JP2023009541A - Rainwater inflow amount prediction device, rainwater inflow amount prediction method, and computer program - Google Patents

Abstract

To provide a rainwater inflow amount prediction device that accurately predicts a rainwater inflow amount.SOLUTION: A rainwater inflow amount prediction device according to one embodiment comprises: calculating main line flow velocity and a main line flow amount at multiple points; calculating a flow time from the multiple points to a pump well 14 from the main line flow velocity at the multiple points and main line extension from the multiple points to the pump well 14; calculating a rainwater inflow amount flowing into the pump well 14 using a water level measured by one of multiple water level gauges and an inflow culvert water gauge installed in the pump well; recording the main line inflow amount and the flow time corresponding to the main line inflow amount in association with each other to calculate an actual flow time corresponding to the main line inflow amount from the rainwater inflow amount and the main line inflow amount; compensating the flow time using a measured flow time corresponding to the main line flow rate; and calculating a predicted value of the rainwater inflow amount using the main line inflow amount, a corrected flow time, and a predicted time required for controlling a rainwater pump installed in the pump well 14.SELECTED DRAWING: Figure 2

Description

TECHNICAL FIELD Embodiments of the present invention relate to a rainwater inflow prediction device, a rainwater inflow prediction method, and a computer program.

In recent years, torrential rains and localized heavy rains (so-called guerrilla rainstorms) have increased, and measures to avoid flood damage are required. In response to this situation, the Ministry of Land, Infrastructure, Transport and Tourism is promoting inundation countermeasure projects through soft countermeasures such as utilization of observation data such as rainfall information, in addition to hard countermeasures such as the development of inundation countermeasure facilities. Against this background, rainwater pump dynamic control technology based on forecasting rainwater inflow to rainwater pumping stations has been developed as a non-structural measure, mainly aimed at reducing the risk of inundation. The rainwater pump dynamic control technology predicts the amount of rainwater flowing into the rainwater pumping station, and changes the start water level and stop water level of the rainwater drainage pump according to the estimated rainwater inflow amount. By changing the starting water level and stopping water level, the rainwater drainage pump can be started and stopped efficiently, aiming to reduce the risk of flooding and the number of times the rainwater drainage pump is started and stopped.

For example, if there is a sudden inflow of rainwater into a rainwater pump station, the rainwater drainage pump will not start until the water level of the rainwater pump well reaches the starting water level, which may increase the risk of flooding. In addition, for example, when the water level of the rainwater pump well has decreased but the rainwater drainage pump in operation does not need to be stopped due to the inflow of new rainwater into the rainwater pump well, the water level of the rainwater pump well falls below the stop water level. There was a possibility that the rainwater drainage pump would stop.

In order to fully exhibit the effect of the rainwater pump dynamic control, it is desirable to calculate the rainwater inflow prediction value with high accuracy.

Japanese Patent Application Laid-Open No. 2021-95711 (Patent Application 2019-226417)

Conventionally, from the main water level at each point, based on Manning's formula, the main flow speed, main flow rate, and the flow time from each point to the pump station are calculated, and the flow time is the same as the predicted time required for dynamic control of the rainwater pump. The flow rate of the main line at the point where the rainwater inflow is predicted is used as the predicted value. However, since the flow time is calculated based on the design information of the main line and pump station, there was a possibility that it differed from the actual measurement value. In addition, since the run-down time changes according to the flow rate in the pipe, it was difficult to accurately calculate the rainwater inflow amount prediction value based on the design information.

The embodiments of the present invention have been made in view of the above circumstances, and it is an object of the present invention to provide a rainwater inflow prediction device, a rainwater inflow prediction method, and a computer program for accurately predicting rainwater inflow. do.

The rainwater inflow prediction device according to the embodiment uses the water levels measured by each of a plurality of water level gauges installed at a plurality of points on the trunk line to calculate the trunk flow velocity and trunk flow rate at the points. a flow time calculation unit configured to calculate a flow time from the plurality of points to the pump well based on the trunk flow velocity at the plurality of points and the extension of the main line from the plurality of points to the pump well; a rainwater inflow calculation unit for calculating the amount of rainwater flowing into the pump well using the water level measured by either the water level gauge of or the inflow culvert water level gauge installed in the pump well; and the flow time corresponding to the main flow rate are recorded in association with each other, and the measured flow time corresponding to the main flow rate is calculated from the rainwater inflow amount and the main flow rate, and the measured flow time corresponding to the main flow rate is calculated. Using a flow time correction unit that corrects the flow time using the flow time, the main flow rate, the corrected flow time, and the predicted time required for controlling the rainwater pump installed in the pump well , and a rainwater inflow predicted value determining unit for calculating a predicted value of the rainwater inflow.

