JP6290072B2 - Unknown water generation area estimation device, unknown water generation area estimation method, and computer program - Google Patents

Description

  Embodiments described herein relate generally to an unknown water generation area estimation device, an unknown water generation area estimation method, and a computer program.

  The sewer is classified into two types of sewers, a combined sewer and a split sewer. The combined sewer system is a sewer system that allows sewage and rainwater to flow into a sewage treatment plant or the like through the same pipe. On the other hand, the diversion sewer system is a sewer system in which sewage and rainwater are allowed to flow down to a sewage treatment plant or the like through different pipes. In recent years, diversion sewerage has become mainstream.

  Originally, it is assumed that only sewage flows into the sewage pipes in the diversion sewerage system. And groundwater may flow in. As a result, more than expected amount of sewage flows into the sewage pipe. Sewage that exceeds this estimated amount is called unknown water. Of this unknown water, unknown water generated due to rain is called intrusion water in rainy weather, and unknown water generated by groundwater or the like is always called intrusion water. Since unknown water affects the quality of sewage, it may affect the operation of sewage treatment at sewage treatment plants. And the influence on the operation of sewage treatment may cause an increase in cost required for sewage treatment. In order to suppress such adverse effects of unknown water, it is necessary to carry out work to repair pipes and correct misconnections in areas where unknown water is generated (hereinafter referred to as “unknown water generation areas”). is there.

  However, it takes a lot of labor and cost to conduct a survey in a wide area to detect damaged pipes and misconnections. Therefore, narrowing the range of the area to be surveyed by some method leads to reduction in labor and cost required for the survey. With this background, techniques for estimating an unknown water generation area have been proposed. For example, as one of methods for estimating an unknown water generation area where intrusion water is generated in rainy weather, a method of performing multiple regression analysis based on the correlation between the amount of rainfall and the amount of unknown water is known.

  However, estimation of the unknown water generation area by the above method requires a lot of sample data for multiple regression analysis. For this reason, it takes time to accumulate sample data, and it may be impossible to estimate the area where unknown water is generated.

Japanese Patent No. 3857670 JP 2011-80347 A

  The problem to be solved by the present invention is to provide an unknown water generation area estimation device, an unknown water generation area estimation method, and a computer program that can estimate an unknown water generation area with less sample data.

  The unknown water generation area estimation device of the embodiment includes an acquisition unit and an estimation unit. The acquisition unit acquires rainfall data indicating rainfall in a small area constituting the target area, and flow data indicating the amount of water flowing out from a predetermined location in the jurisdiction. The estimation unit integrates the rainfall data acquired by the acquisition unit into a plurality of first unit areas wider than the small area, and performs multivariate analysis based on the flow data and the rainfall data for each first unit area. To estimate the first unit area including the unknown water generation area and perform multivariate analysis based on the flow rate data and the rainfall data for each second unit area included in the estimated first unit area. Thus, the second unit area including the unknown water generation area is estimated.

Schematic which shows the example of application of the estimation apparatus 1 of embodiment. The functional block diagram which shows the function structure of the estimation apparatus 1 of embodiment. The figure which shows the specific example of the rainfall table 161. FIG. The figure which shows the specific example of the flow volume table 162. FIG. The figure which shows the specific example of the sample data table 163. FIG. The flowchart which shows the flow of the process which the estimation apparatus 1 of embodiment estimates an unknown water generation | occurrence | production area. The figure which shows the specific example of the 1st unit area set to an object area. The figure which shows the specific example of the 1st unit area set to an object area. The figure which shows the specific example of the method of producing | generating sample data. The figure which shows the specific example of the 2nd unit area set to a 1st unit area. The figure which shows the modification of the 1st unit area set to an object area.

  Hereinafter, an unknown water generation area estimation device, an unknown water generation area estimation method, and a computer program according to embodiments will be described with reference to the drawings.

FIG. 1 is a schematic diagram illustrating an application example of the estimation apparatus 1 according to the embodiment.
The estimation apparatus 1 (unknown water generation area estimation apparatus) is an apparatus that estimates an unknown water generation area from a plurality of areas. The sewage treatment plant 2 treats sewage generated in a jurisdiction area (hereinafter referred to as “jurisdiction area”). In the example of FIG. 1, the area indicated by reference numeral 3 is the jurisdiction area of the sewage treatment plant 2.

