high-precision prediction by making the configuration more complex. I have. However, while these methods have high prediction accuracy, it is difficult to understand why the predicted value was obtained because of the complexity of the configuration.

In particular, short-term dam flow prediction is directly related to dam gate operation, and it is required to clarify the reason for the prediction and the reason for gate operation due to safety problems in the downstream area. Therefore, there is a need for a method that can be easily understood and that can clearly indicate the basis of the predicted value. Also, when constructing a prediction model, it is necessary to analyze the characteristics of the target basin, and it is especially important to calculate the flow time. In the conventional method, the flow time is not always required, and its analysis accuracy is not high.
That is, in the conventional method, only the difference between each peak time of the rainfall amount and the flow rate is obtained. However, when the rainfall peak time is not clear, it is impossible to calculate the falling time, and the peak time is not clear. In such cases, automatic calculation by a computer was difficult.

In a dam into which a plurality of rivers flow,
In the case of FIG. 12 in which the flow rate is predicted for each river and the respective flow rates are summed to obtain the dam flow rate, the average rainfall for each river 22 is obtained by the rain gauge at the stage, and the flow rate of each river 22 is predicted. The flow rate of the dam 24 is predicted by summing the flow rates of the respective rivers 22 obtained by the above. The flow meter 2
Reference numeral 3 is used for acquiring the actual value. According to this method, since the flow rate of the dam 24 is predicted based on the rainfall of the river basin of the target river, the error of the predicted value of the flow rate may increase due to the dispersion of the measured values of the individual rain gauges.

In order to solve this problem, there is a method of first calculating the average rainfall on the assumption that the flow times of all the basins are the same, and then directly predicting the dam discharge. FIG. 13 shows the situation, in which the dam discharge is directly predicted from the average rainfall in the entire basin. However, in this case, when the flow times of the respective rivers 22 are different, the inflow times to the dam 24 are shifted, so that the prediction accuracy tends to be low.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and it is possible to grasp and predict a tendency of a flow rate to increase or decrease by a very simple method, and to accurately reflect a basin characteristic such as a flowing time of each river. An object of the present invention is to provide a flow rate prediction method that can predict with high accuracy by using the prediction model.

[0010]

Means for Solving the Problems To solve the above problems, the invention of claim 1 firstly uses a prediction model constructed by a computer to predict a flow rate of a dam or a river from rainfall in an upstream area. In the method, the actual flow rate value of the past predetermined period based on the present time is graphically displayed on a display screen, and at least the accumulated rainfall in the upstream area of the predetermined period is displayed on the actual flow rate value superimposed, and the accumulated rainfall is displayed. The trend of the flow rate is predicted from the correlation with the actual flow rate value.

According to a second aspect of the present invention, there is provided a flow rate prediction method for predicting a flow rate of a dam or a river from rainfall in an upstream region using a prediction model constructed by a computer. Cumulative rainfall c in the upper reaches
Is calculated, and the predicted flow rate d is calculated by d = current flow rate + A · Δc + B (where A and B are constants).

According to a third aspect of the present invention, the flow rate of a dam or a river is predicted using a prediction model constructed by a computer, while taking into account the flow time, which is the time required for the rainfall amount in the upstream area to be reflected in the flow rate. In the flow rate prediction method, the correlation coefficient of both patterns is calculated while moving the time change pattern of rainfall in the upstream area along the time axis in the direction of the time change pattern of the actual flow rate value thereafter, and the correlation coefficient is maximized. The time corresponding to the amount of movement of the time change pattern of the amount of rainfall at the time of falling is defined as the falling time.

According to a fourth aspect of the present invention, the flow rate of a dam or a river is calculated by using a prediction model constructed by a computer and taking into consideration a flow time, which is a time until rainfall in a plurality of upstream areas is reflected in the flow rate. In the predicted flow rate prediction method, the flowdown time is calculated for each basin according to the invention of claim 3, and the basin average rainfall is calculated by using the average rainfall of each basin and the area of each basin in consideration of these flowdown times. Then, the flow rate is predicted using this basin average rainfall.