Drawing 1 is a figure showing roughly composition of a rainwater pumping station system which installed a rainwater inflow prediction device of one embodiment. FIG. 2 is a block diagram schematically showing one configuration example of the rainwater inflow prediction device of the first embodiment. FIG. 3 is a diagram for explaining variables used in the Manning formula. FIG. 4 : is a figure which shows roughly an example of the table with which the flow-down time correction|amendment part of the rainwater inflow prediction apparatus of one Embodiment is provided. FIG. 5 is a diagram schematically showing an example of time-series data included in the flow-down time correction unit of the rainwater inflow prediction device of one embodiment. FIG. 6 is a diagram for explaining an example of the operation of the rainwater inflow prediction device of the embodiment. FIG. 7 is a diagram for explaining an example of correction of the flow time by the flow time correction unit of the rainwater inflow prediction device of one embodiment. FIG. 8 is a block diagram schematically showing one configuration example of the rainwater inflow prediction device of the second embodiment. FIG. 9 is a diagram for explaining an example of the time-flow distribution data set generated by the time-flow distribution generator of the rainwater inflow prediction apparatus of the second embodiment. FIG. 10 is a diagram for explaining an example of the operation of the rainwater inflow prediction device of the second embodiment.

Hereinafter, a rainwater inflow prediction device, a rainwater inflow prediction method, and a computer program according to embodiments will be described with reference to the drawings.
Drawing 1 is a figure showing roughly composition of a rainwater pumping station system which installed a rainwater inflow prediction device of one embodiment.
The rainwater pumping station system includes water level gauges G0 to GN provided in a main line (for example, a sewage pipe) 10, an inflow culvert 12, a pump well 14, and a rainwater inflow prediction device 100.

The main line 10 is a sewer culvert leading to an inflow culvert 12 and a pump well 14 . Sewage containing rainwater flows through the main line 10 into the pump well 14 as it rains.
A plurality of water level gauges G0 to GN are installed along the downstream direction of the main line 10 to obtain the water level of rainwater on the main line 10 . In this embodiment, "rainwater" is also used in the sense of sewage containing rainwater.
The inflow culvert 12 is a drainage pipe leading from the main line 10 to the pump well 14 , and rainwater from the main line 10 passes through the inflow culvert 12 and flows into the pump well 14 .

The pump well 14 is installed at the most downstream side of the rainwater pumping station, and stores rainwater flowing in from the inflow culvert 12 . In the pump well 14, rainwater is pumped up by a plurality of rainwater drainage pumps (not shown) and discharged into a river.

A plurality of stormwater drainage pumps are controlled to start and stop by a stormwater pump control system (not shown). The plurality of rainwater drainage pumps are, for example, a pump that is activated first when rainwater is discharged from the pump well 14 to the river, and a pump that is selectively activated according to the flow rate of inflow (or expected inflow). including.

The rainwater inflow predictor 100 may be included in a rainwater pump control system. The rainwater inflow predicting device 100 acquires water level signals from the water level gauges G0 to GN, and based on the acquired signals, calculates and outputs a predicted value of the amount of inflowing rainwater. In addition, the rainwater inflow prediction device 100 acquires an operation signal based on the operation of an operation means such as a keyboard, a mouse, a touch panel, etc. by an operator or administrator of the rainwater pump control system, and performs processing according to the operation signal. It is also possible to

The main line water level gauges G1 to Gn are installed at a plurality of points on the main line 10, and measure the main line water level at each of the installed points.
The inflow culvert water level gauge G0 is installed in the inflow culvert 12 and measures the water level at the installed point. In addition, when the water level gauge G0 is not installed in the inflow culvert 12, the trunk water level gauge closest to the inflow culvert 12 may be used instead of the water level gauge G0. In this case, the flow-down time from the nearest main water level gauge to the pump well 14 uses, for example, the flow-down time (initial value) described later. Furthermore, the inflow culvert water level gauge G0 may be a sensor for calculating the amount of rainwater flowing into the pump well 14, and may have a configuration other than the above.

