JPH06102911A - Method and device for predicting inflow water rate in drainage pump place - Google Patents

Abstract

PURPOSE:To precisely predict a rainwater inflow rate to a drainage pump place and to eliminate the unnecessary start/stop of a pump. CONSTITUTION:A rainwater inflow rate prediction part 15 fetches rainfall rate data 5 in a drainage area 1, water level data 8 in a drainpipe groove 2 and water level data 10 in a pump well and calculates the rainwater inflow rate to the pump place 3. A strong rain time rainwater inflow rate prediction model obtained by modelling the rainwater inflow rate flowing to the pump place 3 at the time of strong rain and a regular rainfall time rainwater inflow rate prediction model obtained by modelling the rainwater inflow rate flowing to the pump place 3 at the time of regular rainfall are provided in the rainwater inflow rate prediction part 15. The models are switched at the time of strong rain and regular rainfall so as to execute prediction. The rainwater inflow rate is highly precisely calculated. Thus, the pump is appropriately operated by giving a rainwater prediction rate to a pump number control part 16 as feed forward information.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for predicting an inflow water amount in a drainage pump station, which is capable of accurately predicting an inflow amount of rainwater flowing into the drainage pump station at the time of rainfall and performing an appropriate pump operation.

[0002]

2. Description of the Related Art In order to cope with a sudden inflow of rainwater into a drainage pump station, as shown in, for example, JP-A-64-19402, the inflow amount of rainwater into a pump well is predicted and predicted. A method of controlling a pump based on an inflow amount is known. In this pump control method, the Kalman filter is applied to the time series data of the actual inflow into the pump well to predict the inflow of rainwater.

[0003]

However, in the above-mentioned conventional techniques, the rainwater inflow is not predicted in consideration of the rainfall situation. That is, the point to be noted when handling a rainwater treatment system is that the behavior of the entire system changes greatly depending on the state of rainfall. Therefore, during normal rainfall and during heavy rainfall, “rainfall → reach the ground → sewer pipe The process of "inflow → delivery → delivery to rainwater pump station" is different,
It is to treat them separately. However, this point is not taken into consideration in the conventional technology.

To calculate the actual inflow of rainwater into the pump well, the water level fluctuation data of the pump well is referred to in order to solve the balance equation at the pump well. Is very serious. Therefore, the calculation of the actual rainwater inflow amount is significantly affected by the pump operation (in particular, the operation of the rainwater pump), and the Kalman filter that reflects the pump operation pattern calculates the predicted value from the past actual rainwater inflow amount. Therefore, the error of the actual inflow amount greatly affects the predicted value.

As described above, the above-mentioned conventional technique controls the number of pumps based on the prediction of rainwater inflow into the pump well, but sufficient accuracy is ensured in the accuracy of rainwater inflow prediction into the pump well. However, if the number of pumps is controlled based on the rainwater inflow predicted value, there is a drawback that unnecessary start-up and stop of the pumps are repeated.

An object of the present invention is to provide a method and a device for predicting the inflow water amount in a drainage pump station, which can predict the rainwater inflow amount into the drainage pump station with high accuracy and eliminate unnecessary start and stop of the pump. is there.

[0007]

In order to achieve the above object, the present invention provides at least one of rainfall data in a drainage area, water level or flow rate data in a sewer, and water level data in a pump well. As an input value, in the inflow water amount prediction method that predicts the inflow water amount to the drainage pump station installed in the combined or split type sewer system, the heavy rain modeled the rainwater inflow amount that flows into the drainage pump station during heavy rain. The rainwater inflow prediction model and the rainwater inflow prediction model for normal rainfall that model the rainwater inflow that flows into the drainage pump station during normal rainfall are set in advance, and the rainwater inflow during heavy rain is described above. Based on the prediction model, based on the normal rainfall rainwater inflow prediction model during normal rainfall,
This is to predict the inflow water amount.

