Inflow prediction method and inflow prediction device

Abstract

To provide an inflow prediction method capable of simply and accurately predicting a sewage inflow into a sewage treatment plant.SOLUTION: From a sewage inflow D1 into a sewage treatment plant, a rainfall-impact-excluded estimated inflow D2 is obtained, and a rainfall-impact-increased inflow D3 is obtained from a difference between the sewage inflow D1 and the rainfall-impact-excluded estimated inflow D2. Actual treated sewage amounts F1 to F3 of sewage in respective treatment sections C11 to C15 obtained by dividing a treatment area C1 are obtained, and sewage-treatment-amount-corrected average rainfall amounts r1 to r3 are obtained by performing weighted averaging of section average rainfall amounts fallen in the respective treatment sections in accordance with the respective actual treated sewage amounts. The sewage-treatment-amount-corrected average rainfall amounts are added to each other to obtain a treatment-area weighted average rainfall amount H1 of the treatment area C1. An increased inflow prediction model M1 that reproduces a rainfall-impact-increased inflow as a rainfall-impact-increased predicted inflow D4 from the treatment-area weighted average rainfall amount is constructed. The treatment-area weighted average rainfall amount is obtained from a rainfall amount, a rainfall-impact-increased predicted inflow is obtained using the model M1, and the rainfall-impact-excluded estimated inflow is added to obtain a predicted sewage inflow.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an inflow volume prediction method and an inflow volume prediction device for predicting the amount of sewage flowing into a water treatment facility such as a pumping station, a pump building within a sewage treatment plant, or a sewage treatment plant without a pump building. [Background technology]

[0002] Conventionally, it has been important to predict the amount of sewage inflow into water treatment facilities, such as pumping stations, pump buildings within sewage treatment plants, or sewage treatment plants without pump buildings, in order to stabilize the quality of treated water by operating the water treatment facilities appropriately, to plan appropriate personnel allocation for maintenance during rainy weather, and to prevent overflow inside and outside the water treatment facilities.

[0003] Patent Document 1 discloses a method for predicting the amount of sewage inflow into a water treatment facility. This method involves creating an average fluctuation pattern on non-rainy days, subtracting the average fluctuation pattern from the inflow sewage amount to create residual data (sewage inflow data for rainfall only), creating the relationship between the residual data and rainfall intensity as a system variable in a statistical model, using this to calculate a predicted value for the inflow on rainy days, and predicting the inflow sewage amount by adding the average fluctuation pattern and the predicted value for the inflow on rainy days.

[0004] Patent document 2 also discloses a method for creating an inflow prediction model by selecting an observation mesh for a rainwater radar, receiving weather information using the inflow prediction model, and predicting the amount of rainwater inflow, in which the user can adjust the waveform of the inflow prediction value and the flow time while looking at the calculation results of the inflow prediction value. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Laid-Open No. 2000-345604 [Patent Document 2] Japanese Patent Application Laid-Open No. 2007-255114 Summary of the Invention [Problem to be solved by the invention]

[0006] However, the treatment areas covered by water treatment facilities such as pumping stations are large, and as a result, not only do the rainfall conditions within the treatment areas differ from area to area, but the impact of rainfall in each area on the amount of sewage inflow into the water treatment facilities is also not the same.

[0007] However, Patent Documents 1 and 2 do not accurately take into account the degree to which rainfall in each area within the treatment area (target watershed) responsible for the water treatment facility affects the amount of sewage inflow into the water treatment facility, making it difficult to more accurately predict the amount of sewage inflow into the water treatment facility using a simple method.

[0008] The present invention has been made in consideration of the above points, and its purpose is to provide an inflow prediction method and an inflow prediction device that can predict the amount of sewage inflow into a water treatment facility in a simple, inexpensive, and accurate manner. [Means for solving the problem]

[0009] The present invention provides an inflow prediction method for predicting the sewage inflow of sewage collected from a treatment area to a water treatment facility, which includes: determining an estimated inflow excluding rainfall effects from the sewage inflow into the water treatment facility when there is no effect of rainfall in the treatment area; determining an increased inflow excluding rainfall effects from the difference between the sewage inflow into the water treatment facility when rainfall occurs and the estimated inflow excluding rainfall effects; determining an actual sewage treatment volume of sewage collected in each of a plurality of treatment sections obtained by dividing the treatment area; and determining a sewage treatment volume-corrected average rainfall for each treatment section by averaging the treatment section average rainfall that fell in each treatment section in accordance with the actual sewage treatment volume for each treatment section. The method is characterized in that the treatment area slope average rainfall for the entire treatment area is first calculated by adding these sewage treatment volume corrected average rainfalls together, and an increased inflow prediction model is constructed from the treatment area slope average rainfall to reproduce the rainfall-influenced increased inflow as a rainfall-influenced predicted increased inflow.Next, the treatment area slope average rainfall is calculated from the rainfall input as meteorological information, and the rainfall-influenced predicted increased inflow is calculated from the treatment area slope average rainfall using the increased inflow prediction model.The sewage inflow to the water treatment facility is predicted by adding the rainfall-influenced estimated inflow to the calculated rainfall-influenced predicted increased inflow. Here, the entire method (or series of programs) for carrying out a series of operations to predict the amount of sewage inflow into the water treatment facility will be referred to as an inflow prediction model. According to the present invention, by using the above-mentioned inflow prediction model, the only input data required for predicting sewage inflow during operation is the rainfall amount obtained from meteorological information, i.e., the model can be introduced and operated extremely easily without using actual measurement data of real-time sewage inflow to the water treatment facility or complex calculation means such as machine learning, and the sewage inflow flowing into the water treatment facility can be predicted. In addition, the slope average rainfall for the treatment area is used as the meteorological information to construct the increased inflow prediction model and to input into the increased inflow prediction model, thereby correcting the degree of impact on the sewage inflow for each treatment category.This allows for more accurate prediction of sewage inflow compared to predictions made using an increased inflow prediction model constructed simply using the average rainfall for all rainfall falling in the treatment area.

