CN101387127B - Rainwater drain control system and method - Google Patents

Abstract

The invention discloses a rainwater drain support system and a method which can reduce the number of variables input to an inflow forecast model of an inflow forecast section of the system, can make such forecasting at the inflow forecast section easier, and can forecast the quality of inflow rainwater. The rainwater drain support system is provided with a rainfall measurement section 10 for measuring rainfalls at a plurality of places and a linear mapping section 41 for making the linear mapping of time data matrix for rainfalls measured. An inflow forecast section 42 forecasts the inflow of rainwater by means of the linear mapping data. The rainwater drain support system forecasts the quality of inflow water by a system identification method using a nonlinear Hemmerstein model. The rainwater drain control system is provided with a detecting means to detect water level at a prescribed point upstream the pump facilities where water flows in, and an alteration means to change at least one of the preset rainwater pump start and stop water levels in accordance with water running conditions.

Description

Rain drainage control system and control method

The application be the applicant on December 14th, 2005 submit to, application number for " 200510131677.X ", denomination of invention divides an application for the application for a patent for invention of " rain drainage supporting system and support method, rain drainage control system and control method ".

Technical field

The present invention relates to according to present amount of precipitation of the seasonal effect in time series in a plurality of places or amount of precipitation prediction inflow pump facility (pump field) or the water quality of the inflow water of object facility such as inflow pump facility is used when the rain drainage supporting system used of the influx of the inflow water of object facilities such as water treatment plant or rain-water drainage support method, prediction rainfall down rain drainage supporting system or rain-water drainage support method and rain drainage control system or the rain-water drainage control method of the rainwater pump of object facilities such as pump facility being controlled usefulness in the future.

Background technology

The corresponding technology of the influx of the inflow water that is used to predict inflow pump facility object facilities such as (pump fields) at first, is described.

Flow into the method for the rainwater influx of object facilities such as sewer pump facility and following water treatment facilities as prediction, known for example as consideration based on the pave the way RRL method (refer to Patent Document 1: Japan's patent disclosure is put down 6-No. 322808 communiques) of method of physical model of state and sewer pipe any intermediate position of soil, 2000-No. 257140 communiques of Japan's patent disclosure) and adopt method (to refer to Patent Document 3: 2000-No. 56835 communiques of Japan's patent disclosure) etc. or utilize neural network method (to refer to Patent Document 2: based on the method for black-box model towards block models.In the method based on black-box model, utilize inputoutput data in the past to set up the influx forecast model in advance, this model is used as input variable with the amount of precipitation of Using Radar and Rain Gauge Data and the measurement of a plurality of ground rainfall gauge, simultaneously the rainwater influx is used as output variable, and according to present amount of precipitation or predict in the future amount of precipitation, use this influx forecast model, prediction rainwater influx.

Here, the amount of precipitation of Using Radar and Rain Gauge Data is measured the rainfall intensity in the intensive seizure measuring object of energy zone, thereby compares with the udometric amount of precipitation measurement in a plurality of ground, can obtain careful amount of precipitation information.By utilizing amount of precipitation that Using Radar and Rain Gauge Data measures like this or, can high accuracy carrying out the rainwater influx prediction of object facility according to predicting the input variable of amount of precipitation the future of this amount of precipitation prediction when the influx forecast models such as above-mentioned RRL method, neutral net method of opposing.

The corresponding technology of the quantity of inflow water of object facility of inflow pump facility when being used to predict rainfall and water quality etc. then, is described.

In the sanitation of convergent current type, when rainy, rainwater flows in the sewer pipe usually in the sewer pipe setting area.At the sewer pipe terminal part pump facility etc. is set, the inflow water that will comprise rainwater is discharged into the regulation drainage position.

This sanitation must be handled when the rainwater pump sump directly is discharged into the rivers and creeks, makes river pollution few.That is, need and flow into water quality and the water storage at place and discharging water such as suitably control in the pipe duct respectively rivers and creeks etc. according to the rainwater influx.During rainfall, produce the foul outflow that waits accumulations in sewer pipe and flow into inflow the first of canal at its at initial stage and gush.This first gushing is also needed to catch its generation and suitably processing.

In the past, proposed this sanitation (is for example referred to Patent Document 4: 2004-No. 249200 communiques of Japan's patent disclosure) according to amount of precipitation prediction water quality and the invention of controlling foul composition removal device operational ton.

The following describes and be used for corresponding technology that the rainwater pump of object facilities such as pump facility is controlled.

Control method as the rainwater apparatus for controlling pump, known method has: preestablish opening water level and turn-offing water level of rainwater pump, and reach moment of the water level of these settings at the measured value of the fluviograph of the rainwater pump sump that is located at the rainwater that savings flow down from sewer pipe, make the rainwater pump open or turn-off.

Flow into the rainwater pump sump water level rising that causes in order to suppress rapid rainwater, for example (patent documentation 5: that sees 2000-No. 328642 communiques of Japan's patent disclosure) is such at patent documentation 5, the PREDICTIVE CONTROL device is provided, can predict the rainwater influx of pump facility according to the measured value of ground rainfall gauge and Using Radar and Rain Gauge Data, calculate the required rainwater pump running quantity of draining, real time modifying rainwater pump open, turn-off water level, thereby need append when opening the rainwater pump being judged as, open the rainwater pump from being lower than the water level that preestablishes water level.

Again, the pump facility of convergent current type sewer is for the water pollution in the rivers and creeks that suppresses discharging place, and existent method is by when light rain or the prediction rainwater influx of pump facility when little, and that improves the rainwater pump opens, turn-offs water level, suppresses the discharging of rainwater pump.

The problem of above-mentioned corresponding technology of the influx of the inflow water that is used to predict inflow pump facility object facilities such as (pump fields) at first, is described.In the influx prediction based on black-box model such as above-mentioned neutral net method, use least squares method to differentiate under the situation of coefficient parameter according to inputoutput data when setting up the influx forecast model, when having high dependency relation between the input variable, be difficult to discrimination parameter, have the problem that is difficult to set up high accuracy influx forecast model.There is in the object facility influx prediction problem that is not easy to select suitable parameter, is not easy to study the foundation of the influx forecast model that adapts to this selection in the input variable number of influx forecast model for a long time.When carrying out the amount of precipitation measurement with Using Radar and Rain Gauge Data, the input variable number leap that has high dependency relation mutually increases, thereby significantly the problems referred to above takes place.

Secondly, the problem of above-mentioned corresponding technology of water quality etc. of the inflow water of object facility such as inflow pump facility when being used to predict rainfall is described.In the above-mentioned invention such as patent documentation 4 grades,, there is the numerous and diverse problem of formula owing to work out the inflow water quality prediction method according to the water quality model formula.