FIG. 2 is a block diagram schematically showing one configuration example of the rainwater inflow prediction device of the first embodiment.
The rainwater inflow prediction device 100 of this embodiment includes, for example, at least one processor and a memory in which a program executed by the processor is recorded, and is configured to implement various functions by software. Alternatively, it may be configured to implement various functions by a combination of software and hardware. In addition, the program executed by the rainwater inflow prediction device 100 may be recorded in a computer-readable recording medium and provided.

At least some functions or all functions of the rainwater inflow prediction device 100 may be provided in facilities such as the main line 10, the inflow culvert 12, and the pump well 14, or may be provided remotely. When the rainwater inflow prediction device 100 is installed remotely such as the pump well 14, the rainwater inflow prediction device 100 is based on, for example, information obtained from a sensor via a network or information input by an operator or administrator. may be used to calculate the rainwater inflow prediction value. In this case, the rainwater inflow prediction value calculated by the rainwater inflow prediction device 100 is supplied to the rainwater pump control system via a network, for example, and used for pump control.

The rainwater inflow prediction device 100 of this embodiment includes a main flow velocity main flow rate calculation unit 101, a flow time calculation unit 102, a rainwater inflow calculation unit 103, a flow time correction unit 104, and a rainwater inflow prediction value determination unit. 105 and.

The trunk flow velocity trunk flow rate calculation unit 101 receives as input the trunk water levels h1 to hn measured by the trunk water level gauges G1 to Gn, and from the Manning formula of formula (1) below, the points where the trunk line water level gauges G1 to Gn are installed are calculated. Calculate the trunk flow velocities V1 to Vn.
In the following formula (1), V is the flow velocity (m/sec), n is the roughness coefficient (sec/m ^1/3 ), R is the radial depth (m), I is the energy gradient (-) ≈ physical gradient (- ).
V=1/n×R2/ ³ × ^I1/2 (1)

FIG. 3 is a diagram for explaining variables used in the Manning formula.
A is flow volume (m ² ) and S is wetted area (m). Moreover, h is the water level measured with the water level gauge.
As shown in FIG. 3, the shape of the main line 10 makes it possible to derive the flow volume A (m ² ) and wet side S (m) based on the water levels h1 to hn.
Equation (2) shows the relationship between the diameter depth, the flow volume, and the wetted side.
R=A/S (2)
From the formulas (1) and (2) shown above, the main flow velocity main flow rate calculation unit 101 calculates the rainwater flow velocities V1 to Vn at the points where the main water level gauges G1 to GN are installed from the water levels h1 to hn.

A trunk flow velocity trunk flow rate calculation unit 101 is means for outputting trunk flow rates Q1 to Qn at points where the trunk water level gauges G1 to Gn are installed, based on the rainwater flow velocities V1 to Vn and the acquired water levels h1 to hn. is. From the flow velocities V1 to Vn at the points where the main water level gauges G1 to Gn are installed calculated by the formula (1) and the flow volumes A1 to An at the points where the main water level gauges G1 to Gn are installed, the formula ( 3) is used to calculate the trunk flow rates Q1 to Qn.
Q=V×A (3)

The flow-down time calculation unit 102 receives the main line flow velocities V1 to Vn calculated by the main line flow velocity/main flow rate calculation unit 101 and the main line extensions L1 to Ln. Flow times t1 to tn up to are calculated by the following formula (4). The flow time calculation unit 102 outputs the calculated flow times t1 to tn to the flow time correction unit 104. FIG.
t=L/V (4)
t: flow time (sec) L: length of main line (m)

The rainwater inflow calculation unit 103 receives the water level h0 measured by the inflow culvert water level gauge G0, and calculates the rainwater inflow Q0 flowing into the pump well 14 in the same manner as the main line velocity/main flow rate calculation unit 101 does. Further, when the rainwater inflow of the inflow culvert 12 greatly changes due to the pump operation, the rainwater inflow calculation unit 103 may calculate the rainwater inflow Q0 by also using the log of the pump operation. If the inflow culvert water level gauge G0 is not installed, the main line flow rate at the point closest to the pump well 14 may be substituted to calculate the rainwater inflow Q0. Further, when information on the discharge amount of the pump installed in the pump well 14 can be acquired, the rainwater inflow amount calculation unit 103 calculates the substantial inflow amount together with the operation log of the pump. good.
The rainwater inflow Q0 may be input to the rainwater inflow prediction device 100 from the outside, and in this case, the rainwater inflow calculation unit 103 of the rainwater inflow prediction device 100 can be omitted.