[0010] In addition to the above-mentioned features, the present invention is characterized in that, if the actual sewage inflow into the water treatment facility when there is no influence of rainfall increases or decreases over time compared to the previously determined estimated inflow excluding the influence of rainfall, the estimated inflow excluding the influence of rainfall is corrected to increase or decrease in accordance with the increase or decrease without re-obtaining the estimated inflow excluding the influence of rainfall. The estimated inflow excluding the effects of rainfall used in the inflow prediction model to predict sewage inflow is based on past data, and therefore changes depending on the water usage in the treatment area and changes in the condition of the sewer facilities, etc., and as a result, the predicted sewage inflow using the inflow prediction model may deviate from the actual measured value. According to the present invention, even when the discrepancy between the actual measured value and the predicted value increases due to changes in the estimated inflow excluding the influence of rainfall, the newly acquired data (estimated inflow excluding the influence of rainfall) can be used as past data to correct the predicted value of the sewage inflow to the water treatment facility obtained using the inflow prediction model, thereby making it possible to make highly accurate predictions of sewage inflow over a long period of time.

[0011] In addition to the above features, the present invention is also characterized in that, if the actual rainfall-affected increased inflow increases or decreases over time compared to the rainfall-affected predicted increased inflow calculated using the increased inflow prediction model, a correction is made to increase or decrease the rainfall-affected predicted increased inflow calculated using the increased inflow prediction model in accordance with the increase or decrease, without changing the increased inflow prediction model. The rainfall-influenced increased inflow used to construct the increased inflow prediction model is based on past data, and therefore changes depending on the water usage in the treatment area, changes in the condition of the sewer facilities, changes in the state of residential development, etc., and as a result, the rainfall-influenced increased inflow predicted using the increased inflow prediction model may deviate from the actual measured value. According to the present invention, even when the discrepancy between the actual measured value and the predicted value increases with changes in the rainfall-influenced increased inflow, it is possible to accurately predict sewage inflow over a long period of time by correcting the rainfall-influenced predicted increased inflow obtained using the increased inflow prediction model without recreating the increased inflow prediction model using newly acquired data (rainfall-influenced increased inflow) as past data.

[0012] In addition to the above features, the present invention is also characterized in that if the actual rainfall-influenced increased inflow changes due to temporal fluctuations in the arrival time of the impact of rainfall at the water treatment facility compared to the rainfall-influenced predicted increased inflow calculated using the increased inflow prediction model, the rainfall-influenced predicted increased inflow calculated using the increased inflow prediction model is corrected to match the fluctuations in the arrival time without changing the increased inflow prediction model. The arrival time of increased inflow due to rainfall is based on past data, and may therefore deviate from the actual measured value due to changes in water usage in the treatment area, changes in the condition of sewer facilities, and changes in the state of residential development. According to the present invention, even if the deviation between the actual measured value and the predicted value becomes large as the arrival time of the rainfall-influenced increased inflow changes, it is possible to make highly accurate predictions of sewage inflow over a long period of time by correcting the arrival time of the rainfall-influenced predicted increased inflow obtained using the increased inflow prediction model without recreating the increased inflow prediction model.

[0013] The present invention also provides an inflow volume prediction device for predicting the sewage inflow volume of sewage collected from a treatment area to a water treatment facility, comprising: means for calculating an estimated inflow volume excluding rainfall effects from the sewage inflow volume to the water treatment facility when there is no effect of rainfall in the treatment area; means for calculating an increased inflow volume affected by rainfall from the difference between the sewage inflow volume to the water treatment facility when rainfall occurs and the estimated inflow volume excluding rainfall effects; and means for calculating an actual sewage treatment volume of sewage collected in each of a plurality of treatment sections obtained by dividing the treatment area, and calculating a sewage treatment volume corrected average rainfall for each treatment section by averaging the treatment section average rainfall that fell in each treatment section in accordance with the actual sewage treatment volume for each treatment section. The system is characterized by having: a means for calculating the treatment area slope average rainfall for the entire treatment area by adding each sewage treatment volume corrected average rainfall; a means for constructing an increased inflow prediction model that reproduces the rainfall-influenced increased inflow as a rainfall-influenced predicted increased inflow from the treatment area slope average rainfall; a means for calculating the treatment area slope average rainfall from rainfall input as meteorological information and for calculating the rainfall-influenced predicted increased inflow from the treatment area slope average rainfall using the increased inflow prediction model; and a means for predicting the sewage inflow to the water treatment facility by adding the rainfall-influenced estimated inflow excluding the estimated inflow. Here, the entire means for carrying out a series of operations to predict the amount of sewage inflow into the water treatment facility is referred to as an inflow prediction model. According to the present invention, by constructing the above-mentioned inflow prediction model, the sewage inflow into a water treatment facility can be predicted by a simple method of using only the rainfall obtained from meteorological information when making actual predictions, that is, without using actual measurement data of the real-time sewage inflow of the water treatment facility or complex calculation methods such as machine learning. In addition, the slope average rainfall for the treatment area is used as the meteorological information to construct the increased inflow prediction model and to input into the increased inflow prediction model, thereby correcting the degree of impact on the sewage inflow for each treatment category.This allows for more accurate prediction of sewage inflow compared to predictions made using an increased inflow prediction model constructed simply using the average rainfall for all rainfall falling in the treatment area. [Effects of the Invention]