Further, the present invention is a combined or divided sewer system, wherein at least one of rainfall amount data in a drainage area, water level or flow rate data in a sewer pipe, and water level data in a pump well is used as an input value. In the method of predicting the amount of inflow water to the drainage pump station installed inside, the inflow prediction model for rainwater pump operation that models the inflow amount of rainwater flowing into the drainage pump station during operation of the rainwater pump and the rainwater pump non-operation A rainwater pump non-operating inflow forecasting model that models the rainwater inflow into the drainage pump station during operation, and a sewage pump operating inflow forecasting model that models the rainwater inflow into the drainage pump station during operation , A model for predicting the inflow of non-operating wastewater pumps that models the inflow of rainwater flowing into the drainage pump station when the wastewater pumps are not operating Based on the rainwater pump operating inflow prediction model and the rainwater pump non-operating inflow prediction model during heavy rain, during normal rainfall, the sewage pump operating inflow prediction model and sewage pump not operating This is to predict the inflow water amount based on an inflow amount prediction model.

Further, the present invention is directed to rainfall in drainage areas,
Measuring means for measuring the water level or flow rate in the sewer or the water level of the pump well, and inputting at least one of the measured rainfall data, water level or flow rate data in the sewer, and water level data of the pump well As a value, in the inflow water amount prediction device equipped with an inflow water amount prediction means for predicting the inflow water amount to the drainage pump station installed in the combined or split type sewer system, rainwater flowing into the drainage pump station during heavy rain Heavy rain rainwater inflow prediction model that models inflow, normal rainwater inflow prediction model that models rainwater inflow that flows into the drainage pump station during normal rainfall, and the aforementioned heavy rain rainwater during heavy rain A prediction execution means for predicting the inflow water quantity based on the inflow quantity prediction model based on the normal rainfall rainwater inflow quantity prediction model during normal rainfall. It is intended.

The present invention also provides a measuring means for measuring the amount of rainfall in the drainage area, the water level or flow rate in the sewer pipe, or the water level of the pump well, the measured rainfall data, and the water level in the sewer pipe. Alternatively, at least one of the flow rate data and the water level data of the pump well is used as an input value, and an inflow water amount prediction means for predicting the inflow water amount to the drainage pump station installed in the confluence type or split type sewer system is provided. In the inflow quantity prediction device, the rainwater pump inflow prediction model that models the rainwater inflow quantity that flows into the drainage pump station when the rainwater pump is operating and the rainwater inflow quantity that flows into the drainage pump station when the rainwater pump is not operating are modeled. A model for predicting inflow of rainwater pumps when not operating and a model for predicting inflow of rainwater that flows into the drainage pump station when the wastewater pumps are operating A model, a model for predicting the inflow of non-operating wastewater pumps that models the inflow of rainwater that flows into the drainage pump station when the wastewater pump is not operating, and a model for predicting the inflow of operating rainwater pumps and inflow when the rainwater pump is not operating during heavy rain A prediction execution means for predicting the inflow of water based on a volume prediction model, based on the sewage pump operating inflow prediction model during normal rainfall, and the sewage pump non-operating inflow prediction model Is.

[0011]

According to the present invention, by adopting the rainwater inflow model adapted to the rainfall condition, the randomness peculiar to the rainfall phenomenon and the non-linearity of each rainfall pattern are absorbed, and the rainwater inflow amount to the drainage pump station is absorbed. It can be predicted accurately.

Further, since the operation of the rainwater pump / sewage pump greatly changes the flow condition in the sewer pipe, there are four types of rainwater when the rainwater pump is operating and when it is not operating, and when the wastewater pump is operating and when it is not operating. Considering an inflow quantity prediction model, rainwater inflow quantity prediction is performed by switching the rainwater inflow quantity prediction model according to the rainfall situation. This enables proper pump operation.