According to a fifth aspect of the present invention, there is provided a flow rate prediction method for predicting a flow rate of a dam or a river from a cumulative rainfall amount in a predetermined period in an upstream area using a prediction model constructed by a computer. While moving the time change pattern of the cumulative rainfall in the direction of the time change pattern of the actual flow rate value thereafter along the time axis, calculate all the correlation coefficients of both patterns, and the cumulative rainfall when the correlation coefficient becomes the maximum The accumulated time of the time change pattern is searched, and the accumulated rainfall of the accumulated time is used for the flow rate prediction.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. In general, even if there is rainfall when the ground in the upstream area is dry, the rainfall is absorbed by the ground and does not flow into the dam, and when the ground is wet, such as after a long rain, even a slight rainfall Dam inflow will increase. Thus, the relationship between the dam inflow and the accumulated rainfall due to long-term (over several hours to several days) rainfall is often more important than the relationship with the instantaneous rainfall. Therefore, according to the first aspect of the present invention, the tendency of the increase and decrease of the flow rate is grasped by paying attention to the relationship between the accumulated rainfall rate and the flow rate of a dam or the like, and used for prediction.

The embodiment of the present invention mainly comprises two steps, a step 1 for calculating an optimum cumulative time of the accumulated rainfall, and a step 2 for displaying the accumulated rainfall together with the inflow actual value. In the step 1 for calculating the accumulated time, it is first determined how many hours of rainfall should be accumulated from the past several hours to several months or from similar past flood cases, for example, by using the invention of claim 5 described later. Is calculated.
Next, the flow-down time is calculated for the accumulated rainfall calculated according to the calculated cumulative time. Here, there are various methods for calculating the flow time. For example, a good result can be obtained by using the invention of claim 3 described later. Although the calculation of this flow-down time is not an essential requirement, when the cumulative rainfall and the dam inflow are superimposed and displayed in consideration of the flow-down time, it is considered as the time for shifting the dam inflow along the time axis. .

Next, the process proceeds to step 2 for displaying. Here, at least the accumulated rainfall during the same period is displayed together with the actual flow rate value for the past predetermined period based on the current time.
If necessary, the rainfall amount for each predetermined period from the past to the present, the rainfall forecast value for several hours from the present time, and the future cumulative rainfall forecast value are displayed on the screen. Even when only the accumulated rainfall for the predetermined period up to the present time is displayed, it is possible to grasp the inflow increase / decrease tendency several hours ahead from the correlation between the accumulated rainfall and the actual amount of inflow. By also displaying the forecast value and the future cumulative rainfall forecast value calculated from the rainfall forecast value, it is possible to further grasp the increase / decrease tendency of the future inflow rate.

FIG. 1 shows a display example of a screen according to the present embodiment. As shown in the figure, the inflow amount actual value in the past predetermined period up to the present time is graphically displayed as a thick solid line on the display screen, and the accumulated rainfall in the same period is displayed as a thin broken line. Here, the vertical axis of the graph is obtained by normalizing the actual value on an appropriate scale (the same applies hereinafter). Normally, depending on the case of flooding, the accumulated time having a high correlation with the inflow is different. Therefore, first, as described above, the accumulated time is calculated according to the procedure of the invention of claim 5, and the calculation result is displayed. In the example of FIG. 1, the accumulated time is set to 10 hours. Further, the horizontal axis of FIG. 1 indicates the passage of time in the past 16 days. The specific procedure for calculating and displaying the accumulated time is as follows.

(1) Calculation of the cumulative time The cumulative time is calculated from the rainfall amount and the flow rate (both actual values) for the past several hours. Here, the actual rainfall value is the basin average rainfall (basin basin average rainfall in consideration of the downflow time) determined by the invention of claim 4, and the accumulated time is determined by the invention of claim 5. The calculation of the accumulated time is performed at regular intervals such as every hour or every day.

(2) Display As shown in FIG. 1, the actual inflow amount and the accumulated rainfall are displayed on the screen in a superimposed manner. If the flowing time is calculated to be one hour, the pattern of the dam inflow may be moved to the right along the time axis by one hour relative to the pattern of the accumulated rainfall.