  The rainfall radar 7 measures the amount of rainfall per unit time in a rectangular area of 250 [m] × 250 [m] (hereinafter referred to as “small area”), for example. In FIG. 1, the small area 4 is shown with a broken line 31 as a boundary line. In FIG. 1, reference numeral 8 represents a rain cloud. A dotted line group indicated by reference numeral 81 represents rain falling from the rain cloud 8. The rain radar 7 emits radio waves and measures the amount of rainfall based on the reflected wave reflected by the rain 81. The rainfall radar 7 generates rainfall data indicating the rainfall in each small area from the measured rainfall and transmits it to the estimation device 1.

  The flow meter 5 measures the amount of inflow per unit time of sewage that flows out from the jurisdiction and flows into the sewage treatment plant 2. The flow meter 5 is installed near the connection point (predetermined location) between the sewage treatment plant 2 and the sewage pipe 6-1. The sewage generated in each small area flows down the sewage pipe 6-2 laid in each small area and merges in the sewage pipe 6-1. The joined sewage flows down the sewage pipe 6-1 and flows into the sewage treatment plant 2. The flow meter 5 generates flow rate data indicating the measured inflow amount and transmits it to the estimation device 1.

  First, the estimation device 1 generates sample data used for multivariate analysis such as multiple regression analysis based on the acquired rainfall data and flow data. The estimation device 1 estimates an unknown water generation area from among a plurality of small areas 4 included in the jurisdiction area 3 by performing multivariate analysis using the generated sample data.

Here, an outline of estimation of an unknown water generation area when multiple regression analysis is used as a multivariate analysis technique will be described.
As described above, unknown water due to rainwater intrusion is an event caused by rainfall. Therefore, it is considered that there is a correlation between the amount of unknown water generated and the amount of rainfall. However, since the properties such as the infiltration rate of rainwater and the flow time differ depending on the soil, the time until the rain becomes unclear water and becomes apparent varies depending on the soil. Therefore, the correlation between the rainfall amount and the unknown water amount when viewed in a short period has a strong nonlinearity.

  However, when viewed over a long period of time, between the total amount of unknown water generated within that period (hereinafter referred to as “total unknown water amount”) and the total amount of rainfall (hereinafter referred to as “total rainfall”). Have a certain degree of linearity. Therefore, if the rainfall in a period that is long enough for the correlation between the total unknown water amount and the total rainfall to be linear is considered as one event, the correlation occurs between the rainfall and the unknown water amount. The total amount of unknown water can be estimated. Rain that is captured as one event in this way is called a rain event.

Based on such a relationship between the total unknown water amount and the total rainfall, the unknown water generation region can be estimated from a plurality of small regions by multiple regression analysis. Specifically, a multiple regression analysis is performed with the total rainfall for each small area as an independent variable, and an estimation formula that represents the total unknown water volume as the total rainfall for each small area is determined. In the estimation formula determined in this way, the coefficient of each independent variable can be considered to represent the degree of influence of the total rainfall amount for each area on the total unknown water amount. Therefore, the small area corresponding to the independent variable having a large coefficient can be estimated as the unknown water generation area. By performing such multiple regression analysis, for example, an estimation formula such as the following formula 1 is determined.

In Equation 1, Q represents the total amount of unknown water in the jurisdiction. N is an integer of 1 or more, and represents the number of small areas included in an area to be subjected to multiple regression analysis (hereinafter referred to as “target area”). The target area consists of some or all of the sub-areas included in the jurisdiction. For example, when it is known that an unknown water generation area exists in a certain area, the area in that area may be the target area. r _n (n = 1,..., N) represents the total rainfall in the n-th small area in the target area. a _n is a coefficient of the total rainfall r _n which is determined by multiple regression analysis. a ₀ is the intercept of the regression line determined by multiple regression analysis. ε is a regression error in the multiple regression analysis.

Here, since the coefficient a _n , the intercept a ₀ and the regression error ε need to be obtained in order to determine the estimation formula represented by the formula 1, at least n + 2 pieces of sample data necessary for the multiple regression analysis are required. It becomes. Furthermore, in order to determine an estimation formula that represents the total amount of unknown water with sufficient accuracy, two to three times as many sample data are required. Therefore, depending on the number n of small areas included in the target area, a lot of sample data may be required, and it takes time to collect the sample data.