[0013]

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows an outline of a general combined sewer stormwater treatment system. When the rainwater that has fallen into the drainage area 1 reaches the ground, it will flow out of the ground surface, and after a certain time, it will flow into the sewer 2, and the rainwater will move through the sewer 2 and after a certain time, will be pumped to the pump station. Inflow to 3.
At the pump station 3, the rainfall amount 5 from the ground rainfall gauge 4 or the rainfall amount 5 from the radar device 6 is at least at one point or more, and at least one point from the water level gauge 7 at the drainage channel Water level data 8, inflow basin water level 9, pump well water level 10, river water level 11, and pump operating status 12
Based on, rainwater pump (fixed pump 13, variable pump 1
4) and the sewage pump 19 are operated. The purpose of the rainwater treatment system is to promptly eliminate the rainfall in the drainage area into rivers.

In a general pump station, rainfall to the drainage area passes through the sewer pipe and flows into the pump station by natural flow, so that the pump well is installed at a position lower than the discharge destination river or the like. Therefore, unless the rainwater that reaches the pumping station is properly excluded, the pumping station may be flooded, and in the worst case, the nearby urban area may be flooded. Volume prediction and control of the number of rainwater pumps are important issues in rainwater drainage pump stations.

The rainwater pump control device 17 is composed of a rainwater inflow prediction unit 15 and a pump number control unit 16.
The rainwater inflow prediction unit 15 is installed in the drainage area from the ground rain gauge 4 or the radar device 6 at least one or more points in the drainage area, and is also installed in the drainage pipe in the drainage area. Water level data from at least one point from the water level gauge 7 in the channel, inflow channel water level 9 in the pump station, pump well water level 10, and pump operating condition 12
From this, the predicted value of rainwater inflow reaching the pump station after n minutes (hereinafter referred to as the predicted inflow) is calculated.

The pump number control unit 16 inputs the predicted inflow amount after n minutes output from the rainwater inflow prediction unit 15, the pump well water level 10, the water level information from the river water level 11, and the pump operating state 12, The fixed pumps 13, variable pumps 14, and sewage pumps 19 are operated so that the desired rainwater pump operation is performed based on the evaluation scale such as the water level in the pipe after n minutes, the water level in the pump well, and the leveling of the pump operation. To do.

FIG. 2 shows the detailed structure of the rainwater inflow prediction unit 15. The predicted inflow amount 22 after n minutes is calculated based on the water level data 8 in the pipe, the inflow water level 9, the pump well water level 10, the pump operating state 12, and the pump discharge amount 21. In this embodiment, the predicted inflow amount after 5 minutes was used for controlling the number of pumps, and the predicted inflow amount of 10 minutes or more was used as the guidance information.

The exponential smoothing filter 23 removes noise from various measured value data, removes abrupt changes in various water levels associated with the operation of the rainwater pump, and smoothes the output of the inflow predicting calculator 29. Have a function.

The actual rainfall amount calculating section 24 is provided with the rainfall data 5 and the drainage area area data / runoff coefficient from at least one point in the drainage area from the ground rain gauge 4 or the radar device 6 installed in the drainage area. From the data 25, the actual rainfall data 28 flowing into the sewer pipe in the drainage area is calculated.

Drainage area area data / runoff coefficient data 2
Reference numeral 5 stores the controlled area in each drainage area corresponding to the rainfall amount data 5 from the ground rain gauge 4 or the radar device 6 installed in the drainage area and the runoff coefficient for each controlled area.

The measured data storage memory 26 temporarily stores past data of various water level data 27 and actual rainfall data 28 which have passed through the exponential smoothing filter 23. at least,
Explanatory variable (measured value) H1 used in equation (1) described later
It suffices if the time-series data in the time delays t1 to tn, tr of Hn, R are stored.

The inflow prediction operation unit 29 calculates the predicted inflow 22 from various water level data 27 and actual rainfall data 28 at the current time and past time series data of both of them.
The predicted inflow amount 22 is output to the pump number control unit 16 5
It is divided into the estimated inflow amount after a minute and the guidance information output to the guidance display unit 18.