Next, an embodiment of the present invention will be described. The above-described invention of claim 1 is effective for grasping the increasing / decreasing tendency of the flow rate, but is not suitable for predicting the flow rate itself. Therefore, in the invention of claim 2, the flow rate itself is predicted by a simple method utilizing the fact that the correlation between the change in the accumulated rainfall and the change in the flow rate is high. The present invention shows a prediction basis for an unskilled operator, and uses the following simple formula.

In general, in a dam or a river, when the change in the accumulated rainfall c is Δc, the change in the flow rate has a relationship of A = Δc + B with the change in the flow rate. Focusing on this relationship, the flow rate prediction value d is expressed by the following equation. d = current flow rate + flow rate change = current flow rate + A · Δc + B where A and B in the formula are constants, which can be obtained by using the least square method or the like based on past flood cases at the time of constructing a prediction model. Each time a prediction is made, a value calculated from the latest actual value may be used. Therefore, the flow rate prediction value is calculated from the current flow rate and the change in the flow rate by using the above formula.

2 and 3 are views for explaining the embodiment of the present invention. FIG. 2 is a display screen of the actual inflow amount and the accumulated rainfall corresponding to FIG. 1, and FIG. 3 is an embodiment of the present invention. 6 is a display screen of a predicted value and an actual value of the inflow amount according to FIG. As can be seen from these comparisons, in the display example of FIG. 2, the first inflow peak value almost coincides with the accumulated rainfall peak value, but the second and third inflow peak values and the accumulated rainfall peak value are almost the same. Does not match the value. The second and third inflow peak values and the accumulated rainfall peak values should also be essentially the same. In other words, according to the first aspect of the present invention, when the inflow amount is predicted based on the accumulated rainfall, a prediction error occurs. Will occur. Therefore, in the invention of claim 2, the flow rate prediction without error is performed by using the accumulated rainfall (cumulative rainfall variation).

The equation for predicting the flow rate is the above equation, d = current flow rate + flow rate change = current flow rate + A · Δc + B. Here, the flow rate change is treated as the flow rate change predicted value. Also in this embodiment, the falling time is taken into consideration. For example, if the flowing time is one hour, data that is one hour older is used as the accumulated rainfall change Δc. FIG. 3 shows a predicted value and an actual value of the inflow amount according to the present embodiment. According to this figure, the peak value of the predicted value and the peak value of the actual value almost match.

Next, a third embodiment of the present invention will be described. In general, the parameters of many prediction models require the flow time as the time until the rainfall in the upstream area is reflected on the flow rate of dams, etc., so the analysis of river flow time is indispensable for the construction of a prediction model. . However, it is difficult to automate the method of calculating the difference between the peak time of the rainfall and the peak time of the flow rate, and since the method focuses only on the peak time, the calculation accuracy of the flowing time is not high. Therefore, in the present invention, the falling time is automatically calculated from the correlation coefficient between the rainfall amount pattern and the flow rate pattern.

FIG. 4 shows the principle of the present embodiment. As shown in the figure, the flow rate changes with a delay in the amount of rainfall in the upstream area, but excluding this time delay, the shape of the flow rate pattern is not extremely different from the shape of the rainfall amount pattern. Therefore, as shown in FIG. 4, the rainfall amount pattern is gradually moved in the direction of the arrow, that is, in the direction of delaying the time, until the time coincides with the flow rate pattern best, and the amount of movement (movement time) in the time axis direction is calculated. Time.

At this time, a correlation coefficient is used as an index indicating the degree of coincidence between the rainfall pattern and the flow rate pattern.
The correlation coefficient is an index of linearity that changes in the range of −1 to 1, and is set to 1 if the shape of the rainfall pattern and the flow rate pattern at the time of movement are similar, and is set to 0 if they do not completely match. An appropriate flowing time can be calculated from the amount of movement of the rainfall amount pattern in the time axis direction when the correlation coefficient indicates the maximum value within the above range. These processes are automatically executed by a computer.