Then, the problem that is used for corresponding technology that the rainwater pump of object facilities such as pump facility is controlled is described.The rain that drops on the ground flows into sewer pipe from the face of land, and flows down in sewer pipe and arrive the pump facility.This process is complicated and huge.Therefore, the amount of precipitation measured value is as in the influx forecast model of importing, pump sump forecast model of water level, the rainwater pump running quantitative forecast model, a lot of as the predicted value of its output, produce error sometimes.For example, under the little situation of actual inflow ratio prediction, the rainwater pump of the too much platform number of running quantity, and actual inflow ratio is predicted and water logging danger is taken place when big.

The present invention considers the problems referred to above and finishes, and its purpose is to provide a kind of number that can reduce the influx forecast model input variable of influx prediction section, can makes the convenient and high rain drainage supporting system and the rain-water drainage support method of precision of rainwater influx prediction of this influx prediction section.

Another object of the present invention is to provide a kind of can be convenient and predict the inflow water water quality prediction method of the water quality of the inflow water that rainfall brings exactly, also provide a kind of influx can be with the water quality of this prediction or this water quality and rainfall the time suitably to control to flow into accumulating and discharge of water simultaneously with the rain drainage supporting system and the rain-water drainage support method of minimizing to the influence of environment.

Another purpose of the present invention is to provide a kind of and opens, turn-offs water level so that sharply also can carry out in the apparatus for controlling pump of suitable draining during the change of rainwater influx revising the rainwater pump, make amendment by except that the predicted value of using influx and water level, going back the high index of while dependability, make the rain drainage control system and the rain-water drainage control method of drainage system steady operation.

Summary of the invention

In order to solve above-mentioned problem, the rain drainage supporting system of the present invention of the 1st aspect, the amount of precipitation measurement section that comprises the amount of precipitation of measuring a plurality of places, the amount of precipitation in a plurality of places of measuring according to this amount of precipitation measurement section, obtain the amount of precipitation time series data matrix of each place time series amount of precipitation of expression, go forward side by side and be about to the Linear Mapping that is transformed into less variable data than multivariate data of amount of precipitation time series data matrix and handle, to obtain the Linear Mapping portion of Linear Mapping data matrix, and influx prediction section, this influx prediction section utilizes the Linear Mapping data matrix that Linear Mapping portion is obtained as the influx forecast model of importing, prediction flows into the rainwater influx of object facility, this Linear Mapping portion carries out Linear Mapping by the performance matrix with the inherent vector unit that comprises variance matrix in the amount of precipitation time series data matrix in the past or covariance matrix, obtains the Linear Mapping data matrix.

The rain-water drainage support method of the prediction inflow pump of the present invention field of the 7th aspect and the rainwater influx of following water treatment plant, comprise following steps: the amount of precipitation measurement section is measured the step of the amount of precipitation in a plurality of places, Linear Mapping portion is according to the amount of precipitation in a plurality of places of this amount of precipitation measurement section measurement, obtain the amount of precipitation time series data matrix of each place time series amount of precipitation of expression, go forward side by side and be about to the Linear Mapping that is transformed into less variable data than multivariate data of amount of precipitation time series data matrix and handle, to obtain the step of Linear Mapping data matrix, and the influx prediction section is utilized Linear Mapping data that described Linear Mapping portion the is obtained influx forecast model as input, prediction flows into the step of the rainwater influx of object facility, in the step that obtains the Linear Mapping data matrix, described Linear Mapping portion carries out Linear Mapping by the performance matrix with the inherent vector unit that comprises variance matrix in the amount of precipitation time series data matrix in the past or covariance matrix, obtains the Linear Mapping data matrix.

According to this rain drainage supporting system and rain-water drainage support method, in the Linear Mapping portion, amount of precipitation time series data matrix as k * n matrix is carried out Linear Mapping, be transformed into (the Linear Mapping data matrix of m＜n) as k * m, so can reduce the number of the influx forecast model input variable of influx prediction section, thus the rainwater influx prediction that can make this influx prediction section easily and precision high.

The rain drainage supporting system of the present invention of the 2nd aspect, the amount of precipitation prediction section that comprises to time series the amount of precipitation in prediction a plurality of places in the future, amount of precipitation according to a plurality of places in future of this amount of precipitation prediction section prediction, obtain the amount of precipitation time series data matrix of each place time series amount of precipitation of expression, go forward side by side and be about to the Linear Mapping that is transformed into less variable data than multivariate data of amount of precipitation time series data matrix and handle, to obtain the Linear Mapping portion of Linear Mapping data matrix, and influx prediction section, this influx prediction section is utilized Linear Mapping data matrix that described Linear Mapping portion the is obtained influx forecast model as input, prediction flows into the rainwater influx of object facility, this Linear Mapping portion carries out Linear Mapping by the performance matrix with the inherent vector unit that comprises variance matrix in the amount of precipitation time series data matrix in the past or covariance matrix, obtains the Linear Mapping data matrix.

The rain-water drainage support method of the prediction inflow pump of the present invention field of the 8th aspect and the rainwater influx of following water treatment plant, comprise following steps: the step of the amount of precipitation in amount of precipitation prediction section prediction a plurality of places in the future, Linear Mapping portion is according to the amount of precipitation in a plurality of places in the future of this amount of precipitation prediction section prediction, obtain the amount of precipitation time series data matrix of each place time series amount of precipitation of expression, go forward side by side and be about to the Linear Mapping that is transformed into less variable data than multivariate data of amount of precipitation time series data matrix and handle, to obtain the step of Linear Mapping data matrix, and the influx prediction section is utilized Linear Mapping data that described Linear Mapping portion the is obtained influx forecast model as input, prediction flows into the step of object facility, in the step that obtains the Linear Mapping data matrix, described Linear Mapping portion carries out Linear Mapping by the performance matrix with the inherent vector unit that comprises variance matrix in the amount of precipitation time series data matrix in the past or covariance matrix, obtains the Linear Mapping data matrix.

According to this rain drainage supporting system and rain-water drainage support method, in the Linear Mapping portion, amount of precipitation time series data matrix as k * n matrix is carried out Linear Mapping, be transformed into (the Linear Mapping data matrix of m＜n) as k * m, so can reduce the number of the influx forecast model input variable of influx prediction section, thus the rainwater influx prediction that can make this influx prediction section easily and precision high.

In above-mentioned rain drainage supporting system and the rain-water drainage support method, preferably Linear Mapping portion carries out Linear Mapping by the performance matrix with the inherent vector unit that comprises variance matrix in the amount of precipitation time series data matrix in the past or covariance matrix, obtains the Linear Mapping data matrix.