FIG. 3 shows a detailed structure of the inflow amount prediction calculation unit 29. The model switching unit 31 receives an instruction from the model switching determination unit 32, and the rainwater pump operating inflow prediction model 34 and the rainwater pump non-operating inflow prediction model 3 are provided.
5. Which model of the sewage pump operating inflow prediction model 36 and the sewage pump non-operating inflow prediction model 37 is selected. Furthermore, there is a function to select a model according to the prediction time. That is, the estimated inflow after 5 minutes, 1
A rainwater pump operating inflow prediction model 34, a rainwater pump non-operating inflow prediction model 35, a sewage pump operating inflow prediction model 36, and sewage according to the predicted inflow after 0 minutes and the predicted inflow after n minutes. Inflow prediction model 3 when the pump is not operating
Switch the model parameters of 7.

Based on the data in the model switching information file 33, the model switching determination section 32 instructs the model switching section 31 which rainwater inflow prediction model should be adopted at the present time from the pump operating state 12 and the pump discharge 21. . The model switching information file 33 stores information that determines under what conditions the rainwater inflow prediction model is switched.

A rainwater pump operating inflow prediction model 34,
The rainwater pump non-operating inflow prediction model 35, the sewage pump non-operating inflow prediction model 36, and the sewage pump non-operating inflow prediction model 37 are rainwater inflow prediction models, respectively. Rainwater inflow prediction models 34 and 35 and normal rainwater inflow prediction models 36 and 37 except heavy rain. The heavy rain used here refers to the rainfall that the rainwater pump operates,
Normal rainfall refers to rainfall in which rainwater can be removed only with a sewage pump.

These rainwater inflow prediction models are multiple regression model, AR model, ARMA model, tank model, N
Any model such as the euro model and the Fuzzy model may be adopted. However, before applying this example, the rainfall data in the target drainage area, various water level data, the pump operating state, the pump discharge amount, and the actual rainwater inflow amount were analyzed in advance to determine various model parameters. Need to be kept.

The basic formula of the rainwater inflow prediction model adopted in this embodiment is as shown in the following formula (1).

Qin (t + n) = a0 + a1 H1 (t-t1) + a2 H2 (t-t2) + ..... + an Hn (t-tn) + rR (t-tr) (1) Here , Qin (t + n): Rainwater predicted inflow after n minutes (m ³ / min) a0 ~ an, r: Multiple regression model parameter (-) 1 ~ tn, tr: Qin (t + n) The phase (min) at which the correlation of the explanatory variables is maximum (min) H1 to Hn: water level data (m) R: total rainfall data (m ³ / min) in the drainage area.

The above equation is common to the rainwater pump operating inflow prediction model 34, the rainwater pump non-operating inflow prediction model 35, the sewage pump operating inflow prediction model 36, and the sewage pump non-operating inflow prediction model 37. It is a model for predicting rainwater inflow, but the only difference between the models is the multiple regression model parameters a0 to an, r, t1 to tn, tr.

The method of determining the multiple regression model parameters represented by the above equation (1) is as follows. 1) Cross section of the pump station, plan view, and in addition to the sewer cross-sectional view, the water level-reservoir amount in the pump station + storage curve in the sewer pipe
Create FIG.

2) A phase difference that maximizes the correlation between the actual inflow amount and each explanatory variable is calculated.

3) Calculate the actual inflow amount from the balance equation.

[0033]

[Equation 1]

Here, Qin: Actual inflow (m ³ / min) S: Storage in pump station + Storage in sewer (m ³ ) Qout: Pump discharge (m ³ / min) 4) Heavy rainfall It is classified into four patterns: when the rainwater pump operates during rain, when the rainwater pump does not operate, when the sewage pump operates under normal rainfall except heavy rain, and when the sewage pump does not operate.

5) For each of the four patterns classified above, the actual inflow is used as the dependent variable, the water level data, and the total rainfall data as the explanatory variables, and the regression parameters are calculated while considering the phase difference calculated in 2). . The regression parameter obtained here is used as a model parameter 38.
5 minutes later, 10 minutes later ... If you want to obtain some forecasted inflow rate after n minutes, create a rainwater inflow forecasting model for each time you want to forecast.