FIG. 5 shows a correlation coefficient between a moving time of a rainfall amount pattern and a flow rate pattern for calculating a flowing time of one river in a certain flooding case. In this example,
Since the correlation coefficient is substantially maximum when the moving time is about 4 hours, the flow-down time is calculated to be 4 hours. That is, in this river, it is considered that the flow time from the rainfall in the upstream area to the increase of the dam inflow is 4 hours, and this time may be used for the flow rate prediction.

Next, an embodiment of the present invention will be described. This embodiment assumes that three rivers with different flowing times are flowing into the dam, and calculates the basin average rainfall of these three rivers. When there are a plurality of rivers flowing into the dam, the flooding characteristics (characteristics due to the topography, geology, etc.) of each river are different. For example, in some rivers, the water level rises immediately after rainfall and the inflow to the dam increases.
In some other rivers, it takes considerable time for the inflow to the dam to increase. These differences in time can also be regarded as differences in flow time.

In the embodiment of the present invention, first, the flow time of each river is calculated by applying the above-described embodiment of the third aspect of the present invention. At this time, since the flowing time changes for each flooding case, the average value of the flowing time of many cases is taken for each river. Table 1 shows each flood case and the flow time for each river, and the average value of the flow time is 4.2 hours for river 1, 5.6 hours for river 2, and 3.4 hours for river 3. .

[0031]

[Table 1]

Here, the basin average rainfall of each of the rivers 1 to 3 is calculated according to the obtained flowing time. In this embodiment, if the rainfall information is data in units of one hour, ignoring the decimal point, it is assumed that the river 1 has a flow time longer than the river 3 by about one hour and the river 2 has a flow time longer than the river 3 by about 2 hours. And the basin average rainfall is calculated by the following formula. Basin average rainfall = (basin area of river 1 x 1 of river 1 basin)
Rainfall before time + basin area of river 2 × rainfall two hours before basin of river 2 + basin area of river 3 × current rainfall of basin of river 3) / basin area of all rivers The basin average rainfall is calculated using the weighted average. By using the basin average rainfall obtained in this way, the amount of water flowing into the dam from the rivers 1 to 3 almost at the same time can be predicted, and the flow rate prediction without time delay or inconsistency becomes possible. The basin average rainfall calculated in this way can be applied, for example, in the case of obtaining the accumulated rainfall or the accumulated rainfall change in the inventions of claims 1 and 2.

Next, an embodiment of the invention will be described. The present invention is for calculating an accumulated time having a high correlation with a dam flow rate, that is, an accumulated rainfall. If the accumulated time is too long or too short, the correlation with the flow rate is low and cannot be used for predicting the flow rate. In the present invention, an appropriate cumulative time is calculated by using a correlation coefficient between the cumulative rainfall and the flow rate.

In the following, the term “time cumulative rainfall” means not the cumulative rainfall from the start of rainfall but the hourly cumulative rainfall from n hours to m hours before the river basin. For example, the one-hour accumulated rainfall is the accumulated rainfall from n hours ago to (n-1) hours ago, and the two-hour accumulated rainfall is n hours ago (n-
2) Cumulative rainfall up to the previous hour. Therefore, the two-hour accumulated rainfall based on n hours ago includes the one-hour accumulated rainfall.

FIGS. 6 (a) and 6 (b) are diagrams showing the time change of the accumulated rainfall and the flow rate for calculating the accumulated time of the rainfall. Generally, the relationship between the accumulated rainfall and the flow rate changes depending on the rainfall conditions. That is, when the ground is dry, the flow rate is highly correlated with the accumulated rainfall for a long time (long cumulative time), and when the ground is wet, the flow rate is related to the accumulated rainfall for a short time (short cumulative time). Tends to increase. In addition,
The one-hour cumulative rainfall, the two-hour cumulative rainfall, and the flow rate in FIGS. 6A and 6B are all actual values, and are calculated for the past one hour (FIG. 6A) or two hours (FIG. 6B )) Shows that the change in the flow rate appears later than the change in the accumulated rainfall.