The flow time correction unit 104 calculates the flow time t1 to tn calculated by the flow time calculation unit 102, the main flow rate Q1 to Qn calculated by the main flow velocity main flow calculation unit 101, and the rainwater inflow calculated by the rainwater inflow calculation unit 103. Quantities Q0 are input. The flow time correction unit 104 calculates the actual flow times (measured flow times) tr1 to trn from each point to the pump well 14 based on the temporal changes in the main line flow rates Q1 to Qn and the rainwater inflow amount Q0 at each point. , corrects the flow times t1 to tn using the actually measured flow times tr1 to trn, and outputs the corrected flow times t1 to tn to the rainwater inflow prediction value determination unit 105 .

The rainwater inflow predicted value determination unit 105 compares the corrected flow times t1 to tn calculated by the flow time correction unit 104 with the predicted time required for rainwater pump control, and determines a flow time equivalent to the predicted time. The main flow rate Qk at the point where the main water level gauge k (1≤k≤n) is installed is adopted as the predicted value of the rainwater inflow into the pump well 14, and the predicted rainwater inflow is determined.

The predicted time required for rainwater pump control is, for example, the start-up time of the rainwater drainage pump (not shown) of the pump well 14 . The operator or administrator can input the predicted time (startup time) to the rainwater inflow prediction device 100 in advance from the operation unit. The predicted time input to the rainwater inflow prediction device 100 can be recorded, for example, in a storage unit (not shown), and the rainwater inflow prediction value determination unit 105 acquires information including the activation time input to the storage unit. In this embodiment, it is assumed that a plurality of rainwater drainage pumps are installed in the pump well 14 and, for example, the start-up times of the plurality of rainwater drainage pumps are all the same. The "estimated time" in this embodiment is, for example, the time from when the rainwater drainage pump starts operating until it reaches a steady operating state. Also, the time obtained by adding the control period to the startup time of the rainwater drainage pump may be used as the predicted time.

The predicted rainwater inflow amount determination unit 105 predicts the rainwater inflow amount by machine learning or the like from a data set of the main line flow rate Q1 to Qn and the measured flow time tr1 to trn at the points where the main water level gauges G1 to Gn are installed. A model may be created and a value obtained using a rainwater inflow prediction model may be used as the rainwater inflow prediction value.

Next, an example of the operation of the rainwater inflow prediction device 100 of this embodiment will be described.
The trunk water level gauges G1 to Gn and the inflow culvert water level gauge G0 periodically measure the trunk water levels h1 to hn and the inflow culvert water level h0.
Subsequently, the trunk flow velocity trunk flow rate calculation unit 101 obtains the measured values of the trunk water levels h1 to hn, and calculates the trunk flow velocities V1 to Vn and the trunk flow rates Q1 to Qn using, for example, the Manning formula described above.

The rainwater inflow calculation unit 103 acquires the measured value of the inflow culvert water level h0, and calculates the rainwater inflow Q0 using, for example, the Manning equation described above.
The flow time calculation unit 102 acquires the trunk line flow rates Q1 to Qn and the trunk line extensions L1 to Ln input from the outside, and calculates the flow times (initial values) t1 to tn.

The flow time correction unit 104 has a table storing the flow times t1 to tn with respect to the main flow rate for each point where the main water level gauges G1 to Gn are installed.
FIG. 4 : is a figure which shows roughly an example of the table with which the flow-down time correction|amendment part of the rainwater inflow prediction apparatus of one Embodiment is provided.

The flow time correction unit 104 associates the data of the main flow rate Q1 to Qn with the flow time (initial value) t1 to tn for each point where the main water level gauges G1 to Gn are installed, and creates a table (not shown). to record. At this time, the flow time correction unit 104 may discretize the data recorded in the table, or may record the data as continuous values. In addition, the flow time correction unit 104 identifies a function representing the relationship between the main flow rate Q1 to Qn and the flow time t1 to tn for each point where the main water level gauges G1 to Gn are installed, and replaces the table with the function. may be recorded. The flow time correction unit 104 can record a table or function in a storage unit (not shown).