[0014] According to the present invention, it is possible to predict the amount of sewage inflow into a water treatment facility in a simple, inexpensive and accurate manner. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a diagram illustrating an example of an inflow prediction device 10. FIG. [Figure 2] 1 is a schematic diagram conceptually showing all of the basins C1 in which a sewage treatment plant (water treatment facility) A1 treats sewage. [Figure 3-1] FIG. 10 is a diagram showing an example of a history of a sewage inflow D1 and an estimated inflow D2 excluding the influence of rainfall. [Figure 3-2] This is a diagram showing an example of the history of rainfall-influenced increased inflow D3 and treatment area slope average rainfall H1. [Figure 3-3] FIG. 10 is a diagram showing an example of a history of predicted rainfall-influenced increase inflow D4 and predicted sewage inflow D5. [Figure 4] FIG. 10 is a diagram illustrating a method for calculating the treatment area slope average rainfall amount. [Figure 5] This is a diagram showing the average rainfall over the treatment area slope (including predicted values) and the amount of sewage inflow (including predicted values) flowing into sewage treatment plant A. DETAILED DESCRIPTION OF THE INVENTION

[0016] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. 1 is a diagram showing an example of an inflow volume prediction device 10 according to one embodiment of the present invention. The inflow volume prediction device 10 is a device that predicts the amount of sewage inflow into a sewage treatment plant (sewage treatment facility, water treatment facility) A1, which will be described below, and in this example, a personal computer is used.

[0017] As shown in the figure, the inflow prediction device 10 is configured with a memory unit 110 that stores various information used for prediction, a prediction processing block 130 used to calculate the prediction of the sewage inflow volume flowing from the treatment area C1 described below to the sewage treatment plant A1 described below, an output means 200 that outputs various information, and an input means 220 that inputs various commands and information. The memory unit 110, prediction processing block 130, output means 200, and input means 220 function by driving software (programs) and hardware possessed by the computer that constitutes the inflow prediction device 10.

[0018] The output means 200 is, for example, a screen display device, and displays various screens such as those shown in Figures 3-1 to 3-3 and 5. The output means 200 may also be a means for outputting to a printing device or other equipment.

[0019] The input means 220 is, for example, a means for inputting various commands from a keyboard, mouse, or other device, and instructs the output means 200 to display various screens, and instructs various processes in the prediction processing block 130 described below.

[0020] 2 is a conceptual schematic diagram showing the entire basin (hereinafter referred to as the "treatment area") C1 where sewage treatment plant A1 treats sewage. As shown in the figure, in this example, treatment area C1 has three treatment sections C11, C13, and C15. Sewage flowing down in treatment section C11 is collected at relay pumping station P1 and merges into the next treatment section C13. Sewage flowing down in treatment section C13, together with the sewage from treatment section C11 collected at relay pumping station P1, is collected at relay pumping station P2 and merges into the next treatment section C15. Sewage flowing down in treatment section C15 is then collected at sewage treatment plant A1 together with the sewage from treatment sections C11 and C13 collected at relay pumping station P2. The sewage treated at sewage treatment plant A1 is discharged into, for example, a river. In Figure 2, R1, R2, and R3 represent images of rainfall for treatment categories C11, C13, and C15, respectively.

[0021] Returning to Figure 1, the memory unit 110 of the inflow prediction device 10 receives inputs from outside, such as meteorological information N1, geographical information N2, information on the sewage inflow D1 to the sewage treatment plant A1, and information on the actual sewage treatment volumes F1, F2, and F3 for each treatment section C11, C13, and C15.

[0022] The weather information N1 is weather information distributed over a network. The weather information N1 input to the storage unit 110 is, for example, actual measurement data (including short-term forecast data; the same applies below) such as rainfall at the same time for an area including the processing area C1 (consisting of three processing divisions C11, C13, and C15; the same applies below). Specifically, analyzed rainfall and short-term rainfall forecast data (1-km mesh data) provided by a weather information company are used. As described above, the actual measurement data is meteorological data in mesh units, and information on rainfall, longitude, latitude, and time is linked to each mesh unit, i.e., each area unit (1-km mesh unit) obtained by dividing the area into multiple smaller areas. It is preferable to obtain such data for, for example, several years' worth when creating the various data described below. The weather information N1 is input from a weather information company, for example, every 30 minutes.

[0023] The geographic information N2 is, for example, geographic information about the processing area C1 input from a geographic information system (GIS) distributed over a network.

[0024] The sewage inflow D1 into the sewage treatment plant A1 can be measured using various sensors installed in the sewage treatment plant A1, such as water level meters and flow meters installed in inflow channels, grit basins, pump wells, etc. within the sewage treatment plant A1. The measured sewage inflow D1 is stored in the memory unit 110 as a history of the sewage inflow D1 described below.