  Therefore, the estimation device 1 according to the embodiment sets a plurality of unit areas by integrating small areas included in the target area. The estimation device 1 performs a multiple regression analysis using the total rainfall in each unit area as an independent variable, and estimates a unit area including an unknown water generation area from a plurality of unit areas. The estimation device 1 reduces the size of the unit area estimated to include the unknown water generation area by repeatedly setting the unit area and performing multiple regression analysis using the estimated unit area as a new target area. To go. As a result, it becomes possible to narrow the unknown water generation area to a specific range. The range for narrowing the unknown water generation area is arbitrary. For example, if the above repeated processing is performed until the estimated unit area becomes the unit of the small area, the unknown water is finally generated in one small area. It can also be estimated as an area.

  Specifically, the estimating apparatus 1 first sets a plurality of first unit areas by integrating the small areas in the target area. And the estimation apparatus 1 performs the multiple regression analysis by making the total rainfall in each 1st unit area into an independent variable, and estimates the 1st unit area containing an unknown water generation | occurrence | production area from several 1st unit areas.

  And the estimation apparatus 1 integrates the small area contained in the estimated 1st unit area, and sets a 2nd unit area in the estimated 1st unit area. The second unit area may be an area that coincides with the small area, or may be an area wider than the small area. The estimation device 1 performs a multiple regression analysis using the total rainfall in each second unit area as an independent variable, and estimates a second unit area including an unknown water generation area from a plurality of second unit areas.

  As described above, the estimation device 1 according to the embodiment reduces the size of the unit area estimated to include the unknown water generation area by repeatedly executing the setting of the unit area and the multiple regression analysis twice or more. I will do it. And the estimation apparatus 1 estimates the unit area finally obtained as an unknown water generation area. The estimation apparatus 1 according to the embodiment can estimate an unknown water generation area with less sample data by performing the above-described processing. Hereinafter, the configuration of the estimation apparatus 1 of the embodiment will be described in detail.

FIG. 2 is a functional block diagram illustrating a functional configuration of the estimation apparatus 1 according to the embodiment.
The estimation device 1 includes a CPU (Central Processing Unit), a memory, an auxiliary storage device, a communication interface, and the like connected by a bus. The estimation apparatus 1 includes a communication unit 11 as a communication interface. The communication unit 11 is configured using a communication interface such as a LAN (Local Area Network). The estimation device 1 communicates with the flow meter 5 and the rainfall radar 7 via the communication unit 11. The estimation device 1 includes a storage unit 16 as an auxiliary storage device. The storage unit 16 is configured using a storage device such as a magnetic hard disk device or a semiconductor storage device. The storage unit 16 stores an estimation program in advance. The CPU of the estimation device 1 reads the estimation program stored in the storage unit 16 into the memory and executes it. The estimation device 1 functions as a device including a data acquisition unit 12, a zone setting unit 13, a data generation unit 14, and an estimation unit 15 by executing an estimation program.

  All or some of the functions of the estimation device 1 may be realized by using hardware such as an application specific integrated circuit (ASIC), a programmable logic device (PLD), or a field programmable gate array (FPGA). The estimation program may be recorded on a computer-readable recording medium. The computer-readable recording medium is, for example, a portable medium such as a flexible disk, a magneto-optical disk, a ROM, a CD-ROM, or a storage device such as a hard disk built in the computer system. The estimation program may be transmitted via a telecommunication line.

  The data acquisition unit 12 (acquisition unit) acquires rainfall data indicating the rainfall amount for each small area and flow rate data indicating the inflow amount of sewage flowing out from a predetermined location in the jurisdiction into the sewage treatment plant. The data acquisition unit 12 acquires rainfall data from the rainfall radar 7. The data acquisition unit 12 acquires flow rate data from the flow meter 5. The data acquisition unit 12 stores the acquired rainfall data and flow rate data in the storage unit 16.

The area setting unit 13 sets a plurality of unit areas in the target area. The unit area is an area that is a unit for obtaining the total rainfall that is an independent variable of the multiple regression analysis. The unit area is composed of one or more small areas. Specifically, the zone setting unit 13 sets a unit zone based on the target zone information and the setting information. The target area information is information indicating the target area. The setting information is information indicating what unit area is set in the target area.
For the unit area set as the target area, the area setting unit 13 generates configuration information indicating the configuration of the small areas included in each unit area. The area setting unit 13 outputs the generated configuration information to the data generation unit 14.

  The setting information may be information that can be dynamically changed, or may be static information that is registered in the estimation device 1 in advance. For example, when the setting information can be dynamically changed, the size of the unit area may be changed by the operator, or the size of the unit area may be changed according to the rainfall. Further, for example, when the setting information is static information, the unit area may be set in a stepwise manner in advance for the target area. In this case, the estimation device 1 may be configured by omitting the zone setting unit 13.