The model parameter file 38 includes a rainwater pump operating inflow prediction model 34, a rainwater pump non-operating inflow prediction model 35, a sewage pump operating inflow prediction model 36, and a sewage pump non-operating inflow prediction model 36. 37
The multiple regression model parameters of are stored. For example, 5
When it is desired to calculate the four predicted inflow amounts after minutes, 10 minutes, 15 minutes, and 20 minutes, it is sufficient to store 16 model parameters of 4 models × 4 points.

When two types of inflow rate prediction models selected by the model switching section 31 are selected, the model synthesis processing section 39 weights the rainwater inflow rates predicted by the respective models and performs synthesis processing. The result is used as the rainwater inflow predicted value 22.

FIG. 4 shows a water level-pump station storage amount + sewer drain storage amount curve. This storage volume curve is calculated from the civil structure drawing for both the storage volume in the pump station and the storage volume in the sewer pipe.

FIG. 5 further illustrates the model synthesis processing unit 39. The output of the model selected by the model switching unit 31 is the inflow predictive value QinBo during heavy rain.
Two predicted inflow values, i.e., n and QinBoff, and predicted inflow values QinSon and QinSoff during normal rainfall except heavy rain are output.

Inflow forecast values QinBon, QinBoff during heavy rain
Let QinB be the forecasted inflow rate adopted from QinS, and QinS be the forecasted inflow rate adopted from normal rainfall forecast values QinSon and QinSoff except during heavy rain. If it is between the vertical axes QinSC and QinBC of FIG. 5, the model integrator process is performed on the predicted inflow amount Qinp by the following equation.

Qinp = ωS (QinS) * QinS + ωB (QinB) * QinB (3) {QinSC ≦ QinS ≦ QinBC, QinSC ≦ QinB ≦ QinBC} where ωS and ωB are weighting factors of 0 to 1.

The predicted value of each inflow is the vertical axis QinS in FIG.
When it is not between C and QinBC, equations (4) and (5) are inflow predictive values.

Qinp = QinS (4) {QinS ≦ QinSC} Qinp = QinB (5) {QinBC ≦ QinB} However, in the model synthesis processing unit 39, the instruction from the model switching unit 32 causes {QinS ≦ QinSC , QinBC ≤
The situation like QinB} does not occur.

FIG. 6 shows the output result of the model synthesis processing section 39. In the model synthesizing process, each rainwater inflow prediction value is shown in a shaded area in the figure, that is, the outputs of the two rainwater inflow prediction models are QinSC to QinBC.
If it is within the range of QinSC, either the sewage pump operating inflow prediction model 36 or the sewage pump non-operating inflow prediction model 37 is adopted below QinSC, and the rainwater pump operating inflow prediction is above QinBC. Either the model 34 or the rainwater pump non-operating inflow prediction model 35 is adopted.

By performing the model synthesizing process in this way, it is possible to adopt the predicted value of the most accurate region of each rainwater inflow model, which contributes to the improvement of the accuracy of rainwater inflow prediction. In addition, the output value from the inflow amount prediction calculation unit 29 does not become discontinuous when the model is switched.

FIG. 7 shows a selection processing procedure of the rainwater inflow prediction model. The rainfall condition is divided into heavy rainfall and normal rainfall, and the inflow prediction model is selected according to the rainfall condition. However, in order to classify rain conditions into heavy rain and normal rain, field data such as rainfall amount data in the drainage area where the rainwater is to be removed, water level data in the culvert, pump station data, and pump operation patterns are recorded in advance. It is necessary to analyze from. In addition, the classification of the rainfall state is performed by classifying the values of the actual rainfall data 28 as shown in FIG.

FIG. 8 shows classification of models based on actual rainfall. Here, RSC and RBC are the threshold values for model switching, and it is necessary to analyze and determine the field data in advance.