According to this rain drainage supporting system and rain-water drainage support method, the performance matrix combination that is used for Linear Mapping is the variance matrix of amount of precipitation time series data matrix or the inherent vector unit of covariance matrix in the past, thereby can make the unit of the Linear Mapping data matrix that Linear Mapping obtains become mutual incoherent linear independent time sequence data.

In above-mentioned rain drainage supporting system and the rain-water drainage support method, preferably according to the accumulation contribution rate of calculating with the eigenvalue of this inherent vector unit in the inherent vector that constitutes described performance matrix greater than predetermined threshold.

According to this rain drainage supporting system and rain-water drainage support method, set up the performance matrix, make according to and the accumulation contribution rate calculated of the eigenvalue of inherent vector unit greater than predetermined threshold, thereby the mode that the information that can have with the amount of precipitation time series data matrix before the Linear Mapping is not lost as far as possible obtains the Linear Mapping data matrix.

In above-mentioned rain drainage supporting system and the rain-water drainage support method, preferably Linear Mapping portion carries out Linear Mapping by the performance matrix of forming with the load matrix that obtains in the master unit process of analyzing past amount of precipitation time series data matrix, obtains the Linear Mapping data matrix.

According to this rain drainage supporting system and rain-water drainage support method, the mode of not losing as far as possible except the information that can have with the amount of precipitation time series data matrix before the Linear Mapping obtains the Linear Mapping data matrix, can also make the unit of the Linear Mapping data matrix of acquisition become mutual incoherent linear tolerance time series data.

In above-mentioned rain drainage supporting system and the rain-water drainage support method, preferably also comprise the past amount of precipitation time series data matrix and the time in the past sequence rainwater influx that are transformed into few variable data according to multivariate data, set up the model evaluation portion of influx forecast model amount of precipitation time series data matrix.

According to this rain drainage supporting system and rain-water drainage support method, the number of past amount of precipitation time series data that is input to model evaluation portion is few, can carry out the foundation of influx forecast model easily.And, set up the influx forecast model owing to independently pass by the amount of precipitation time series data according to mutual incoherent linearity, can obtain the high influx forecast model of precision.

The prediction of the present invention of the 13rd aspect is from the rain-water drainage support method of the water quality of the water of sewer pipe inflow pump field and following water treatment plant, comprise following steps: in accordance with regulations the step of the amount of precipitation of the setting area of the described sewer pipe of period measurement, in accordance with regulations period measurement inflow pump field or down the water quality of the inflow water of water treatment plant step and according to the power value of present water quality, some water quality in the past, some amount of precipitations in the past and some amount of precipitations in the past by using system's discrimination method of Nonlinear Hammerstein model, the step of prediction water quality in the future.

The rain drainage supporting system of the pump of the present invention field of the 14th aspect or following water treatment plant, comprise the amount of precipitation measurement mechanism of measuring the sewer pipe setting area, measure the water quality measurement device of the water quality of the inflow water that flows into described pump field or following water treatment plant, periodically collecting and store the amount of precipitation of described amount of precipitation measurement mechanism and water quality measurement device periodic measurement and the data of inflow water quality respectively collects and storage device, according to the present water quality in the data of this data collection and memory device stores, some water quality in the past, the power value of some amount of precipitations in the past and some amount of precipitations in the past is by using system's discrimination method of Nonlinear Hammerstein model, the prediction unit of prediction water quality in the future, and the inflow water quality of predicting with this prediction unit produces relevant sewage pump well, the drive support device of the running of accumulating and the discharging instruction of the inflow water of rainwater pump sump or rainwater plash.

In the above-mentioned rain drainage supporting system, best described prediction unit prediction water quality in the future, simultaneously also by using system's discrimination method of Nonlinear Hammerstein model, according to the present influx in the memory device stores data, some influxs in the past, some amount of precipitations in the past and some power value prediction rainwater influxs of amount of precipitation in the past; The drive support device is not only with the running of accumulating and the discharging instruction that flows into water quality but also produce the inflow water of relevant sewage pump well, rainwater pump sump or rainwater plash with the rainwater influx of prediction unit prediction.

In the above-mentioned rain drainage supporting system, be preferably in the water quality measurement device is set in the sewer pipe, data are collected with storage device and are also periodically collected and be stored in the water quality of measuring in the sewer pipe; Prediction unit is by the system's present sewer pipe of discrimination method basis that uses the Nonlinear Hammerstein model interior water quality and the power value prediction inflow pump field of the water quality in some sewer pipes in the past, some past amount of precipitations and some past amount of precipitations or the water quality in the sewer pipe before the following water treatment plant; The drive support device also produces the running of accumulating and the discharging instruction of the inflow water of relevant sewage pump well, rainwater pump sump or rainwater plash with the water quality in the sewer pipe before inflow pump field of predicting or the following water treatment plant.

In the above-mentioned rain drainage supporting system, be preferably in flow measurement device is set in the sewer pipe, data are collected with storage device and are also periodically collected and be stored in the flow of measuring in the sewer pipe; Prediction unit is by the system's present sewer pipe of discrimination method basis that uses the Nonlinear Hammerstein model interior flow and the power value prediction inflow pump field of the flow in some sewer pipes in the past, some past amount of precipitations and some past amount of precipitations or the flow in the sewer pipe before the following water treatment plant; The drive support device also produces the running of accumulating and the discharging instruction of the inflow water of relevant sewage pump well, rainwater pump sump or rainwater plash with the flow in the sewer pipe before inflow pump field of predicting or the following water treatment plant.

In the above-mentioned rain drainage supporting system, be preferably in water level measuring amount device in the pipe is set in the sewer pipe, data are collected and the water level of measurement in the sewer pipe is also periodically collected and store to storage device, and can obtain and store the weather information from Meteorological Information System; Prediction unit obtains fine water level in sewer pipe from storage device, calculating the height of sand and deposit in the sewer pipe, and according to the flow in the rainy day sewer pipe and water quality prediction inflow pump field or the flow and the water quality of water treatment plant down.

In the above-mentioned rain drainage supporting system, preferably the drive support device produces the running instruction, in the scope that is no more than the amount of precipitation upper limit threshold when the water-quality ratio predetermined threshold value difference of the inflow water of prediction, make the foul water pump running that is located at sewage pump well, and the rainwater pump that is located at the rainwater pump sump does not turn round.

In the above-mentioned rain drainage supporting system, preferably the drive support device is differentiated first gushing according to the influx of the inflow water of prediction, and generation running instruction, make this water that just gushes accumulate in the rainwater plash, and amount of precipitation is during less than predetermined value, and the pump that is located at the rainwater plash makes the described water that accumulates return sewage pump well.