FIG. 5 is a diagram schematically showing an example of time-series data included in the flow-down time correction unit of the rainwater inflow prediction device of one embodiment.
The flow time correction unit 104 records as time-series data that associates the time, the main flow rate at each point where the main water level gauges G1 to Gn are installed, and the rainwater inflow calculated by the rainwater inflow calculation unit 103. do. The flow time correction unit 104 can set the time series data sampling interval to any time interval.

The flow time correction unit 104 refers to the time series data, and based on the time difference between the waveforms of the main water flow Q1 to Qn at each point where the main water level gauges G1 to Gn are installed and the waveform of the rainwater inflow Q0. Flow times tr1 to trn are calculated. Since the measured flow times tr1 to trn depend on the flow rate, the measured flow times tr1 to trn cannot be obtained simply from the time difference between the waveforms. Therefore, the flow time correction unit 104, for example, extracts the maximum value (maximum value) of the waveforms of the flow rates Q1 to Qn at each point, extracts the maximum value (maximum value) of the waveform of the rainwater inflow Q0, and extracts the flow rate Q1 The difference between the time at which ~Qn reaches its maximum value and the time at which the rainwater inflow amount Q0 reaches its maximum value is calculated as actually measured run-down times tr1 to trn.

FIG. 6 is a diagram for explaining an example of the operation of the rainwater inflow prediction device of the embodiment.
The flow-down time correction unit 104, for example, determines the maximum value (local maximum value) in each of the trunk flow rates Q1 to Qn at the points where the trunk water level gauges G1 to Gn are installed and the rainwater inflow Q0. A quantity threshold may be set. This is because the accuracy deteriorates when the flow time correction unit 104 catches minute fluctuations in the time-series data. By setting the threshold value of the amount of change in the time-series data in consideration of the average variation at each point, the flow time correction unit 104 can accurately calculate the maximum value (local maximum value). For example, the flow time correction unit 104 records the amount of change in the main line flow rate Q1 to Qn per unit time when the rainfall intensity is 1 mm/h observed by XRAIN (High Performance Radar Rain Gauge Network) or the like, By using this amount of change as a threshold, it can be used as the sensitivity of the flowmeter for measuring the trunk flow rates Q1 to Qn and the rainwater inflow rate Q0.

In addition, the flow time correction unit 104 corrects the waveforms of the main line flow Q1 to Qn and the rainwater inflow Q0 at each point, instead of the maximum value (local maximum) of the time-series data, using a characteristic waveform (flow time is known). Measured flow times tr1 to trn may be set as the flow times corresponding to characteristic waveforms that match each other in comparison with a certain waveform). The flow time correction unit 104 calculates the distance between waveforms by, for example, dynamic time warping (DTW), and can determine whether or not the waveforms match by comparing the calculated distance with a threshold value. The flow time corresponding to the characteristic waveform is recorded in advance in the flow time correction unit 104, and when it is determined that the waveforms match, the flow time corresponding to the characteristic waveform may be set as the measured flow time tr1 to trn. good.

When the measured flow times tr1 to trn are calculated, the flow time correction unit 104 associates the flow rates Q1 to Qn at each point with the measured flow times tr1 to trn, and calculates the flow times recorded in the table (initial value ) t1 to tn (or already corrected updated flow times t1 to tn) are corrected with the actually measured flow times tr1 to trn.

FIG. 7 is a diagram for explaining an example of correction of the flow time by the flow time correction unit of the rainwater inflow prediction device of one embodiment.
For example, when any one of the measured flow times tr1 to trn can be calculated, the flow time correction unit 104 corrects the corresponding values of the flow times t1 to tn recorded in the table and updates the table.

In the example shown in FIG. 7, when the trunk flow rate is 5 (m ³ /sec) at the trunk water level gauge G1, the measured flow time is 4 (sec), and at the trunk water level gauge Gn, the trunk flow rate is 3 (m ³ /sec). ) is calculated to be 2 (sec).

When correcting the values recorded in the table, the flow time correction unit 104 calculates the average value of the initial value of the flow time to be corrected and the measured flow time, and calculates the average value as the flow time at the corresponding trunk flow rate. can be corrected as In this case, when the trunk flow rate is 5 (m ³ /sec) at the trunk water level gauge G1, the flow time after correction is 5 (sec), and at the trunk water level gauge Gn, the trunk flow rate is 3 (m ³ /sec). The flow time after the correction is 3 (sec).