  The data generation unit 14 (generation unit) generates sample data used for multiple regression analysis based on the rainfall data, the flow rate data, and the configuration information. First, the data generation unit 14 determines a rain event based on rainfall data. The data generation unit 14 calculates the total rainfall in each unit area for each rain event period based on the rainfall data. Further, the data generation unit 14 calculates the total amount of unknown water in the jurisdiction for each period of the rain event based on the flow rate data. The data generation unit 14 generates the total rainfall amount and the total unknown water amount calculated for each period of the rain event as sample data of the rain event. The data generation unit 14 stores the generated sample data of each rain event in the storage unit 16.

  The estimation unit 15 performs a multiple regression analysis using the sample data generated by the data generation unit 14, and determines an estimation formula that represents the total unknown water amount in the jurisdiction area by the total rainfall in each unit area. The estimation unit 15 estimates a unit area including an unknown water generation area from among a plurality of unit areas included in the target area based on the determined estimation formula. For example, the estimation unit 15 repeatedly estimates the unit area using the estimated unit area as a new target area until the estimated unit area becomes a predetermined size.

  The storage unit 16 stores, for example, a rainfall table 161, a flow rate table 162, and a sample data table 163.

FIG. 3 is a diagram illustrating a specific example of the rainfall table 161.
The rainfall table 161 is a table that holds rainfall data. The rainfall data is acquired every unit time by the rainfall radar 7 and transmitted to the estimation device 1.
The rainfall table 161 has a rainfall record for each time information. The rainfall record has each value of time information and rainfall r _n (n = 1, 2,..., N). N is an integer of 1 or more, and represents the number of sub-areas in the jurisdiction. The time information represents the date and time when the rainfall radar 7 acquired the rainfall data. Rainfall r _n represents the amount of rainfall per unit time in the n-th subregion of the N subregions. The rainfall record is generated by the data acquisition unit 12 based on the rainfall data and registered in the rainfall table 161.

FIG. 4 is a diagram illustrating a specific example of the flow rate table 162.
The flow rate table 162 is a table that holds flow rate data. The flow rate data is acquired every unit time by the flow meter 5 and transmitted to the estimation device 1.
The flow rate table 162 has a flow rate record for each time information. The flow rate record has each value of time information and inflow amount q. The time information represents the date and time when the flow meter 5 acquires the flow data. The inflow q represents the inflow per unit time of sewage flowing into the sewage treatment plant. The flow rate record is generated by the data acquisition unit 12 based on the flow rate data and registered in the flow rate table 162.

FIG. 5 is a diagram illustrating a specific example of the sample data table 163.
The sample data table 163 is a table that holds sample data. The sample data table 163 has a sample data record for each data ID. The sample data record has values of data ID, period information, total rainfall R _m (m = 1, 2,..., M) and total unknown water quantity Q. M is an integer of 1 or more, and represents the number of unit areas included in the target area. The data ID is identification information of the sample data record. Since the sample data is generated for each rain event, the data ID is also identification information for the rain event. The period information represents the period of the rain event indicated by the data ID. The total rainfall R _m is the m-th each unit area Rm of the M unit areas included in the target area, represents the total rainfall which occurred in the period of rainfall events indicated by the period information. The total unknown water amount Q represents the total amount of unknown water that has flowed into the sewage treatment plant in the period of the rain event indicated by the period information. The period of the rainfall event is a period set so that the correlation between the total rainfall amount and the total unknown water amount has a certain degree of linearity. Therefore, the total unknown water amount Q is the total amount of unknown water that has flowed out of the jurisdiction during the rain event. The sample data record is generated by the data generation unit 14 based on the rainfall data, the flow rate data, and the configuration information, and is registered in the sample data table 163.

FIG. 6 is a flowchart illustrating a process flow in which the estimation apparatus 1 according to the embodiment estimates an unknown water generation area.
First, the data acquisition unit 12 acquires rainfall data and flow data from the flow meter 5 and the rainfall radar 7 (step S101). The data acquisition unit 12 generates a rainfall record based on the acquired rainfall data and registers it in the rainfall table 161. Further, the data acquisition unit 12 generates a flow rate record based on the acquired flow rate data and registers it in the flow rate table 162.

  The area setting unit 13 sets a plurality of unit areas in the target area based on the target area information and the setting information (step S102). For example, the area setting unit 13 sets a first unit area as described below based on the setting information.