[0025] The actual sewage treatment volume F1 for treatment section C11 is measured at or near relay pumping station P1. The actual sewage treatment volume F2 for treatment section C13 (the inflow volume obtained by subtracting the actual sewage treatment volume F1 from the total inflow volume) is measured at or near relay pumping station P2. These measurements are performed using a water level gauge, flow meter, etc., similar to the method used to measure the sewage inflow volume D1 at sewage treatment plant A1. The actual sewage treatment volume F3 for treatment section C15 is calculated by subtracting the actual sewage treatment volumes F1 and F2 from the total sewage inflow volume D1 measured at sewage treatment plant A1. Note that if the sewage collected at relay pumping station P1 or P2 is transported directly to sewage treatment plant A1 independently, this subtraction is not necessary. These actual sewage treatment volumes F1, F2, and F3 are used to calculate the gradient-averaged rainfall for the treatment area described below, but historical data, such as the past year's data, is used as the actual sewage treatment volume.

[0026] Next, the memory unit 110 stores the created increased inflow prediction model M1, the history of sewage inflow D1 to the sewage treatment plant A1, the history of actual sewage treatment volumes F1, F2, and F3 for each treatment section C11, C13, and C15, geographic information N2, and the history of meteorological information N1.

[0027] The created increased inflow prediction model M1 is a model that is created in advance before the actual sewage inflow prediction in the prediction processing block 130 described below. This increased inflow prediction model M1 is a program that uses meteorological information N1 input to the memory unit 110 to calculate the predicted increased inflow due to rainfall (described below) to the sewage treatment plant A1 that will increase due to rainfall.

[0028] The history of sewage inflow D1 is historical data of past sewage inflow D1 that has flowed into the sewage treatment plant A1, and can be obtained from the sewage inflow D1 to the sewage treatment plant A1 that is input into this memory unit 110 (of course, instead of inputting the sewage inflow D1 directly into the memory unit 110, it is also possible to configure the memory unit 110 to store historical data of sewage inflow D1 that has been obtained and created by another system).

[0029] Figure 3-1(A) is a diagram showing an example of the history of sewage inflow D1. This history of sewage inflow D1 shows the history of sewage inflow D1 from August 23 to November 7 of a certain year, and it can be seen that the sewage inflow D1 to the sewage treatment plant A1 increases on days when it rains.

[0030] Returning to Figure 1, the history of actual sewage treatment volumes F1, F2, F3 for each treatment section C11, C13, C15 is the actual sewage treatment volumes F1, F2, F3 for each treatment section C11, C13, C15 stored over time in memory unit 110.

[0031] The geographic information N2 is geographic information relating to the processing area C1 (three processing divisions C11, C13, C15) input from the geographic information system (GIS).

[0032] The history of the weather information N1 is data in which the weather information N1 input to the storage unit 110 is stored at predetermined time intervals.

[0033] The prediction processing block 130 includes an increased inflow prediction model construction block 140 that creates an increased inflow prediction model (hereinafter also referred to as "model") M1, which is a program that predicts the increased sewage inflow into the sewage treatment plant A1 that increases due to rainfall using rainfall data, a sewage inflow prediction block 160 that predicts the increased sewage inflow into the sewage treatment plant A1 that increases due to rainfall using the increased inflow prediction model M1, and various prediction correction blocks 180 that correct the predicted value of the sewage inflow.

[0034] The increased inflow prediction model construction block 140 and the sewage inflow prediction block 160 are collectively referred to as the inflow prediction model MD. In other words, the inflow prediction model MD refers to the entire method (model of the inflow prediction method) for performing a series of operations from constructing the increased inflow prediction model M1 to predicting the sewage inflow to the sewage treatment plant A, or the entire series of programs (model of the inflow prediction program), or the entire means for performing the series of operations (model of the inflow prediction device).

[0035] The increased inflow prediction model construction block 140 has a rainfall-influence-excluding estimated inflow calculation function 141, a rainfall-influenced increased inflow calculation function 143, a treatment area slope average rainfall calculation function 145, and an increased inflow prediction model construction function 147.

[0036] The rainfall-free estimated inflow calculation function 141 calculates the estimated sewage inflow (rainfall-free estimated inflow D2) to the sewage treatment plant A1, assuming no rainfall influence in the treatment area C1. This rainfall-free estimated inflow D2 is set (estimated) by collecting multiple years of historical sewage inflow D1 data, as shown in Figure 3-1(A), and selecting data for periods when rainfall does not influence the inflow, taking into account the influences of season, day of the week, and time of day. The setting (estimation) method may be, for example, automatic setting by a program or manual setting by collecting data for periods when rainfall does not influence the inflow. A period affected by rain refers to the sewage inflow for several days (e.g., two days) following rainfall (e.g., rainfall of 0.5 mm / h or more). The data from that period is replaced with data from another year without rainfall, during the same season, on the same day of the week, and at the same time. Figure 3-1(B) shows historical data illustrating an example of the set rainfall-free estimated inflow D2.

[0037] The rainfall-influenced increased inflow calculation function 143 is a functional unit that calculates the difference between the past sewage inflow D1 to the sewage treatment plant A during rainfall in the treatment area C1 and the rainfall-influence-free estimated inflow D2 to obtain the rainfall-influenced increased inflow D3. In this example, for example, the history of the sewage inflow D1 shown in Figure 3-1(A) is used as the past sewage inflow D1, and the rainfall-influence-free estimated inflow D2 shown in Figure 3-1(B) is subtracted from this history of the sewage inflow D1 with the same season, day of the week, and time of day set. Figure 3-2(C) shows data showing an example of the rainfall-influenced increased inflow D3 obtained as described above.