The rain drainage control system of the present invention of the 21st aspect, flow decision rainwater pump running quantity according to rainwater inflow pump field, and change control rainwater pump according to pump field middle water level and turn on and off, comprise the checkout gear of the water level of flow point on the regulation that detects water inflow pump facility and predetermined water level and the predetermined modifier that turn-offs at least one side in the water level opened of revising the rainwater pump according to the situation that current get off.

In the above-mentioned rain drainage supporting system, when the water level of best described regulation point surpassed predetermined threshold, described modifier was revised the predetermined at least one side who opens in water level and the predetermined shutoff water level of rainwater pump.

In the above-mentioned rain drainage supporting system, when the water level of best described regulation point was lower than predetermined threshold, described modifier was revised the predetermined at least one side who opens in water level and the predetermined shutoff water level of rainwater pump.

In the above-mentioned rain drainage supporting system, preferably the SEA LEVEL VARIATION speed of the water level of described regulation point and described regulation point all is used as the index of the water level of the described regulation point of expression.

In the above-mentioned rain drainage supporting system, the required rainwater pump running quantity Calculation device of water that also comprises the prediction unit of the rainwater influx of predicting described pump facility and calculate the rainwater influx of discharging prediction.

The rain-water drainage control method of the present invention of the 26th aspect comprises following steps: the step that turns on and off according to the step of the prediction rainwater flow of inflow pump facility decision rainwater pump running quantity, according to the SEA LEVEL VARIATION control rainwater pump of pump facility and will revise the step that water level and shutoff water level stagger opening of rainwater pump in time opened of rainwater pump near the water level in the flow point place sewer pipe on the regulation of pump field as index.

The rain-water drainage support method of the present invention of the 27th aspect, prediction is to the pump field or the following influx of the inflow water of water treatment plant of sewer pipe, comprise following steps: the step of the amount of precipitation of the setting area of the described sewer pipe of period measurement in accordance with regulations, period measurement flows into the described pump field or the following step of the influx of the inflow water of water treatment plant in accordance with regulations, and with the present influx in these measured values, some influxs in the past, some amount of precipitations in the past and some power values of amount of precipitation in the past, by the system's discrimination method that uses non-linear Hammerstien model, prediction flows into the step of the influx of water in the future.

The rain drainage control system of the present invention of the 28th aspect, rainwater influx with the prediction of the rain-water drainage support method described in the present invention of the 7th aspect, determine the quantity of the rainwater pump of running according to the rainwater influx that flow into the pump facility, and control turning on and off of described rainwater pump according to the SEA LEVEL VARIATION of described pump facility, wherein comprise and detect the checkout gear of water outlet toward the water level in the place, regulation upper reaches that described facility flows into, revising predetermined rainwater pump according to the situation under the current opens water level and turn-offs the modifier of at least one side in the water level, and the device of the required rainwater pump running quantity of the water of rainwater influx of calculate discharging prediction.

According to rain drainage supporting system of the present invention and rain-water drainage support method, in the Linear Mapping portion, amount of precipitation time series data matrix or prediction amount of precipitation time series data matrix in the future as k * n matrix are carried out Linear Mapping, be transformed into k * m (m＜n), information is not lost as far as possible, and form mutual incoherent variable, so can reduce the input variable number of influx forecast model in the influx prediction section, thus the rainwater influx prediction that can make this influx prediction section easily and precision high.

Another rain drainage supporting system and rain-water drainage support method according to the present invention utilize system's discrimination method (non-linear Hammerstien model) to go into water quality according to predicted flows such as flowing into water quality measurement value, past amount of precipitation, thereby can carry out this prediction easily.Control flows into accumulating of water and discharges owing to can utilize these to predict the outcome suitably, and its utilization can reduce the influence to environment.

According to rain drainage control system of the present invention and rain-water drainage control method, on pump facility influx base of prediction, revise opening water level and turn-offing water level of rainwater pump with situation under the current, thereby can make the drainage system steady operation.

Description of drawings

Fig. 1 is the key diagram of composition that the drainage system against rain of embodiment of the present invention is shown.

Fig. 2 is the block diagram that the control system main assembly of the drainage system against rain of key diagram 1 is used.

Fig. 3 is the block diagram that the composition of influx projecting body in the control system of Fig. 2 is shown.

Fig. 4 is that the performance matrix that illustrates in the influx projecting body of Fig. 3 is set up the flow chart of the effect of portion.

Fig. 5 illustrates the block diagram that flows into the composition of water quality prediction mechanism in the control system of Fig. 2.

Fig. 6 is the block diagram of composition that the control system moderate rain water pump controlling organization of Fig. 2 is shown.

Fig. 7 is the flow chart of effect that the rainwater pump controlling organization of Fig. 6 is shown.

Fig. 8 illustrates to revise the ideograph that the rainwater pump turns on and off the situation of water level in the rainwater pump controlling organization of Fig. 6.

Fig. 9 illustrates to make many rainwater pumps turn on and off the ideograph of the situation of water level displacement in the rainwater pump controlling organization of Fig. 6.

Figure 10 is the table of expression water level rising index in the rainwater pump controlling organization of Fig. 6.

Figure 11 is the figure that in the rainwater pump controlling organization of Fig. 6 the water level time series data and the dirty water level time series data of sewer pipe at sewer pipe upper reaches is compared.

Figure 12 is the block diagram that another composition of influx projecting body in the drainage system against rain of the present invention is shown.

Figure 13 is the block diagram that another composition of rainwater pump controlling organization in the drainage system against rain of the present invention is shown.

Label declaration

The 10th, amount of precipitation measurement section, the 11st, Using Radar and Rain Gauge Data, the 12nd, ground rainfall gauge, the 15th, the amount of precipitation prediction section, the 20th, sewer pipe, the 21st, flow into canal, 23, the 27th, fluviograph, 24, the 25th, water quality analyser, the 26th, flow meter, the 30th, sand trap rainwater pump sump, the 31st, sewage pump well, the 32nd, the rainwater plash, the 34th, rivers and creeks, the 35th, rainwater pump, the 36th, foul water pump, the 37th, rainwater plash pump, the 38th, fluviograph, 40,40p is the influx projecting body, the 41st, and Linear Mapping portion, the 42nd, the influx prediction section, the 43rd, the performance matrix is set up portion, and the 44th, model evaluation portion, 50,50p flows into water quality prediction mechanism, the 52nd, transacter, 52a is the Using Radar and Rain Gauge Data data, and 52b is ground rainfall gauge data, and 52c is a waterlevel data, 52d flows into water quality data, 52e is a pump information, and 52f is a discharge rate, and 52p is a transacter, the 53rd, data storage device, the 54th, prediction unit, the 55th, arithmetic unit, the 56th, drive support device, the 57th, control device, the 58th, synoptic model prediction unit, the 60th, rainwater pump controlling organization, the 62nd, operating number calculation element, the 63rd, open and turn-off water level modifier, the 64th, fluviograph selecting arrangement.