7 and 8 are diagrams illustrating specific examples of the first unit area set as the target area.
FIG. 7 shows a specific example of the target area. In the case of the example in FIG. 7, the target area is represented as a rectangular area in which 16 vertical and horizontal small areas are arranged, and is configured by a total of 256 small areas r1 to r256. FIG. 8 shows an example of a unit area set as the target area of FIG. In the case of the example in FIG. 8, the first unit areas R1 to R16 are areas in which four sub areas in the vertical and horizontal directions are integrated in the target area in FIG. In this case, for example, the setting information indicates the configuration of the small areas included in the first unit areas R1 to R16. For example, the setting information indicates that the first unit area R1 is composed of small areas r1 to r4, r17 to r20, r33 to r36, and r49 to r52. The setting information may be set in advance in the estimation device 1 by the user, or may be input to the estimation device 1 by the user when the estimation process is executed. Note that the setting information may not be information that holds the correspondence between the unit area and the small area as described above. For example, the setting information may be a rule or algorithm for setting a unit area. In this case, the estimation device 1 may be configured to automatically set the unit area based on the setting information.

Returning to the description of FIG.
Next, the zone setting unit 13 generates configuration information indicating the configuration of the small zone in the unit zone set in the target zone (step S103). The area setting unit 13 outputs the generated configuration information to the data generation unit 14.

The data generation unit 14 determines a rain event based on the rainfall data (step S104). Specifically, the data generation unit 14 refers to the rainfall table 161 and determines whether or not each rainfall record is a non-rainfall record in the order of the time indicated by the time information. The non-rainfall records, the value of rainfall r _n of all subregions are rainfall records is less than the reference value is determined to no rain. The data generation unit 14 acquires a period in which no-rainfall records are continuous by determining each rainfall record. The data generation unit 14 determines a period having a predetermined length or more of the acquired periods as a no-rain period. Then, the data generation unit 14 determines a rain event by setting a period between the no-rain periods as a rain event period.

  Next, the data generation unit 14 generates sample data of the rainfall event determined in step S104 based on the rainfall data, the flow rate data, and the configuration information (step S105). Specifically, the data generation unit 14 generates sample data as follows.

FIG. 9 is a diagram illustrating a specific example of a method for generating sample data.
In FIG. 9, the vertical axis represents time. Code 100-11～100-14 shows rainfall data obtained at time _t 11 _{~t 14} respectively. Times t _{11 to} t ₁₄ are times included in the period of the rain event 1. Similarly, reference numerals 100-21 and 100-22 show the rainfall data obtained at time _{t 21} and _{t 22,} respectively. Times t ₂₁ and t ₂₂ are times included in the period of the rain event 2. Similarly, reference numeral 100-p1~100-p3 shows rainfall data obtained at time _t P1 _{~t P3} respectively. Times t _{P1 to} t _P3 are times included in the period of the rain event P (P is an integer of 1 or more). The rain event p (p = 1, 2,..., P) is the rain event determined in step S104. Hereinafter, for convenience of explanation, each rainfall data is described as rain data 100 unless otherwise specified.
Each rainfall data 100 has a value of rainfall r _n of each subregion. Data generating unit 14, for each rain event p, calculates the average value of rainfall r _n of subregions included in the target area, and total rainfall R _mp in each unit area Rm. For example, when the unit area is set as in the example of FIG. 8, the total rainfall R _m in each unit area Rm is calculated by the following equation 2.

Further, in FIG. _9, q 11 _{to q 14} is the inflow indicated by the obtained flow rate data at time _t 11 _{~t 14} respectively. Similarly, q ₂₁ and q ₂₂ are the inflow amounts indicated by the flow rate data acquired at times t ₂₁ and t ₂₂ , respectively. _Similarly, q P1 _{to q P3} is the inflow indicated by the obtained flow rate data at time _t P1 _{~t P3} respectively. Hereinafter, for convenience of explanation, these inflows are described as inflows q unless otherwise distinguished. Data generating unit 14 calculates the sum of the inflow q each rainfall event p, calculates the total inflow Q _'p in rainfall event p. The total inflow Q ′ _p includes the total amount of normal sewage generated by domestic wastewater and the like and the total amount of unknown water. It is thought that the total amount of normal sewage is generated at a certain amount every day regardless of whether there is rainfall. Therefore, when the total amount of normal sewage is Q _s and the total unknown water amount of each rainfall event is Q _p , the total inflow amount Q ′ _p can be expressed as the sum of Q _s and Q _p . Then, if the total unknown water amount Q _p = 0 in fine weather is considered, the total amount Q _s of normal sewage becomes the total inflow amount Q ′ _{p in} fine weather. Therefore, the total amount Q _s of normal sewage can be calculated from the flow rate data in fine weather. Data generating unit 14, by the total flow rate Q _'p subtracting the total amount Q _s of the normal sewage, it calculates the total unknown amount of water Q _p of each rain event _p. Data generating unit 14, a total rainfall R _mp and total unknown amount of water Q _p calculated, to generate a sample data rainfall event p. The data generation unit 14 registers the generated sample data in the sample data table 163.
In general, it is known that the total amount of normal sewage involves seasonal fluctuations. Therefore, data for the same season may be used for calculating the total amount Q _s of ordinary sewage.