[0038] The treatment area gradient average rainfall calculation function 145 does not simply calculate the average amount of rainfall that fell in the treatment area C1 (treatment area average rainfall), but calculates the treatment area gradient average rainfall by adding up the average rainfall that fell in each of the treatment sections C11, C13, and C15 that make up the treatment area C1, weighted by the amount of sewage inflow that actually flows into the sewage treatment plant A from each of the treatment sections C11, C13, and C15. Rain that fell in each of the treatment sections C11, C13, and C15 does not affect the amount of sewage inflow into the sewage treatment plant A with the same probability, but rather focuses on the fact that the degree of influence on the amount of sewage inflow that increases with rainfall differs depending on the average sewage inflow of each of the treatment sections C11, C13, and C15. Based on past data, the gradient average rainfall that is weighted for each of the treatment sections C11, C13, and C15 relative to the amount of sewage inflow into the sewage treatment plant A is calculated. For example, this takes into consideration the fact that the amount of sewage inflow from each treatment section C11, C13, C15 to sewage treatment plant A that increases with rainfall will differ depending on the degree of residential development in each treatment section C11, C13, C15, the amount of rain that flows into the river (the amount that does not flow into the sewer), etc.

[0039] 4 is a diagram specifically illustrating a method for calculating the gradient-average rainfall for a treatment area. As shown in the figure, the past (predetermined period, e.g., one year) actual sewage treatment volumes F1, F2, and F3 for each treatment section C11, C13, and C15 are read from the storage unit 110 and used. Furthermore, past (predetermined period, e.g., one year) meteorological information N1 for each treatment section C11, C13, and C15 is read from the storage unit 110 to calculate the average rainfall volumes R1, R2, and R3 for each treatment section. The past average rainfall volumes R1, R2, and R3 for each treatment section are calculated from the history of meteorological information N1 from the time when the past actual sewage treatment volumes F1, F2, and F3 were measured.

[0040] The sewage treatment volume-adjusted average rainfall amounts r1, r2, and r3 are calculated for each treatment section C11, C13, and C15 using the formula shown in Figure 4. Specifically, the sewage treatment volume-adjusted average rainfall amounts r1, r2, and r3 are the average rainfall amounts for each treatment section C11, C13, and C15, adjusted by weighting the treatment section average rainfall amounts R1, R2, and R3 with the actual sewage treatment volumes F1, F2, and F3. The sum of these sewage treatment volume-adjusted average rainfall amounts r1, r2, and r3 is calculated as the treatment area gradient average rainfall amount. Figure 3-2(D) shows an example of the treatment area gradient average rainfall amount H1. Using the treatment area gradient average rainfall amount H1, the average rainfall amount can be estimated, taking into account the impact of rainfall in each treatment section C11, C13, and C15 on the sewage inflow amount flowing into sewage treatment plant A.

[0041] The incremental inflow prediction model construction function 147 is a function for constructing an incremental inflow prediction model M1 for reproducing the rainfall-influenced incremental inflow D3 shown in Figure 3-2(C) from the treatment area slope average rainfall H1 as shown in Figure 3-2(D). Specifically, it finds the relationship between the average rainfall and rainfall duration at the treatment area slope average rainfall H1 in Figure 3-2(D) and the rainfall-influenced incremental inflow D3 in Figure 3-2(C), that is, it constructs the incremental inflow prediction model M1 consisting of a program for finding the rainfall-influenced incremental inflow D3 as the rainfall-influenced predicted incremental inflow D4 as shown in Figure 3-3(E) from the average rainfall and rainfall duration at the treatment area slope average rainfall H1.

[0042] The above is the function of the increased inflow prediction model construction block 140. By adding the predicted increased inflow D4 affected by rainfall shown in Figure 3-3(E) and the estimated inflow D2 excluding the influence of rainfall shown in Figure 3-1(B), the predicted sewage inflow D5 shown in Figure 3-3(F) is obtained. This predicted sewage inflow D5 is similar to the history of the sewage inflow D1 in Figure 3-1(A), which means that it is possible to predict the sewage inflow that is close to the actual sewage inflow.

[0043] Returning to Figure 1, the sewage inflow prediction block 160 is a functional part that predicts the sewage inflow flowing into the sewage treatment plant A using the weather information N1 obtained from outside and the constructed increased inflow prediction model M1, and is configured to have a weather information input function 161, a treatment area slope average rainfall calculation function 163, a rainfall-influenced predicted increased inflow calculation function 165 using the increased inflow prediction model M1, and a sewage inflow calculation function 167.

[0044] The weather information input function 161 is a functional unit for inputting weather information N1 distributed over the network. The weather information N1 to be input may be, for example, weather information for the past 10 days, current weather information, and weather forecast information for the next 15 hours.

[0045] The treatment area gradient average rainfall calculation function 163 calculates the treatment area gradient average rainfall from the average rainfall over the past few days to the future for each treatment area C11, C13, and C15 using input weather information N1 covering the past few days to the present and the predicted future several hours from the present, and the geographic information N2 stored in the storage unit 110. The average rainfall is the rainfall amount obtained by averaging the rainfall amount per mesh (1-km mesh unit) for the corresponding treatment area. The treatment area gradient average rainfall (r1 + r2 + r3) is calculated using the same formula as shown in Figure 4 above. The actual sewage treatment volumes F1, F2, and F3 used in this calculation are the past actual sewage treatment volumes F1, F2, and F3 used to calculate the treatment area gradient average rainfall shown in Figure 4 above. Therefore, the only newly acquired data used to predict the sewage inflow in the sewage inflow prediction block 160 is the weather information N1.