Note that the method of calculating the correction value of the flow time is not limited to the above. Further, in order to improve robustness, the flow times t1 to tn may be corrected using a Kalman filter.

In addition, the flow time correction unit 104 creates a regression equation that uses weather conditions such as temperature at the relevant time, rainfall data observed by XRAIN, etc., event information of the relevant area, etc. as variables, and calculates the correction value. can be calculated. Alternatively, the flow time correction unit 104 may present the correction value to the user on a display unit (not shown), ask the user for approval of the correction, and adopt the correction value after the user's approval.

The flow time corrector 104 outputs the corrected flow times t1 to tn for the trunk flow rates Q1 to Qn to the rainwater inflow prediction value determiner 105. FIG.
The rainwater inflow predicted value determination unit 105 compares the corrected flow times t1 to tn calculated by the flow time correction unit 104 with the predicted time required for rainwater pump control, and determines a flow time equivalent to the predicted time. The main flow rate at the point where the main water level gauge is installed is adopted as the predicted value of the rainwater inflow to the rainwater pumping station, and the predicted value of the rainwater inflow is determined.

Note that if there is no point where the main water level gauge is installed and the flow time t1 to tn is equal to the predicted time, the rainwater inflow prediction value determination unit 105 determines the main flow rate of the point where the main water level gauge is installed. A rainwater inflow prediction model may be created by machine learning or the like from a data set of Q1 to Qn and corrected flow times t1 to tn, and used as rainwater inflow prediction values.

As described above, according to the present embodiment, the main line flow rates Q1 to Qn measured at the points where the main line water level gauges G1 to Gn are installed, and the flow times t1 to tn at each point calculated based on the design information is corrected using the actually measured flow times tr1 to trn, the accuracy of the flow times t1 to tn recorded in the table can be improved. Furthermore, by improving the accuracy of the flow times t1 to tn recorded in the table, it is possible to improve the prediction accuracy of the rainwater inflow amount. Also, although the actually measured flow times tr1 to trn contain noise and the like, they are not robust values, but by applying them as a correction process, robust corrected flow times t12 to tn can be obtained. That is, when correcting the flow times (initial values) t1 to tn, it is possible to improve the robustness of the correction value by adopting a format such as a median value, an average value, a regression equation, or user approval. This makes it possible to improve the prediction accuracy of rainwater inflow.

That is, according to the present embodiment, it is possible to provide a rainwater inflow prediction device, a rainwater inflow prediction method, and a computer program for accurately predicting rainwater inflow.

Next, the rainwater inflow prediction device, rainwater inflow prediction method, and computer program of the second embodiment will be described in detail with reference to the drawings.
In addition, in the following description, the same reference numerals are given to the same configurations as those of the above-described first embodiment, and the description thereof will be omitted.

FIG. 8 is a block diagram schematically showing one configuration example of the rainwater inflow prediction device of the second embodiment.
The rainwater inflow prediction device 100 of this embodiment is different from the above-described first embodiment in that it includes a flow time distribution generation unit 106 and corrects the flow time using the flow time distribution and the flow time correction policy. ing.

The flow time distribution generation unit 106 receives as inputs the main line flow rates Q1 to Qn calculated by the main line velocity main line flow rate calculation unit 101 and the rainwater inflow amount Q0 calculated by the rainwater inflow amount calculation unit 103. Actual flow times (measured flow times tr1 to trn) at each point are calculated from changes over time in the main line flow rates Q1 to Qn and rainwater inflow Q0. Note that the method of calculating the actually measured run-down times tr1 to trn is the same as that of the above-described first embodiment.

Furthermore, the flow time distribution generation unit 106 uses the accumulation period information input from the outside to determine the flow times of the trunk line flow rates Q1 to Qn and the measured flow times tr1 to trn at each point during a certain period (accumulation period). Record the distribution dataset. The flow-down time distribution data set can be generated for each point where the main water level gauges G1 to Gn are installed, and recorded as a table, for example, in a storage unit (not shown).

The accumulation period information may be information specifying a period, or may be information specifying an accumulation period in real time with a period start signal and a period end signal. For example, by designating a period in which a heavy rain forecast is issued based on the accumulation period information, it is possible to accumulate actually measured run-down times tr1 to trn under specific conditions.