Returning to the description of FIG.
The estimation unit 15 refers to the sample data table 163 and determines whether or not the number of sample data necessary for the multiple regression analysis has been accumulated (step S106). Specifically, the estimation unit 15 performs the above determination using the number of sample records registered in the sample data table 163 as the number of sample data. When the number of sample data necessary for the multiple regression analysis is not accumulated (step S106-NO), the process returns to step S101, and the estimation device 1 acquires the next rainfall data and flow data.

On the other hand, when the number of sample data necessary for the multiple regression analysis is accumulated (step S106-YES), the estimation unit 15 performs the multiple regression analysis using the accumulated sample data, and determines the total unknown water amount Q in the jurisdiction. and determining an estimated equation representing total rainfall R _m of each unit area within the target area (step S107). For example, when the unit area is set as in the example of the first unit area in FIG. 8, the estimation unit 15 performs the multiple regression analysis to calculate the total unknown water amount Q in the jurisdiction area as in the following Expression 3. Determine the estimation formula to represent.

In Formula _3, R 1 _{to R 16} represents the total rainfall in each unit area. a _{1 to} a ₁₆ are coefficients of total rainfall R _{1 to} R ₁₆ determined by multiple regression analysis, respectively. a ₀ is the intercept of the regression line determined by multiple regression analysis. ε is a regression error in the multiple regression analysis. In order to determine the estimation formula represented by Expression 2, it is necessary to determine the coefficients a _{1 to} a ₁₆ , the intercept a _0, and the regression error ε. Therefore, the sample data necessary for the multiple regression analysis is , At least 18. Based on the determined coefficient of the estimation formula, the estimation unit 15 estimates a unit area including an unknown water generation area from each unit area (step S108).

  Next, the estimating unit 15 determines whether or not the estimated unit area is a small area (step S109). When the estimated unit area is a small area (step S109-YES), the estimation unit 15 ends the process with the estimated unit area as an unknown water generation area. On the other hand, when the estimated unit area is not a small area (step S109-NO), the estimation unit 15 updates the target area information as information indicating the estimated unit area (step S110). The estimating apparatus 1 repeatedly executes the processes in steps S102 to S108 until the unit area estimated in step S108 has a predetermined size.

  For example, when the first unit area R3 is estimated from the first unit area set as in the example of FIG. 8 in step S108, the estimation unit 15 uses the target area information as information indicating the first unit area R3. Update. The area setting unit 13 sets a new second unit area composed of small areas included in the unit area R3 in the first unit area R3 indicated by the target area information based on the setting information. For example, when the setting information indicates that each small area included in the first unit area R3 is a new second unit area, the area setting unit 13 sets a new second unit area as follows. .

FIG. 10 is a diagram illustrating a specific example of the second unit area set in the first unit area.
The area setting unit 13 sets the second unit areas R1 to R16 with the small areas r9 to 12, r25 to 28, r41 to 44, and r57 to 60 included in the first unit area R3 as new second unit areas. To do. Then, the zone setting unit 13 generates configuration information indicating the second unit zone as illustrated in FIG. 10 and outputs the configuration information to the data generation unit 14.

  The data generation unit 14 performs the same process as step S105, and generates sample data used for multiple regression analysis for the second unit area R3 indicated by the target area information.

The estimation unit 15 performs multiple regression analysis using the sample data generated by the data generation unit 14, and similarly to step S107, calculates the total unknown water amount Q of the jurisdiction area in each of the second unit areas R1 to R1 in the target area. The estimation formula represented by the total rainfall R ′ _{1 to} R ′ ₁₆ of R16 is determined. The estimation formula determined here is expressed as the following formula 4, for example.