The specific embodiment

Below, with reference to the description of drawings embodiment of the present invention.The embodiment that the present invention not the following describes limits, and comprises the various embodiments that contain technological thought of the present invention.

Fig. 1 is the key diagram of composition that the drainage system against rain of embodiment of the present invention is shown, and Fig. 2 is the block diagram that the main assembly of control system of the drainage system against rain of key diagram 1 is used.

As shown in Figure 1, the drainage system against rain of present embodiment has in the setting area to sewer pipe 20 grades a plurality of places and carries out the measuring amount of precipitation measurement section 10 of amount of precipitation.The dirty side of sewer pipe 20, the pump facility that comprises inflow canal 21 and rainwater pump sump 30 is set.

Amount of precipitation measurement section 10 comprises for example Using Radar and Rain Gauge Data 11 or a plurality of ground rainfall gauge 12.When amount of precipitation measurement section 10 is made up of Using Radar and Rain Gauge Data 11, utilizes the straightline propagation of electric wave and run into the just characteristic of reflection of rain, a radar base station is measured the amount of precipitation in n zone in batch.As shown in Figure 1, being divided into amount of precipitation measuring object zone for example by observation grid, n limit is that roughly hundreds of rice is measured the amount of precipitation in n zone of this division respectively to the square of several kms.

Here, when amount of precipitation measurement section 10 is made up of Using Radar and Rain Gauge Data 11, compare when forming, can divide amount of precipitation measuring object zone, can increase the quantity n in rainfall gauge measuring object zone with finer observation grid by a plurality of ground rainfall gauge 12.

Sewer pipe 20 is made up of for example convergent current type pipeline, and the rainwater that lands in its setting area is past dirty together with daily life sewage in this zone and industrial wastewater.Dirty side at this sewer pipe 20 is provided with inflow pump facility (pump field), in this pump facility, is provided with and flows into canal 21.This inflow canal 21 temporarily accumulates the rainwater that sewer pipe 20 is sent here.As shown in figure 11, near the sewer pipe 20 the inflow canal 21 water quality analyser 25, flow meter 26 and fluviograph 27 are set respectively.

In the pump facility, be provided with sand trap rainwater pump sump 30 the dirty side who flows into canal 21.Sewage pump well 31 and rainwater plash 32 also are set in the pump facility.Inflow water from sewer pipe 20 flows into these sand trap rainwater pump sumps 30, sewage pump well 31 and rainwater plash 32 by flowing into canal 21 and sand trap (not shown).Usually between sand trap rainwater pump sump 30 and inflow canal 21 overflow weir is arranged, inflow water at ordinary times flow into sewage pump well 31.Otherwise, comprising during rainfall that a large amount of inflow current of rainwater are fashionable, the water level that flows into canal 21 raises, and inflow water overflows the high overflow weir of water level and flows to sand trap rainwater pump sump 30.Rainwater plash 32 accumulates the inflow water of pump field, and by be located at and sewage pump well 31 between the overflow weir (not shown) and/or the lock that is located at and flows between the canal 21 admit to flow into water.

At sand trap rainwater pump sump 30 rainwater pump 35 is set, the inflow water (rainwater) that utilizes rainwater pump 35 will flow into sand trap rainwater pump sump 30 is discharged into places such as rivers and creeks 34.Be provided with foul water pump 36 at sewage pump well 31, utilize foul water pump 36 will flow into inflow water (sewage) that sewage pump well 31 accumulates and deliver to unshowned water treatment facilities down in addition after the water treatment, be discharged into rivers and creeks etc. and locate.At this moment, the sewage former state of the treating capacity of water treatment facilities is overflowed from sewage pump well 31 under surpassing, thereby is sent to 32 stops of rainwater plash.Be provided with rainwater plash pump 37 at rainwater plash 32, utilize described rainwater plash pump 37 to send the stop water that this rests on rainwater plash 32 back to sewage pump well 31 in the stage that the inflow water that flows into canal 21 fully reduces.

Promptly, will from the inflow water of sewer pipe 20 by being located at its terminal part inflow canal 21 and after unshowned sand trap rests on the either party in sand trap rainwater pump sump 30, sewage pump well 31 and the rainwater plash 32 or this 3 side stopped respectively, be discharged into the regulation drainage position.

At inflow canal 21 fluviograph 23 and water quality analyser 24 are set respectively.Fluviograph 23 is measured the water level that flows into the inflow water that accumulates in the canal 21.Water quality analyser 24 is measured the inflow water quality that flows into the inflow water in the canal 21.

As shown in Figure 1, be provided with the fluviograph 38 that the water level of measuring this sand trap rainwater pump sump 30 is used at sand trap rainwater pump sump 30.Can utilize in real time and be located at fluviograph 23 that flows into canal 21 and the measured value that is located at the fluviograph 38 of sand trap rainwater pump sump 30.

Then, with the control system in the block diagram illustrations drainage system against rain of Fig. 2.

As shown in Figure 2, drainage system against rain have the influx projecting body 40 used with the influx of the inflow water of the present amount of precipitation of the seasonal effect in time series in a plurality of places prediction inflow pump facility (pump field) the object facility of etc.ing, with the inflow water quality prediction mechanism 50 of the usefulness such as water quality of the inflow water of object facilities (for example flowing into canal 21) such as described amount of precipitation prediction inflow pump facility and with the rainwater pump controlling organization 60 of opening at least one side in water level and the shutoff water level of the influx adjustment rainwater pump 35 of the inflow water of predictions such as influx projecting body 40.

The detailed composition of influx projecting body 40 at first, is described with Fig. 3 and Fig. 4.

As shown in Figure 3, the amount of precipitation time series data matrix of influx projecting body 40 with a plurality of places that described amount of precipitation measurement section 10 is measured carries out the Linear Mapping portion 41 of Linear Mapping and the influx prediction section 42 that flows into the rainwater influx of subject area according to the Linear Mapping data prediction that Linear Mapping portion 41 obtains.The performance matrix that Linear Mapping portion 41 connects the performance matrix of setting up the Linear Mapping that is used for this Linear Mapping portion 41 is set up portion 43.Influx prediction section 42 connects the model evaluation portion 44 of the influx forecast model of setting up the rainwater influx prediction that is used for this influx prediction section 42.