The calculation of the accumulated time in this embodiment is performed according to the following procedure. (1) The one-hour cumulative rainfall pattern shown in FIG.
While moving one hour at a time in the flow pattern direction along the time axis, the correlation coefficient of both patterns is determined. That is, FIG.
And the correlation coefficient is also obtained as an index of linearity that changes in the range of -1 to 1. 1 in this case
The moving time of the pattern of the time accumulated rainfall corresponds to the above-mentioned falling time.

(2) Similarly, the two-hour cumulative rainfall shown in FIG. 6B, the three-hour cumulative rainfall (not shown), the four-hour cumulative rainfall,... It is moved by time, and the correlation coefficient between the flow pattern and the flow pattern is obtained. How many types of cumulative rainfall patterns are used for the cumulative time (how long the above-mentioned n is up to how many hours)
Is appropriately determined based on experience and the like. (3) Next, the maximum value is selected from the obtained correlation coefficients, and the accumulated time corresponding to the maximum value is searched. That is, the "-time" of the time-to-time accumulated rain amount at which the correlation coefficient becomes the maximum is the accumulated time as it is.

For example, when the correlation coefficient obtained by moving the accumulated rainfall for one hour to 12 hours in a certain river basin one hour at a time is shown in Table 2, the moving time is one hour for the 10-hour accumulated rainfall.
The correlation coefficient “0.934” at the time is the largest among all the correlation coefficients. Therefore, for this river, the accumulated time for calculating the accumulated rainfall is set to 10 hours. Thereafter, using the accumulated rainfall for 10 hours, the flow rate is predicted in consideration of the flow-down time.

[0039]

[Table 2]

Here, FIG. 7 shows the average rainfall obtained by simply averaging the measured values of a plurality of rain gauges installed in the upstream area of the dam in order to explain the effectiveness of the inventions of claims 3 to 5. FIG. 3 is a diagram in which the actual values of the dam inflow are superimposed. The actual inflow amount value and the rainfall amount, which are the basis of the display in FIG. 7, are the same as those in the example of FIG. Because the rain falls and stops and varies widely,
A simple averaging of the rain gauge readings does not eliminate the variation. As can be seen from FIG. 7, the average rainfall may decrease even when the inflow is increasing, and it is difficult to grasp the increasing or decreasing tendency of the inflow from the average rainfall.

On the other hand, according to the invention of claim 3, the flow time for each river basin is automatically calculated, and according to the invention of claim 4, the basin average rainfall is calculated from the weighted average weighting the basin area. According to the fifth aspect of the present invention, if the accumulated rainfall is converted into the accumulated rainfall and displayed, the correlation between the change in the accumulated rainfall and the increasing / decreasing tendency of the inflow becomes clear, and the increasing / decreasing tendency of the inflow from the accumulated rainfall can be easily determined. Can be used for prediction of flow rate.

Finally, a prediction result when the embodiment of the fifth aspect of the present invention is applied to an inflow rate prediction model will be examined. Here, a neural network was used as an inflow estimation model. When applying the neural network to inflow estimation, rainfall information for a plurality of hours is required. In addition, as a method of determining the input time of how many hours of rainfall data to use, a method empirically determined by an operator (see the 1993 IEEJ Power and Energy Competition N
O.185, Ichiyanagi et al.), A method to determine the rainfall time (Japanese Patent Application
10-187108), there is a method of determining the cumulative rainfall time having a high correlation with the flow rate (the invention of claim 5).

Here, two examination cases 1 and 2 will be examined. In the study example 1, using the prediction model shown in FIG. 8, a basin average rainfall difference is input to the input layer I every six hours before the current time every one hour. It is to be noted that II is an intermediate layer, and III is an output layer that outputs a change in the inflow amount one hour after the present time (corresponding to A.Δc + B, which is the above-mentioned change in flow rate), as a predicted value. The time up to 6 hours before is based on the flow time in the target basin and the empirical value of the operator.

In the study example 2, using the prediction model shown in FIG. 9, the difference of the basin average rainfall is input every hour from the present time to 20 hours before. The value of 20 hours is the cumulative time calculated according to the invention of claim 5. The cumulative time differs depending on the flood case, but here, the longest time was selected from past cases. In an average case, the value is about 10 hours as shown in Table 2 above. Table 3 summarizes the input and output factors of the two prediction models in FIGS.
Under these input-output factors, past cases were sufficiently learned and two flood cases were predicted.