FIG. 9 shows an example in which the rainwater inflow model prediction result Qinpn and the actual inflow amount Qin during heavy rain according to the present invention and the prediction result Qinpo by the conventional rainwater inflow prediction model are shown.

In the prediction result Qinpo by the conventional rainwater inflow prediction model, the tendency of the rainwater inflow prediction error differs depending on whether the rainwater pump is operating or not when compared with the actual inflow quantity Qin. In this example, the prediction accuracy during rainwater pump operation is particularly poor. However, the prediction result Qinpn by the rainwater inflow prediction model according to the present embodiment is close to the actual inflow Qin and the prediction accuracy is good. In this way, when operating the rainwater pump,
The effect was produced by separating the non-operating time and creating a rainwater inflow prediction model. Further, even in the case of rainfall that can be eliminated only by the sewage pump, the inflow prediction accuracy can be improved by separating the sewage pump operation and non-operation.

[0050]

As described above, according to the present invention,
The rainwater inflow amount can be accurately predicted by switching the rainwater inflow prediction model according to the rainfall state and further switching according to the pump operating state. Then, the predicted rainwater inflow is provided as feedforward information to the number-of-pumps control unit so that the pump is properly operated, and the rainwater inflow is presented to the operator as guidance to reduce the mental burden during manual pump operation. Can be planned.

[Brief description of drawings]

FIG. 1 is an explanatory diagram showing an outline of a rainwater treatment system.

FIG. 2 is a configuration diagram of a rainwater inflow prediction unit.

FIG. 3 is a configuration diagram of an inflow amount prediction calculation unit.

FIG. 4 is a diagram showing a relationship between a water level and a storage amount in a pump station and a storage amount curve in a sewer pipe.

FIG. 5 is a diagram illustrating a processing function of a model synthesis processing unit.

FIG. 6 is a diagram showing an output result of a model synthesis processing unit.

FIG. 7 is a diagram showing a processing procedure of inflow prediction.

FIG. 8 is a diagram showing an example of model classification based on actual rainfall.

FIG. 9 is a diagram showing the predicted inflow amount with respect to the actual inflow amount in comparison between the present invention and the prior art.

[Explanation of symbols]

 1 Drainage area 2 Sewer pipe 3 Pump station 4 Ground rain gauge 5 Rainfall data 6 Radar equipment 7 Water level gauge in the drainage pipe 8 Water level data in the drainage pipe 9 Inlet drainage water level 10 Pump well water level 12 Pump operating condition 13 Fixed pump 14 Variable Pump 15 Rainwater inflow prediction unit 16 Number of pumps control unit 17 Rainwater pump control device 18 Guidance display unit 19 Sludge pump

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Internal reference number FI technical display location G05D 7/06 Z 9324-3H (72) Inventor Mikio Yoda 5-2 Omika-cho, Hitachi City, Ibaraki Prefecture No. 1 Incorporated company Hitachi, Ltd. Omika factory (72) Inventor Hitoshi Onizawa 3-2-1, Saicho-cho, Hitachi-shi, Ibaraki Hitachi Engineering Co., Ltd.

Claims (7)