Linear Mapping portion 41 is according to the amount of precipitation data in a plurality of places of amount of precipitation measurement section 10 measurements, obtain the amount of precipitation time series data matrix X of the time series amount of precipitation in each place of expression, also this amount of precipitation time series data matrix X is carried out Linear Mapping simultaneously, multivariate data is transformed into few variable data, to obtain Linear Mapping data matrix Y.

Particularly, Linear Mapping portion 41 carries out the computing of following formula (1).

Y=XP ... formula (1)

Wherein, X is the matrix of k (discrete time the point) * n (amount of precipitation detected object zone number) of amount of precipitation time series data, and P is the performance matrix of setting forth later n * m matrix of setting up portion 43 and setting up (performance matrix of m＜n).

Among the amount of precipitation time series data matrix X, the regional amount of precipitation of each of each time point t is corresponding to row vector x _t(t=1 ..., k).Here, up-to-date amount of precipitation data are defined as x ₁, the amount of precipitation data of q (1≤q≤k-1) before the stage are defined as x _Q+1

Obtain Linear Mapping data matrix Y by the computing of carrying out above-mentioned formula (1) in Linear Mapping portion 41 as k * m matrix.

The Linear Mapping data that influx prediction section 42 obtains Linear Mapping portion 41 are as input, the influx forecast model prediction rainwater influx of utilizing the model evaluation portion 44 that sets forth later to set up.In the rainwater influx prediction of this influx prediction section 42, as shown in Figure 3, also with reference to the influx time series data.

Here, the influx forecast model that the present invention uses is the black-box model by the relation decision of the amount of precipitation data of past measurement and the pairing influx data of this amount of precipitation, the influx forecast model that the utilization of influx prediction section produces in advance according to the relation of past amount of precipitation data and influx data, and Linear Mapping data of amount of precipitation in the past that obtain with reference to Linear Mapping portion 41 and influx time series data in the past, prediction rainwater influx in future.

The performance matrix is set up the principal component analysis that portion 43 utilizes as a kind of multi-variables analysis method, obtain according to past amount of precipitation time series data and to calculate the load matrix that this principal component is used, and the performance matrix P that utilizes its characteristic to migrate and use in the Linear Mapping of Linear Mapping portion 41 in the matrix of obtaining.Particularly, make up the variance of past amount of precipitation time series data, the inherent vector of covariance matrix, thereby set up the performance matrix P of n * m matrix.

Describe the details that this performance matrix is set up portion 43 in detail with Fig. 4.

The past amount of precipitation time series data in n place when at first, the step 11 of setting up portion 43 by the performance matrix is according to the amount of precipitation measuring object is calculated variance, covariance matrix.Use k _h* n matrix X _hExpression is the amount of precipitation time data in the past, and the unit that then constitutes performance matrix P forms X _hVariance, covariance matrix S _hN inherent vector p _j(j=1 ..., n) in p ₁..., p _m(m＜n) arrange as column vector.Wherein, to inherent vector p _j(j=1 ..., n) use corresponding eigenvalue λ _j(j=1 ..., n) satisfy λ ₁〉=λ ₂〉=λ _m〉=... 〉=λ _nRelation.

In this step 11, shown in following formula (2), the 1st principal component ~ m principal component of discrete time point i is expressed as k _hM row vector t of * m matrix T _i(i=1 ..., k).

X _h=TP '+E ... formula (2)

Wherein, E is because of (m＜n) principal component is carried out the error that truncation produces at m." ' " be the label of expression permutation matrix.

Then, as shown in Figure 4, calculate accumulation contribution rate a by step 12.This accumulation contribution rate a is that how many source matrix X expression principal component T has _hThe value of the information that comprises.Calculate this accumulation contribution rate a by following formula (3).

[several 1]

$a=\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{\mathrm{\λ}}_i/\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\λ}}_i$ Formula (3)

Then, select inherent vector, make accumulation contribution rate a greater than pre-set threshold by step 13.At last, arrange the inherent vector of selecting, thereby obtain the performance matrix P of n * m matrix by step 14.

Like this, the performance matrix is set up portion 43 and is just being pass by amount of precipitation time series data matrix X _hThe process of principal component analysis in obtain the performance matrix P that forms by load matrix.

Model evaluation portion 44 bases are carried out Linear Mapping and multivariate data are transformed into a plurality of amount of precipitation time series datas of past and a plurality of influx time series datas of past of few variable data, set up the influx forecast model.

The influx forecast model that the present invention uses is a black-box model, but in this model evaluation portion, the application system discrimination method carries out differentiating coefficient parameter etc. with least squares method etc., will carry out the input-output relationship modelization of the rainwater influx time series data of past amount of precipitation time series data after the Linear Mapping and object facility.As modeling method, the method for available Hammerstein model of for example setting forth later etc. based on neural network method and employing.

Then, with Fig. 5 the detailed composition that flows into water quality prediction mechanism 50 is described.As shown in Figure 5, flow into water quality prediction mechanism 50 and have transacter 52, data storage device 53, prediction unit 54, arithmetic unit 55, drive support device 56 and control device 57.In addition, also synoptic model prediction unit 58 can be set.

Transacter 52 periodically respectively collection obtain amount of precipitation (Using Radar and Rain Gauge Data data 52a and ground rainfall gauge data 52b) that above-mentioned amount of precipitation measurement section 10 measures, flow into the inflow water in a canal position 52c that the water yield uses and the pump information 52e of the operating condition that flows into water quality 52d and relevant above-mentioned each pump 35,36,37 and the discharge value 52a of each pump 35,36,37.Data storage device 53 is stored the various data that transacter 52 is periodically collected respectively.

The various data of described data storage device 53 storages of prediction unit 54 usefulness flow into water quality and flow into the water yield by the Nonlinear Hammerstein model prediction.The Nonlinear Hammerstein model simplification is represented then this model is made up of following modular form.That is, obtain inflow water quality, obtain the inflow water yield by formula (5) by following formula (4).