[0046] The rainfall-affected predicted increase in inflow calculation function 165 is a functional part that calculates the rainfall-affected predicted increase in inflow from the past to the future using the calculated treatment area slope average rainfall (r1+r2+r3) from the past to the future and the increased inflow prediction model M1 previously calculated in the increased inflow prediction model construction block 140.

[0047] The sewage inflow calculation function 167 is a functional part that calculates the predicted value of the sewage inflow amount flowing into the sewage treatment plant A from the past to the future. In this functional part, the predicted increase in inflow amount due to rainfall is calculated by adding the part of the estimated inflow amount excluding the influence of rainfall D2 shown in Figure 3-1(B) that corresponds to the season, day of the week, and time of day.

[0048] Figure 5 shows the gradient-averaged rainfall over the treatment area (including predicted values) and the sewage inflow (including predicted values) flowing into sewage treatment plant A. Figure 5(A) shows the gradient-averaged rainfall over the treatment area from the past (9:00 on July 14th) to the future (1:00 on July 16th) as viewed from the current time (8:00 on July 15th). The solid line shows the measured values ​​input from meteorological information N1 from the past to the present. The dashed line shows the forecast values ​​input from meteorological information N1 from the present to the future.

[0049] Figure 5(B) shows the sewage inflow into WWTP A from the past (9:00 on July 14) to the future (1:00 on July 16) as viewed from the present (8:00 on July 15). The dashed-dotted line shows the predicted increase in inflow due to rainfall, and the dotted line shows the predicted sewage inflow. The solid line shows the measured sewage inflow that was added later for use in correcting the prediction described below after calculating the predicted sewage inflow.

[0050] As shown in the figure, the time of rainfall and the time when the predicted sewage inflow increases are different, and the predicted sewage inflow and the actual measured sewage inflow are almost the same.

[0051] As described above, this embodiment is an inflow prediction method for predicting the sewage inflow of sewage collected from a treatment area C1 to a sewage treatment plant (water treatment facility) A1, and involves calculating an estimated inflow D2 excluding the influence of rainfall from the sewage inflow to the sewage treatment plant A1 when there is no influence of rainfall in the treatment area C1, and calculating an increased inflow D3 influenced by rainfall from the difference between the sewage inflow to the sewage treatment plant A1 during rainfall and the estimated inflow D2 excluding the influence of rainfall. The actual sewage treatment volumes F1, F2, and F3 of the sewage collected in each of the treatment sections C11, C13, and C15 are calculated, and the average rainfall volumes R1, R2, and R3 that fell in each of the treatment sections C11, C13, and C15 are averaged according to the actual sewage treatment volumes F1, F2, and F3 for each of the treatment sections C11, C13, and C15 to calculate the adjusted average rainfall volumes r1, r2, and r3 for each of the treatment sections C11, C13, and C15. These adjusted average rainfall volumes r1, r2, and r3 are then added together. The rainfall-influence-excluded estimated inflow D2 is added to the obtained rainfall-influence-excluded estimated inflow D2, thereby predicting the predicted sewage inflow to the sewage treatment plant A1. Therefore, the sewage inflow to the sewage treatment plant A1 can be predicted by a simple method in which only the rainfall obtained from the weather information N1 is used as the information used to actually predict the predicted sewage inflow. In other words, the sewage inflow to the sewage treatment plant A1 can be predicted by a simple method in which the only information used to actually predict the predicted sewage inflow is the rainfall obtained from the weather information N1. In other words, the sewage inflow to the sewage treatment plant A1 can be predicted by a simple method in which the only information used to actually predict the predicted sewage inflow is the rainfall obtained from the weather information N1. In other words, the sewage inflow to the sewage treatment plant A1 can be predicted by a simple method in which the only information used to actually predict the predicted sewage inflow is the rainfall obtained from the weather information N1. In other words, without using complicated simulations using machine learning or actual measurement data of the real-time sewage inflow to the sewage treatment plant A1.

[0052] Furthermore, the treatment area slope average rainfall H1 is used to construct the increased inflow prediction model M1, which corrects the degree of impact on the sewage inflow collected at sewage treatment plant A for treatment divisions C11, C13, and C15. This allows for more accurate prediction of sewage inflow compared to predictions using an increased inflow prediction model constructed simply using the average rainfall of all rain falling in treatment area C1.

[0053] Returning to Figure 1, the prediction correction block 180 is a functional part that corrects the predicted value of the sewage inflow while using the constructed inflow prediction model MD itself without any changes, and has a rainfall influence excluding estimated inflow correction function 181, a rainfall influence increase predicted inflow correction function 183, and a rainfall influence increase predicted inflow arrival time correction function 185.

[0054] Because the constructed inflow prediction model MD is based on limited past data, there is a possibility that the deviation between the actual measured value and the predicted value will become large due to changes in the water usage status of the treatment area C1 and the condition of the sewer facilities. This prediction correction block 180 has the function of correcting the predicted sewage inflow value without changing the inflow prediction model MD itself, even if a deviation between the actual measured value and the predicted value occurs. In other words, it is possible to continue using the same inflow prediction model MD over a long period of time without having to recreate the inflow prediction model MD using newly acquired data as past data.