FIG. 9 is a diagram for explaining an example of the time-flow distribution data set generated by the time-flow distribution generator of the rainwater inflow prediction apparatus of the second embodiment.
Here, the main line flow rates Q1 to Qn and the measured flow times tr1 to trn of the flow time distribution data set are represented by histograms.

The flow time distribution generation unit 106 generates the distribution (flow time distribution) of each main line flow rate Q1 to Qn and the measured flow time tr1 to trn at each point where the main water level gauges G1 to Gn are installed, to the flow time correction unit 104. output to The flow time distribution generation unit 106 may output all individual data of the flow time distribution to the flow time correction unit 104, or may identify the distribution and output distribution parameters such as mean and variance to the flow time correction unit 104. You may The flow time distribution generation unit 106 may generate a flow time distribution data set excluding outliers, for example, using an externally given threshold value.

The flow time correction unit 104 calculates the flow times (initial values) t1 to tn calculated by the flow time calculation unit 102 and the main water level gauges G1 to Gn calculated by the flow time distribution generation unit 106 at each point where the water level gauges G1 to Gn are installed. At each point where the main water level gauges G1 to Gn are installed, according to the flow time correction policy input from the outside, using the flow time distribution data of the main flow rate Q1 to Qn and the measured flow time tr1 to trn is corrected, and output to the rainwater inflow prediction value determination unit 105 as corrected flow times t1 to tn.

In addition, the flow time correction unit 104 stores (not shown) distribution data of the main flow rates Q1 to Qn and the measured flow times tr1 to trn, and distribution data of the main flow rates Q1 to Qn and the corrected flow times t1 to tn. may be recorded in the section. The flow time correction unit 104 uses the distribution data recorded in the storage unit to adjust the characteristics (emergency, safe, normal, etc.) of the flow time t1 to tn after correction according to the situation such as heavy rain forecast. can be set.

In the above-described first embodiment, the flow time correction unit 104 corrects the flow time initial values t1 to tn and the average value of the measured flow times tr1 to trn each time the measured flow times tr1 to trn are input. The downstream times t1 to tn are calculated and the downstream times (initial values) t1 to tn are updated. In the present embodiment, the actually measured downstream times tr1 to trn input to the downstream time correction unit 104 are distribution data.

FIG. 10 is a diagram for explaining an example of the operation of the rainwater inflow prediction device of the second embodiment. Here, the flow time correction unit 104 calculates the flow time (initial value ) An example of the operation for correcting t1 to tn will be described.

The flow time correction unit 104 sets the flow times (initial values) t1 to tn calculated by the flow time calculation unit 102 to, for example, the average value of a normal distribution, and sets the variance to an arbitrary value (for example, 1) to determine the flow time (initial value) value) Distribution data of t1 to tn. The flow time correction unit 104 updates (corrects) the initial value distribution data based on the distribution data of the measured flow times tr1 to trn input from the flow time distribution generation unit 106 . The flow time correction unit 104 can use, for example, a Bayesian estimation method as a method of updating the distribution data.

Subsequently, the flow time correction unit 104 calculates post-correction flow times t1 to tn from the updated distribution data in accordance with the corrected flow time correction policy. The correction policy can be determined using statistical parameters such as the average value and variance of the distribution corresponding to the water level state of the pump well 14. For example, if normal, the average value is calculated as the post-correction run-down time t1 to tn. However, if the water level of the pump well is in a dangerous state (the water level exceeds a predetermined threshold value, etc.), correction such as calculating a value that is 1σ smaller than the average value as the flow time t1 to tn after correction is performed. Contains policy information. By inputting a correction policy from the outside, it is possible to specify a complicated policy that takes into consideration all kinds of conditions such as weather conditions and events.

The rainwater inflow prediction device of this embodiment is the same as that of the first embodiment described above except for the configuration described above. According to the rainwater inflow prediction device of the present embodiment, by calculating the post-correction flow times t1 to tn using the distribution data of the actually measured flow times tr1 to trn, it is possible to further improve the robustness of the correction values. This makes it possible to improve the prediction accuracy of rainwater inflow.
That is, according to the present embodiment, it is possible to provide a rainwater inflow prediction device, a rainwater inflow prediction method, and a computer program for accurately predicting rainwater inflow.

While several embodiments of the invention have been described, these embodiments have been presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and modifications can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the scope of the invention described in the claims and equivalents thereof.