Predicted flows go into water quality=present situation (t) flow into water quality * α 1+ go over (t-1) flow into water quality * α 2+ go over (t-2) flow into water quality * α 3+

+ past (t-1) amount of precipitation * β 1+ goes over (t-2) amount of precipitation * β 2+

+ (past (t-1) amount of precipitation) ²* γ 1+ (past (t-2) amount of precipitation) ²* γ 2+

+ (past (t-1) amount of precipitation) ³* δ 1+ (past (t-2) amount of precipitation) ³* δ 2+

Formula (4)

Predicted flows go into the water yield=present situation (t) flow into the water yield * α 11+ go over (t-1) flow into the water yield * α 12+ go over (t-2) flow into the water yield * α 13+

+ past (t-1) amount of precipitation * β 11+ goes over (t-2) amount of precipitation * β 12+

+ (past (t-1) amount of precipitation) ²* γ 11+ (past (t-2) amount of precipitation) ²* γ 12+

+ (past (t-1) amount of precipitation) ³* δ 11+ (past (t-2) amount of precipitation) ³* δ 12+

Formula (5)

In above-mentioned formula (4), (5), t represents current point in time, t-1 represents the 1st time in the past, t-2 represents the 2nd time in the past, α 1, α 2, α 3 ..., α 11, α 12, α 13 ..., β 1, β 2 ..., β 11, β 12 ..., γ 1, γ 2 ..., γ 11, γ 12 ..., δ 1, δ 2 ..., δ 11, δ 12 ... be respectively coefficient, each equipment is according to sewer 20 and settings such as scale that flows into canal 21 grades and characteristic.

Promptly, flow into water quality and flow into the water yield by imitating with the regulation non-linear relation of past amount of precipitation, thereby flowing into water quality, some amount of precipitations in the past, some power values of amount of precipitations in the past with flowing into now water quality, some past, prediction flows into water quality in the future according to above-mentioned modular form (4).Equally, also flow into the water yield, some amount of precipitations in the past, some power values of amount of precipitations in the past, and obtained according to above-mentioned modular form (5) and flow into the water yield in the future with the present inflow water yield, some past.

The volume of traffic (operating number and rotary speed) of each pump 35,26,37 shown in arithmetic unit 55 is calculated with the inflow water yield according to the inflow water quality of prediction unit 54 predictions etc.Here, the volume of traffic that elaborates rainwater pump 35 in the explanation of the rainwater pump controlling organization 60 of setting forth is in the back calculated.Drive support device 56 instructs with shutoff with the running of accumulating and discharging that inflow water quality produces the inflow water of relevant sand trap rainwater pump sump 30, sewage pump well 31, rainwater plash 32 according to the inflow water yield of prediction unit 54 predictions.Control device 57 produces according to described running instruction the control of relevant device is exported.Synoptic models such as synoptic model prediction unit 58 common rainfall models of prediction and torrential rain model can utilize the synoptic model of prediction to revise the value that arithmetic unit 55 is obtained.

Below, the detailed composition of rainwater pump controlling organization 60 is described with Fig. 6.

Here, rainwater pump controlling organization 60 is meant programmable devices such as programmable logic controller (PLC) (PLC), work station, personal computer, microcomputer.Rainwater pump controlling organization is a kind of function of the computer system of multi-section computer composition sometimes.

As shown in Figure 6, rainwater pump controlling organization 60 has running quantity calculation element 62, opens, turn-offs water level modifier 63 and fluviograph selecting arrangement 64.

Number of calculations calculation element 62 is with rainwater influx predicted value Q, the pump sump water level H of described influx projecting body 41 predictions _p, pump discharge value Q _pIn at least 1 as input, the pump running quantity that rainwater influx predicted value Q calculate is needed.

Then, open, turn-off each upper reaches water level H that water level modifier 63 utilizes sewer pipe ₁..., H _nIn at least 1, the pump running quantity of prediction and pump sump water level H now _pWater level is opened, turn-offed to the pump that modification is predetermined.

Then, fluviograph selecting arrangement 64 is selected to the fluviograph that water level modification index was opened, turn-offed to pump according to the plane rainfall distribution in the object sewer zone of 10 measurements such as Using Radar and Rain Gauge Data 11 measurement section such as amount of precipitation such as grade.

Rainwater pump controlling organization 60 also sends to open, to turn-off the various control signals that (break-make) signal is representative, with the pump plant equipment of control prime mover and valve etc. rainwater pump 35.

The effect of above-mentioned such present embodiment of forming then, is described.

The effect of influx projecting body 40 at first, is described with Fig. 3 and Fig. 4.

As shown in Figure 1, at first measure the amount of precipitation in n the zone of pressing the observation grid division respectively in amount of precipitation measurement section 10.

On the other hand, as shown in Figure 3, utilize the performance matrix to set up the performance matrix P that portion 43 sets up Linear Mapping.Particularly, as shown in Figure 4, set up in the portion 43 at this performance matrix, at first calculate variance, covariance matrix according to the past amount of precipitation time series data in n the place in amount of precipitation measuring object zone, also calculate the inherent vector p of variance, covariance matrix simultaneously by step 11 _j(j=1 ..., n) and corresponding eigenvalue λ _j(j=1 ..., n).

Then, by step 12 according to eigenvalue λ _jAfter utilizing above-mentioned formula (3) to calculate accumulation contribution rate a, select eigenvalue λ by step 13 _j, make accumulation contribution rate a greater than pre-set threshold.At last, arrange the eigenvalue λ that selects by step 14 _jThereby, set up n * m matrix (performance matrix P of m＜n).

Then, Linear Mapping portion 41 obtains the amount of precipitation time series data matrix X that represents each place time series amount of precipitation according to the amount of precipitation data in present a plurality of places that amount of precipitation measurement section 10 is measured, also amount of precipitation time series data matrix X is carried out Linear Mapping simultaneously according to above-mentioned formula (1), multivariate data is transformed into few variable data, thereby obtains Linear Mapping data matrix Y.

Here, amount of precipitation time series data matrix X is k * n matrix (k: discrete time point, n: amount of precipitation measuring object zone number), the performance matrix P that performance square portion 43 sets up is n * m matrix (m＜n), thereby the Linear Mapping data matrix Y of gained is k * m matrix.

Like this, just can utilize Linear Mapping portion 41 to make the quantity of amount of precipitation time series data reduce to k * m from k * n.

Model evaluation portion 44 sets up the influx forecast model according to carrying out Linear Mapping and multivariate data being transformed into a plurality of amount of precipitation time series datas of past and a plurality of influx time series datas in the past after few variable data.

Then, the Linear Mapping data that influx prediction section 42 obtains Linear Mapping portion 41 are as input, the influx forecast model prediction rainwater influx of setting up with model evaluation portion 41.

Then, with Fig. 5 the effect that flows into water quality prediction mechanism 50 is described.