[0055] The rainfall-influence-excluded estimated inflow correction function 181 compares, for example, in Figure 5(B), the predicted sewage inflow shown by the dotted line with the actually measured sewage inflow shown by the solid line added later, and if there is a difference between the predicted sewage inflow when there is no rainfall and the actually measured sewage inflow, the solid line and the dotted line will be separated in the vertical axis (up and down direction), so the predicted sewage inflow shown by the dotted line is moved in the vertical axis direction to match them, and a correction is made in the direction of increase or decrease in flow rate to increase or decrease. This correction is made, for example, by displaying Figure 5 as an image on the screen of the output means 200 and operating the input means 220 while looking at this screen (the same applies to the rainfall-influence increase predicted inflow correction function 183 and the rainfall-influence increase predicted inflow arrival time correction function 185 described below. Note that this correction may also be configured to be performed automatically).

[0056] For example, in Figure 5(B), the rainfall-affected predicted inflow correction function 183 compares the predicted sewage inflow shown by the dotted line with the actual measured sewage inflow shown by the added solid line, and if there is a difference between the predicted sewage inflow at the time when it increases due to rainfall (from 1:00 on July 15th) and the actual measured sewage inflow, the time when it is affected by rain (from 1:00 on July 15th) on the solid line and the dotted line is separated along the vertical axis, so the rainfall-affected predicted inflow increase shown by the dot-dash line in Figure 5(B) is moved along the vertical axis and a correction is made in the direction of increase or decrease of the flow rate so that the actual measured sewage inflow and predicted sewage inflow at the time when it increases due to rainfall (from 1:00 on July 15th) match.

[0057] The rainfall-affected increased inflow arrival time correction function 185 compares, for example, the predicted sewage inflow volume shown by the dotted line in Figure 5(B) with the actual measured sewage inflow volume shown by the added solid line, and if there is a temporal difference between the predicted sewage inflow volume when it increases due to rainfall and the actual measured sewage inflow volume, causing a deviation in the horizontal axis direction (left and right direction), it moves the rainfall-affected increased inflow volume shown by the dotted line in Figure 5(B) in the horizontal axis direction (time direction, hourly direction) to make a correction so that the actual measured sewage inflow volume and the predicted sewage inflow volume match.

[0058] As explained above, by using the rainfall-influence-excluded estimated inflow correction function 181, if the actual rainfall-influence-excluded estimated inflow increases or decreases over time compared to the rainfall-influence-excluded estimated inflow D2 at the time the inflow prediction model MD was constructed, the predicted value of the sewage inflow to the water treatment facility A1 (predicted sewage inflow) calculated using the inflow prediction model MD is corrected to increase or decrease in accordance with the increase or decrease without changing the inflow prediction model MD.Therefore, even if the discrepancy between the actual measured value and the predicted value of the rainfall-influence-excluded estimated inflow, the same inflow prediction model MD can be continuously used over a long period of time without having to rebuild the inflow prediction model using newly acquired data as past data.

[0059] Furthermore, by using the rainfall-influenced increased predicted inflow correction function 183, if the actual rainfall-influenced increased inflow increases or decreases over time compared to the rainfall-influenced increased predicted inflow calculated using the increased inflow prediction model M1, the rainfall-influenced increased predicted inflow calculated using the increased inflow prediction model M1 is corrected to increase or decrease in accordance with the increase or decrease without changing the increased inflow prediction model M1. Therefore, even if the discrepancy between the actual measured value and the predicted value of the rainfall-influenced increased inflow becomes large, the same increased inflow prediction model M1 can be continuously used over a long period of time.

[0060] Furthermore, by using the rainfall-influenced increased predicted inflow arrival time correction function 185, if the actual rainfall-influenced increased inflow changes due to temporal fluctuations in the arrival time of the impact of rainfall at the water treatment facility compared to the rainfall-influenced increased predicted inflow calculated using the increased inflow prediction model M1, the rainfall-influenced increased predicted inflow calculated using the increased inflow prediction model M1 is corrected in the time direction to match the fluctuations in the arrival time without changing the increased inflow prediction model M1, so that even if the discrepancy between the actual measured value and the predicted value due to fluctuations in the arrival time of the rainfall-influenced increased inflow becomes large, the same increased inflow prediction model M1 can be continuously used over a long period of time.

[0061] As described in the above embodiment, the present invention is an inflow volume prediction method and an inflow volume prediction device for predicting the amount of sewage inflow into a water treatment facility, but it may also be configured to automatically and directly control the operation of the sewage treatment plant A1 (such as adjusting the amount of sewage inflow into the sewage treatment plant A1) in accordance with the sewage inflow volume predicted using the present invention. In this case, the present invention becomes an operation control method and an operation control device for a water treatment facility.

[0062] Furthermore, it is preferable to configure the system so that, if the predicted amount of sewage inflow into the water treatment facility exceeds, for example, a predetermined caution standard value, a caution alarm (such as sound or email) is issued.