DESCRIPTION OF SYMBOLS 10... Main line (sewage pipe), 12... Inflow culvert, 14... Pump well, 100... Rainwater inflow prediction apparatus, 101... Main line velocity/main line flow rate calculation part, 102... Flow time calculation part, 103... Rainwater inflow calculation part, 104... Flow time correction unit, 105... Rainwater inflow prediction value determination unit, 106... Flow time distribution generation unit, A1 to An... Flow volume, G0... Inflow culvert water level gauge, G1 to GN... Trunk water level gauge, L1 to Ln Main line length Q0 Rainwater inflow Q1 to Qn Main line flow rate V1 to Vn Main line flow velocity h0 Inlet water level h1 to hn Main line water level t1 to tn Flow time (initial value) After correction flow time, tr1 to trn: measured flow time

Claims (8)

a trunk flow velocity and trunk flow rate calculation unit that calculates the trunk flow velocity and the trunk flow rate at a plurality of points on the trunk line using the water levels measured by each of a plurality of water level gauges installed at a plurality of points on the trunk line;
a flow-down time calculation unit that calculates the flow-down time from the plurality of points to the pump well based on the trunk flow velocity at the plurality of points and the extension of the main line from the plurality of points to the pump well;
a rainwater inflow amount calculation unit that calculates an inflow amount of rainwater flowing into the pump well using the water level measured by one of the plurality of water level gauges and an inflow culvert water level gauge installed in the pump well;
The main flow rate and the flow time corresponding to the main flow rate are recorded in association with each other, and the actually measured flow time corresponding to the main flow rate is calculated from the rainwater inflow amount and the main flow rate, and the flow time corresponding to the main flow rate is calculated. a flow-down time correction unit that corrects the flow-down time using the measured flow-down time;
Rainwater inflow predicted value determination for calculating the predicted value of the rainwater inflow using the trunk flow rate, the corrected flow time, and the predicted time required for controlling the rainwater pump installed in the pump well. A rainwater inflow prediction device comprising: 2. The rainwater inflow prediction device according to claim 1, wherein said flow time correction unit calculates a difference between a time when said rainwater inflow reaches a maximum value and a time when said trunk flow rate reaches a maximum value as said measured flow time. . 3. The rainwater inflow predicting device according to claim 2, wherein said flow time correction unit can set conditions for determining a maximum value of said rainwater inflow and said trunk flow. The flow time correction unit sets the flow time after correction to an average value of the flow time and the measured flow time, or a sum of weighted values of the flow time and the measured flow time. , The rainwater inflow predicting device according to claim 1. a flow time distribution generating unit that calculates the measured flow time from the rainwater inflow amount and the main flow rate, and generates distribution data for a predetermined accumulation period in which the main flow rate and the measured flow time are associated. ,
2. The rainwater inflow according to claim 1, wherein the flow time correcting unit corrects the distribution of the flow time using the distribution data, and calculates the corrected flow time using the corrected flow time distribution. Quantity predictor. The flow time correction unit calculates the corrected flow time from the distribution of the corrected flow time according to a correction policy input from the outside,
6. The rainwater inflow prediction apparatus according to claim 5, wherein said correction measure is determined using statistical parameters such as an average value and variance of a distribution corresponding to the water level state of said pump well. Using the water levels measured by each of a plurality of water level gauges installed at a plurality of points on the trunk line, calculate the trunk flow velocity and the trunk flow rate at the points,
calculating the flow time from the plurality of points to the pump well from the main line flow velocity at the plurality of points and the extension of the main line from the plurality of points to the pump well;
calculating the amount of rainwater flowing into the pump well using the water level measured by one of the plurality of water level gauges and the inflow culvert water level gauge installed in the pump well;
The main flow rate and the flow time corresponding to the main flow rate are recorded in association with each other, and the actually measured flow time corresponding to the main flow rate is calculated from the rainwater inflow amount and the main flow rate, and the flow time corresponding to the main flow rate is calculated. Correcting the flow time using the measured flow time,
A rainwater inflow prediction method, wherein the predicted value of the rainwater inflow is calculated using the main flow rate, the corrected flow time, and the predicted time required for controlling the rainwater pump installed in the pump well. . A computer program that causes a computer to execute the rainwater inflow prediction method according to claim 7 .

Images