In Equation 4, R ′ _{1 to} R ′ ₁₆ represent the total rainfall of the newly set second unit areas R1 to R16, respectively. Where each second unit area, because it is set in units of sub-regions, the total rainfall R _'1 ~R' ₁₆ of the second unit section R1~R16 has a total rainfall of each subregion Become. Specifically, the total rainfall R ′ _{1 to} R ′ ₁₆ is the total rainfall of the small areas r9 to 12, r25 to 28, r41 to 44, and r57 to 60, respectively. b _{1 to} b ₁₆ are coefficients of total rainfall R ′ _{1 to} R ′ ₁₆ determined by multiple regression analysis. b ₀ is the intercept of the regression line determined by multiple regression analysis. ε is a regression error in the multiple regression analysis. In order to determine the estimation formula represented by Expression 4, it is necessary to determine the coefficients b _{1 to} b ₁₆ , the intercept b _0, and the regression error ε. Therefore, the sample data necessary for the multiple regression analysis is , At least 18. The estimation unit 15 estimates a unit area including an unknown water generation area from each unit area based on the coefficient of the determined estimation formula. That is, in this case, the estimation unit 15 estimates one small area as an unknown water generation area from each small area in the second unit area R3.

  The estimation unit 15 performs the process of step S109 again, determines that the estimated second unit area is a small area, and ends the process.

  The estimation apparatus 1 of the embodiment configured as described above first integrates the small areas included in the target area, and sets a plurality of first unit areas in the target area. And the estimation apparatus 1 performs the multiple regression analysis which makes the total rainfall of each 1st unit area the independent variable about the object area, and the 1st unit area including an unknown water generation area from each 1st unit area Is estimated. Furthermore, the estimation device 1 sets a second unit area within the estimated first unit area. The estimation device 1 performs a multiple regression analysis using the total rainfall of each second unit area as an independent variable, and estimates a unit area including an unknown water generation area from each second unit area. Thus, the estimation apparatus 1 of the embodiment can estimate the unknown water generation area with a smaller number of sample data by repeatedly setting the unit area in the target area and performing multiple regression analysis. .

  For example, when the total rainfall in the unit area set as shown in FIG. 8 is used as an independent variable, 18 sample data are required for the multiple regression analysis. Further, when the total rainfall in the unit area set as shown in FIG. 9 is used as an independent variable, 18 pieces of sample data are required as in FIG. That is, the multiple regression analysis requires the number of unit areas set as the target area + 2 sample data. Therefore, when the unknown water generation area is estimated using the multiple regression analysis for the target area composed of 256 small areas as in the example of FIG. 8, the conventional estimation method requires at least 258 sample data. On the other hand, the estimation apparatus 1 of the embodiment can estimate the unknown water generation area using at least 18 pieces of sample data.

  Hereinafter, modified examples of the estimation device 1 of the embodiment will be described.

  The unit area set as the target area is not limited to the aspect of FIG. For example, the unit area may be set as shown in FIG.

  In order to estimate the unknown water generation area with a smaller number of sample data, it is desirable that the number of unit areas set in the target area is constant. For example, when one small area is estimated by two multiple regression analyses, the number of unit areas set for the first multiple regression analysis is x, and the unit is set for the second multiple regression analysis. Let y be the number of zones. In this case, the number of necessary sample data is the larger value of x or y, and the number of small areas in the target area is expressed as x × y = R constants. The larger value of x or y in the set of x and y satisfying such a relationship is the minimum when x = y. This relationship is considered to be the same when three or more multiple regression analyzes are performed. Therefore, it is desirable that the number of unit areas set in the target area is constant.

Moreover, the estimation apparatus 1 may estimate the unknown water amount which generate | occur | produces in each unit area by determining the estimation formula showing the total unknown water amount. The unit of the total unknown water quantity Q in Formula 1 is m ³ . The unit of rainfall r _n in each subregion is in mm. Therefore, the unit of the coefficient _{a n} ^is m 3 / mm. That, a _n in Formula 1 is considered unknown amount of water generated by 1mm rainfall in n-th subregion. Based on the same idea, the estimation device 1 can estimate the unknown water amount in an area where the possibility of generating unknown water is high by performing the multivariate analysis by applying the stepwise method. In the stepwise method, it is possible to determine an estimation expression represented only by an independent variable having a high correlation. Therefore, the coefficient in the estimation formula determined by the stepwise method indicates an area where there is a high possibility of generating unknown water. The estimation device 1 can estimate the coefficient value determined by the stepwise method as the unknown water amount in the unknown water generation area.