[0063] Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments and various modifications are possible within the scope of the claims and the technical concept described in the specification and drawings. Furthermore, any configuration not directly described in the specification and drawings is within the scope of the technical concept of the present invention as long as it achieves the functions and effects of the present invention. For example, in the above embodiments, a sewage treatment plant was used as an example of a water treatment facility. However, the water treatment facility may be a relay pumping station or a pipeline facility with an inflow measurement device. The sewage treatment plant may be a sewage treatment plant with a pump building or may not be a sewage treatment plant with a pump building. Similarly, the relay pumping stations P1 and P2 may instead be sewage treatment plants or pipeline facilities with an inflow measurement device.

[0064] Furthermore, the embodiments described above and shown in the drawings can be combined with each other as long as there is no contradiction in their purpose, configuration, etc. Furthermore, even a part of the description described above and the drawings can be an independent embodiment, and the embodiment of the present invention is not limited to a single embodiment combining the description described above and the drawings. [Explanation of symbols]

[0065] A1... sewage treatment plant (water treatment facility), C1... treatment area, C11, C13, C15... treatment classification, D1... sewage inflow, D2... estimated inflow excluding the influence of rainfall, D3... increased inflow influenced by rainfall, D4... predicted increased inflow influenced by rainfall, F1, F2, F3... actual sewage treatment volume, R1, R2, R3... average rainfall for treatment classification, r1, r2, r3... average rainfall adjusted for sewage treatment volume, H1... average rainfall on the treatment area slope, M1... increased inflow prediction model, MD... inflow prediction model, N1... weather information.

Claims

1. 1. A method for predicting the amount of sewage inflow collected from a treatment area to a water treatment facility, comprising: In advance, calculating an estimated inflow excluding the influence of rainfall from the sewage inflow into the water treatment facility when there is no influence of rainfall in the treatment area; and calculating an increased inflow amount affected by rainfall from the difference between the sewage inflow amount into the water treatment facility during rainfall and the estimated inflow amount excluding the influence of rainfall; The actual sewage treatment volume of sewage collected in each of the plurality of treatment divisions into which the treatment area is divided is calculated, and the treatment division average rainfall that fell in each treatment division is averaged according to the actual sewage treatment volume of each treatment division to calculate the sewage treatment volume corrected average rainfall for each treatment division, and these sewage treatment volume corrected average rainfalls are added together to calculate the treatment division average rainfall for the entire treatment area, An increased inflow prediction model is constructed to reproduce the rainfall-affected increased inflow as a rainfall-affected predicted increased inflow from the slope-average rainfall in the treatment area, next, The slope-average rainfall amount for the treatment area is calculated from the rainfall amount input as meteorological information, and the rainfall-influenced predicted inflow amount is calculated from the slope-average rainfall amount for the treatment area using the increased inflow amount prediction model. The amount of sewage inflow into the water treatment facility is predicted by adding the estimated inflow excluding the influence of rainfall to the obtained predicted increase in inflow influenced by rainfall. A method for predicting inflow volume.

2. The inflow prediction method according to claim 1, An inflow prediction method characterized in that, when the actual sewage inflow into the water treatment facility when there is no influence of rainfall increases or decreases over time compared to the previously obtained estimated inflow excluding the influence of rainfall, the estimated inflow excluding the influence of rainfall is corrected to increase or decrease in accordance with the increase or decrease without re-obtaining the estimated inflow excluding the influence of rainfall.

3. The inflow prediction method according to claim 1, An inflow prediction method characterized by the fact that, when the actual rainfall-affected increased inflow increases or decreases over time compared to the rainfall-affected predicted increased inflow obtained using the increased inflow prediction model, the rainfall-affected predicted increased inflow obtained using the increased inflow prediction model is corrected to increase or decrease in accordance with the increase or decrease without changing the increased inflow prediction model.

4. The inflow prediction method according to claim 1, An inflow prediction method characterized by the fact that, when the actual rainfall-influenced increased inflow changes due to temporal fluctuations in the arrival time of rainfall at the water treatment facility compared to the rainfall-influenced predicted increased inflow calculated using the increased inflow prediction model, the rainfall-influenced predicted increased inflow calculated using the increased inflow prediction model is corrected to match the fluctuations in the arrival time without changing the increased inflow prediction model.

5. An inflow prediction device for predicting the amount of sewage inflow collected from a treatment area to a water treatment facility, means for calculating an estimated inflow amount excluding the influence of rainfall from the sewage inflow amount to the water treatment facility when there is no influence of rainfall in the treatment area; means for calculating an increased inflow amount influenced by rainfall from the difference between the sewage inflow amount into the water treatment facility during rainfall and the estimated inflow amount excluding the influence of rainfall; a means for calculating the actual sewage treatment volume of sewage collected in each of a plurality of treatment sections obtained by dividing the treatment area, calculating a sewage treatment volume-corrected average rainfall for each treatment section by averaging the treatment section average rainfall that fell in each treatment section according to the actual sewage treatment volume for each treatment section, and then calculating the treatment section average rainfall for the entire treatment area by adding up each of these sewage treatment volume-corrected average rainfalls; A means for constructing an increased inflow prediction model that reproduces the rainfall-affected increased inflow as a rainfall-affected predicted increased inflow from the slope-average rainfall in the treatment area; means for calculating a treatment area slope average rainfall from rainfall input as meteorological information, and calculating a rainfall-influenced predicted inflow amount from the treatment area slope average rainfall using the increased inflow amount prediction model; a means for predicting the amount of sewage inflow into the water treatment facility by adding the estimated inflow excluding the influence of rainfall to the calculated predicted increase in inflow influenced by rainfall; An inflow prediction device comprising:

Images