[0045]

[Table 3]

In Table 3, the basin average rainfall in the case of obtaining the basin average rainfall difference is a value which is not adjusted by the flowing time when there are a plurality of rivers. In other words, it is a value obtained by adding all values obtained by multiplying the rainfall amount of each river at the same time by the basin area of each river, and dividing the value by the total basin area. Also, for example, the basin average rainfall difference up to one hour ago is a value obtained by subtracting the basin average rainfall one hour ago from the current basin average rainfall.

FIG. 10A shows the predicted inflow amount and the actual value for the flood case 1 based on the prediction model of FIG.
FIG. 10B similarly shows the predicted inflow amount and the actual value for the flood case 2. Similarly, FIG.
(B) shows the predicted inflow amount and the actual value for the flood case 1 and the flood case 2 based on the prediction model of FIG. 9. A specific method for predicting the inflow amount is a change in the inflow amount one hour later (prediction value) obtained by inputting the difference of the average basin rainfall into the prediction model.
Is added to the current inflow amount to predict the inflow amount one hour later, and this is repeated along the time axis direction in FIGS.

Tables 4 and 5 show the results of evaluating the prediction errors for the study examples 1 and 2. Table 4 shows flood cases 1 and 5
Is for flood case 2. The absolute value average error [%] and the correlation coefficient (correlation coefficient between the predicted value and the actual value) were used as the evaluation index of the prediction error. Here, it is better that the absolute value average error is smaller and the correlation coefficient is closer to 1.0.

[0049]

[Table 4]

[0050]

[Table 5]

As is apparent from Tables 4 and 5, Case 2 in which the difference data corresponding to the accumulated time in the invention of Claim 5 is used as the basin average rainfall difference is the same in both Tables 4 and 5. Absolute average error is smaller than Study Example 1,
The correlation coefficient is close to 1.0. That is, it was confirmed that selecting the cumulative time as the rainfall input time yielded better results.

[0052]

As described above, first, claims 1 and 2 are described.
The described invention can be used as an aid to a high-precision prediction model, and it is possible to grasp and predict the increase / decrease tendency of the flow rate by a simple method of displaying the accumulated rainfall together with the actual flow rate value. Since only a small amount of information, such as accumulated rainfall and flow rate, is enough to grasp the increase / decrease trend of the flow rate, it is easy for operators to understand and even when operating the dam gate based on the predicted value, the basis is clear. Can be

The third, fourth, and fifth aspects of the present invention are techniques that can be used when constructing a prediction model. In constructing a prediction model, it is necessary to understand the characteristics of the basin, and the rainfall time in the upstream basin is an important factor. According to the invention of claim 3, the falling time can be calculated with high accuracy, and even in the case where the peak time of the rainfall amount and the flow rate is unclear as in the related art, the calculation of the falling time is easily performed. Calculation is possible. Further, since automation is easy by numerical processing of a correlation coefficient or the like, it is particularly effective for a river having a characteristic in which a flowing time changes in a short time.

The fourth aspect of the present invention relates to a method for calculating a basin average rainfall. When multiple rivers are converging, a method of predicting the discharge for each river is known,
The accuracy of individual river forecasts is not high due to the dispersion of individual forecast values, and there is a limit to improving the accuracy of target river discharge forecast values. On the other hand, in the method of predicting a plurality of rivers as a whole, the variation in data is reduced, but the accuracy is not high when there is a difference in characteristics between the rivers. According to the present invention, a basin average rainfall can be calculated in consideration of a flowing time of each river, and highly accurate prediction can be performed by using the basin average rainfall.