[Claims] 1. Installed in a confluence type or a diversion type sewer system using at least one of rainfall data in a drainage area, water level or flow rate data in a sewer pipe, and water level data in a pump well as an input value. In the method of predicting the amount of inflow water to the drainage pump station, a model for predicting the inflow of rainwater during heavy rain that models the inflow of rainwater flowing into the drainage pump station during heavy rain, and to the drainage pump station during normal rainfall A rainwater inflow prediction model for normal rainfall that models inflowing rainwater inflow is set in advance, and based on the rainwater inflow prediction model for heavy rain during heavy rain, the rainwater inflow during normal rain is introduced based on the heavy rain rainwater inflow prediction model. A method for predicting an inflow water amount at a drainage pump station, comprising: predicting the inflow water amount based on an amount prediction model. 2. The method for predicting inflow of rainwater according to claim 1, wherein the model for predicting inflow of rainwater during heavy rain is a model in which a rainwater pump in the drainage pump station operates, and the model for predicting inflow of rainwater during normal rainfall is A method for predicting an inflow of water at a drainage pump station, which is a model in which only a sewage pump in the drainage pump station operates. 3. Installed in a confluence type or a diversion type sewer system using at least one of rainfall data in the drainage area, water level or flow rate data in the sewer pipe, and water level data in the pump well as an input value. In the method of predicting the inflow of water into the drainage pump station, a model for predicting the inflow of rainwater pumps that models the inflow of rainwater flowing into the drainage pumps during operation of the rainwater pump,
Rainwater pump non-operating inflow forecasting model that models the inflow of rainwater into the drainage pump station when the rainwater pump is not operating, and operating inflow of the sewage pump that models rainwater inflow into the drainage pumping station when the sewage pump is operating A predictive model and a model for predicting the inflow of rainwater that flows into the drainage pump station when the wastewater pump is not operating are set in advance, and the model for predicting the inflow of the rainwater pump when the rainwater is operating is set in advance. Based on the rainwater pump non-operation inflow prediction model and the normal rain, the sewage pump operating inflow prediction model and the sewage pump non-operation inflow prediction model are used to predict the inflow amount. A characteristic method for predicting the inflow of water at drainage pump stations. 4. The method for predicting inflow of water according to claim 1 or 3, wherein when the water level in the sand basin or a sewer pipe near the sand basin is equal to the pump well water level, the sand basin and the sand basin are A method of predicting inflow of rainwater at a drainage pump station, comprising predicting the amount of inflow of water, including the amount of water in a nearby sewer. 5. The inflow water amount prediction method according to claim 1, wherein the rainwater inflow amount prediction value is presented as guidance to an operator at least when the rainwater pump is operating. . 6. Measuring means for measuring the amount of rainfall in the drainage area, the water level or flow rate in the sewer pipe, or the water level of the pump well, the measured rainfall data, the water level or flow rate data in the sewer pipe, Inflow water amount prediction means for predicting the inflow water amount to the drainage pump station installed in the combined or split type sewer system using at least one of the water level data of the pump well as an input value In the equipment, a rainwater inflow prediction model that models the rainwater inflow into the drainage pump station during heavy rain and a normal rainwater inflow that models the rainwater inflow into the drainage pump station during normal rainfall Prediction model, based on the rainwater inflow prediction model during heavy rain during heavy rain, based on the rainwater inflow prediction model during normal rainfall during normal rain, Inflow water amount estimating device in the drainage pump stations, characterized in that it includes a prediction execution means for performing measurement, the. 7. Measuring means for measuring the amount of rainfall in the drainage area, the water level or flow rate in the sewer pipe, or the water level of the pump well, and the measured rainfall data, the water level or flow data in the sewer pipe, Inflow water amount prediction means for predicting the inflow water amount to the drainage pump station installed in the combined or split type sewer system using at least one of the water level data of the pump well as an input value In the equipment, a model for predicting the inflow of rainwater pumps that models the inflow of rainwater that flows into the drainage pump station when the rainwater pumps are operating,
A model for predicting the inflow of rainwater pumps when the rainwater pump flows into the drainage pump station when the rainwater pump is not operating, and an inflow during operation of the wastewater pump that models the rainwater inflow flow into the drainage pump station when the wastewater pump is operating Prediction model, model for predicting rainwater inflow into drainage pump station when sewage pump is not operating, model for predicting inflow when sewage pump is not operating, and model for predicting inflow when rainwater pump is operating and when rainwater pump is not operating during heavy rain A prediction execution means for predicting the inflow water quantity based on the inflow quantity prediction model, and based on the sewage pump operating inflow quantity prediction model and the sewage pump non-operation inflow quantity prediction model during normal rainfall. A device for predicting inflow of water at a drainage pump station, which is characterized in that