  According to at least one embodiment described above, the unknown water generation area estimation device is configured to provide rainfall data indicating the amount of rainfall in a small area constituting the target area, and a flow rate indicating the amount of water flowing out from a predetermined location in the jurisdiction. The data acquisition unit (data acquisition unit 12), and the rainfall data acquired by the acquisition unit are integrated into a plurality of first unit areas wider than the small area in the target area, and the flow rate data and the first By conducting multivariate analysis based on rainfall data integrated for each unit area, the first unit area including the unknown water generation area is estimated from the plurality of first unit areas, and the flow data and the unknown water generation area An unknown water generation area from a plurality of second unit areas by performing multivariate analysis based on rainfall data for each of a plurality of second unit areas included in the first unit area estimated to contain Estimation unit that estimates a second unit area containing a (zone setting unit 13 and the estimation unit 15), by having a can estimate the unknown water generation area with less sample data.

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

DESCRIPTION OF SYMBOLS 1 ... Estimation apparatus, 11 ... Communication part, 12 ... Data acquisition part, 13 ... Area setting part, 14 ... Data generation part, 15 ... Estimation part, 16 ... Memory | storage part, 161 ... Rainfall table, 162 ... Flow rate table, 163 ... Sample data table, 2 ... sewage treatment plant, 3 ... jurisdiction, 4 ... sub-region, 5 ... flow meter, 6-1, 6-2 ... sewage pipe, 7 ... rain radar, 8 ... rain cloud, 81 ... rain

Claims (7)

An acquisition unit for acquiring rainfall data indicating rainfall in a small area constituting the target area, and flow data indicating the amount of water flowing out from a predetermined location in the jurisdiction;
By integrating the rainfall data acquired by the acquisition unit into a first unit area wider than the small area, and performing multivariate analysis based on the flow rate data and the rainfall data for each first unit area, it is unknown Unknown water generation by estimating the first unit area including the water generation area and performing multivariate analysis based on the flow rate data and the rainfall data for each second unit area included in the estimated first unit area An estimation unit for estimating a second unit area including the area;
An unknown water generation area estimation device. The estimation unit performs multivariate analysis on the same number of the first unit area and the second unit area.
The unknown water generation | occurrence | production area estimation apparatus of Claim 1. A generator for generating sample data used for the multivariate analysis based on the rainfall data and the flow data;
The generation unit determines a plurality of rainfall events based on the rainfall data, with the rainfall that has occurred in a period that is long enough for the correlation between the total rainfall and the total unknown water in the target area to be linear. The total rainfall amount for each of the first unit area or the second unit area generated by one rain event and the total unknown water amount of the target area generated during the rain event are Generating sample data of the plurality of rainfall events as sample data;
The unknown water generation | occurrence | production area estimation apparatus of Claim 1 or Claim 2. The estimation unit performs the multivariate analysis by multiple regression analysis, determines an estimation formula that represents the total unknown water amount of the target area by the total rainfall of the first unit area or the second unit area, and the estimation formula Estimating the first unit area or the second unit area including the unknown water generation area based on the coefficient of
The unknown water generation | occurrence | production area estimation apparatus as described in any one of Claims 1-3. The estimation unit performs the multivariate analysis using a stepwise method, and has a high correlation with the total unknown water amount in the total rainfall of the first unit area or the second unit area in the target area. Determining an estimation formula representing the total unknown water amount by rainfall,
The unknown water generation | occurrence | production area estimation apparatus of Claim 4. An acquisition step of acquiring rainfall data indicating rainfall in a small area constituting the target area, and flow data indicating the amount of water flowing out from a predetermined location in the jurisdiction;
By integrating the rainfall data acquired in the acquisition step into a first unit area wider than the small area, and performing multivariate analysis based on the flow data and the rainfall data for each first unit area, it is unknown Unknown water generation by estimating the first unit area including the water generation area and performing multivariate analysis based on the flow rate data and the rainfall data for each second unit area included in the estimated first unit area An estimation step for estimating a second unit area including the area;
An unknown water generation area estimation method having an unknown water generation area. An acquisition step of acquiring rainfall data indicating rainfall in a small area constituting the target area, and flow data indicating the amount of water flowing out from a predetermined location in the jurisdiction;
By integrating the rainfall data acquired in the acquisition step into a first unit area wider than the small area, and performing multivariate analysis based on the flow data and the rainfall data for each first unit area, it is unknown Unknown water generation by estimating the first unit area including the water generation area and performing multivariate analysis based on the flow rate data and the rainfall data for each second unit area included in the estimated first unit area An estimation step for estimating a second unit area including the area;
A computer program for causing a computer to execute.

Images