According to a fifth aspect of the present invention, the cumulative time for calculating the cumulative rainfall is calculated. Claims 1 and 2 described above.
In the present invention, the accumulated rainfall having a high correlation with the flow rate is required as data. According to the fifth aspect of the present invention, since a highly correlated accumulated rainfall is obtained, for example,
If the invention is applied to the invention, relatively high-precision prediction can be performed.
Also, in other prediction models, rainfall for several hours is often used as input information. However, according to the invention of claim 5, the input range of the optimal rainfall (how many hours of rainfall To use for prediction) can be determined. As described in detail above, the inventions of claims 1 to 5 are versatile and easily automated methods applicable to rivers of any shape.

[Brief description of the drawings]

FIG. 1 is a diagram showing a display example of a screen according to an embodiment of the present invention.

FIG. 2 is a diagram showing a display example of a screen according to the embodiment of the first invention.

FIG. 3 is a diagram showing a display example of a screen according to the embodiment of the second invention;

FIG. 4 is an explanatory view of the principle of the invention of claim 3;

FIG. 5 is a diagram showing a relationship between a flow time and a correlation coefficient in the embodiment of the third invention.

FIG. 6 is a diagram showing a change over time of a cumulative rainfall and a flow rate for calculating a cumulative time in the embodiment of the invention of claim 5;

FIG. 7 is a diagram in which an average rainfall obtained by simply averaging the measured values of a plurality of rain gauges and a dam inflow actual value are superimposed.

FIG. 8 is a conceptual diagram of a prediction model to which the embodiment of claim 5 is applied.

FIG. 9 is a conceptual diagram of a prediction model to which the embodiment of claim 5 is applied.

FIG. 10 is a diagram showing prediction results for each flooding case.

FIG. 11 is a diagram showing prediction results for each flood case.

FIG. 12 is an explanatory diagram of a conventional dam flow prediction method.

FIG. 13 is an explanatory diagram of a conventional dam flow prediction method.

Claims (5)

[Claims] 1. A flow rate prediction method for predicting a flow rate of a dam or a river from rainfall in an upstream area using a prediction model constructed by a computer, wherein a flow rate actual value of a past predetermined period based on a present time is displayed. Along with displaying the graph on the top, at least the accumulated rainfall in the upstream area of the predetermined period is displayed so as to be superimposed on the actual flow rate value, and an increase or decrease tendency of the flow rate is predicted from the correlation between the accumulated rainfall rate and the actual flow rate value. Method for predicting discharges in dams or rivers. 2. A flow rate prediction method for predicting a dam or river flow rate from a rainfall rate in an upstream area using a prediction model constructed by a computer, the method comprising the steps of: A flow rate prediction method for a dam or river characterized by calculating a change Δc in the flow rate, and calculating a flow rate predicted value d by d = current flow rate + A · Δc + B (where A and B are constants). 3. A flow rate prediction method for predicting a flow rate of a dam or a river using a prediction model constructed by a computer and considering a flow time, which is a time required for the rainfall amount in the upstream area to be reflected in the flow rate. Calculate the correlation coefficient of both patterns while moving the temporal change pattern of the rainfall in the basin along the time axis in the direction of the time change pattern of the actual flow rate after that, and the rainfall when the correlation coefficient becomes the maximum Wherein the time corresponding to the movement amount of the time change pattern is defined as the flow-down time. 4. A flow rate prediction method for predicting a flow rate of a dam or a river using a prediction model constructed by a computer and considering a flow-down time which is a time until rainfall in a plurality of upstream areas is reflected in the flow rate. According to the invention of claim 3, a basin average rainfall is calculated for each basin, and a basin average rainfall is calculated using an average rainfall of each basin and an area of each basin in consideration of the downflow time. A method for predicting discharge in a dam or a river, comprising predicting discharge using a method. 5. A flow rate prediction method for predicting a flow rate of a dam or a river from a cumulative rain rate for a predetermined period in an upstream area using a prediction model constructed by a computer, wherein a time change pattern of the cumulative rain rate for a plurality of types of cumulative times. Are calculated along the time axis in the direction of the time change pattern of the actual flow rate value, and calculate the correlation coefficients of both patterns, and calculate the cumulative time corresponding to the accumulated rainfall when the correlation coefficient is the maximum. A flow prediction method for dams or rivers, characterized by searching and using the accumulated rainfall during the accumulation time for discharge prediction.