In the sanitation, first the gushing at general rainfall initial stage becomes problem.Just gush when occurring in initial rainfall, but do not know what mud that the sewer pipe duct piles up flow out constantly and arrive the pump field at.Therefore, flow into and to constitute the inflow water quality and the influx of inflow water that contains the mud of accumulation by prediction such as when the rainfall in the water quality prediction mechanism 50, also can do suitably processing just gushing.For example,, suppress water to be delivered to sanitation as far as possible, observe the draining rule, carry out the little processing of environmental loads toward the rivers and creeks draining to just gushing the rainwater that contains mud etc. by accumulate in the sewer pipe duct or in the rainwater plash as far as possible.

Here, the water quality prediction method that flows into the following current entry in the water quality prediction mechanism 50 is obtained inflow water quality according to amount of precipitation, with amount of precipitation or rainfall intensity as input, flow into water quality (coliform, COD, BOD, phosphorus, nitrogen etc.) as output, the system discrimination method of utilization provides and flows into the water quality prediction value.Promptly, according to above-mentioned modular form (4), flow into the water quality measurement value with present situation inflow water quality measurement value and some past and carry out recursive operation, also weigh recursive operation simultaneously with some measured values of rainfall in the past, and then add power (for example 2 powers, the 3 powers) value that some past flow into the water quality measurement value, thereby prediction forms the inflow water quality of non-linear relation to rainfall.

Like this owing to can predict the water quality of the inflow water that flows into canal 21 like this, can by drive support device 56 according to the inflow water quality convection current entry of prediction accumulate and discharging is suitably controlled.In the prediction unit 54, can flow into the inflow water yield of canal 21, thereby can go into water quality and carry out more suitable running and control by go into the water yield and predicted flows with this predicted flows according to above-mentioned modular form (5) prediction.Be elaborated below.

In the sanitation, when the rainfall of the setting area of sewer pipe 20, as indicated above, rainwater flows in the sewer pipe interior 20, accumulates in sand trap rainwater pump sump 30, sewage pump well 31, rainwater plash 32 from the inflow canal 21 that is located at its terminal part by unshowned sand trap.Sand trap rainwater pump sump 30 has the overflow weir that is higher than sewage pump well 31, thereby inflow water at ordinary times flows into sewage pump well 31, is discharged into down water treatment facilities by foul water pump 36, be subjected to down the water treatment facilities processing after, be discharged into places such as rivers and creeks 34.

During rainfall, flow into a large amount of inflow water that comprise rainwater, the water level that then flows into canal 21 raises, and flows into water and also flows into sand trap rainwater pump 30.The rainwater that accumulates in sand trap rainwater pump sump 30 is discharged into places such as rivers and creeks 34 by rainwater pump 35.Yet, first during rainfall such as gushes at the inflow water that causes flows out the accumulation mud of piling up in the sewer pipe 20, so this water flows into sand trap rainwater pump sump 30 and when being discharged into rivers and creeks 34 by rainwater pump 36, discharge the rainwater that contains the water quality inferiority of piling up mud, and is undesirable aspect environment.Therefore, contain the inflow water quality of just gushing, and prediction water influx down, thereby judge and accumulate or discharge, carry out the control of safety and consideration environment according to the rainfall condition predicting.

As indicated above, during rainfall, measure the amount of rainfall in the amount of precipitation measurement section 10 of Using Radar and Rain Gauge Data 11 and a plurality of ground rainfall gauge 12 etc., collect their data 52a, 52b with transacter 52, store data store 53 into.Equally, also collect waterlevel data 52c with acquisition inflow water in a canal position, and flow into water quality data 52d, store these data into data storage device 53 from water quality analyser 24 collections from being located at the fluviograph 23 that flows into canal 21.Cycle is carried out the data collection of these transacters 52 in accordance with regulations, and the data that each cycle is collected store data storage device 53 respectively into.

Prediction unit 54 utilizes above-mentioned non-linear Hammerstien modular form (4), (5) convection current to go into water quality and flows into the water yield and predict computing according to the amount of precipitation of data storage device 53 records and the data of inflow water quality.Non-linear Hammerstien model is to be conceived to amount of precipitation and to flow into water quality or the model of the non-linear relation of the inflow water yield, flows into water quality when amount of precipitation is increased and is used to flow into the water quality and the inflow water yield with the structure that the inflow water yield is directly proportional with the power (for example 2 powers, 3 powers) of amount of precipitation.

Drive support device 56 is according to the predicted value of the inflow water quality of prediction like this, the break-make of the break-make of the accumulating of the pipe duct that the accumulating, comprise of decision rainwater plash 32 flows into canal 21, rainwater pump 35 and running quantity, foul water pump 36 and running quantity etc.For example, during the threshold difference of the water-quality ratio regulation of the inflow water of prediction, in the scope that is no more than the amount of precipitation upper limit threshold, sending the running instruction, the rainwater pump 35 that is located at sand trap rainwater pump sump 30 is not turned round, and the foul water pump 36 that is located at sewage pump well 31 turns round.Be predicted as when flowing into the water quality variation, also can make its as ahead of time down the feedforward of the running of water treatment facilities (advanced processes, conventional actived sludge, oxidation channel process etc.) etc. operate.

Then, with Fig. 6 to Figure 11 rainwater pump controlling organization 60 is described.

Particularly, the execution in step of rainwater pump controlling organization 60 comprises the running content that flow chart shown in Figure 7 is recorded and narrated.

At first, step 21 is carried out the influx prediction by described influx projecting body 40 after, calculate the running quantity of rainwater pump 35 according to its result in step 22.About being used for the running quantity calculation element 62 of this step, in 2000-No. 328642 communiques of Japan's patent application etc. detailed content is arranged, but it is as follows to set forth its outline.

As input parameter, prediction is also calculated the suitable pump running quantity that a few minutes places (for example locating in 5 minutes) (for example located) to the dozens of minutes place in 30 minutes to running quantity calculation element 62 with the pump facility rainwater influx of influx projecting body 40 predictions and the level measuring value of sand trap rainwater pump 30 etc.Then, running quantity calculation element 62 is according to formula (6), and the influx predicted value Qr ' at the stipulated time point t+1 place of usefulness influx projecting body 40 outputs is the running quantity of calculating pump (t+1).

N’(t+1)＝Qr’(t+1)/R ……(6)

Wherein, N (t) is the running quantity (platform) of pump, and R is the rated value (m of pump ³/ s), Qr ' is a stipulated time point t+1 place influx predicted value (t+1).

Then, the modification of opening water level and turn-offing water level of description of step 23.Open water level and turn-off water level modifier 63 in the time need appending pump running quantity, revise the predetermined water level of opening of the predetermined pump of running, with ahead of time or postpone opening, turn-offing of pump the required running quantity of the water of discharging prediction influx.

Shown in the example of Fig. 8, to be judged as pump that needs open the predetermined water level H that opens according to running quantity with this pump _On, turn-off water level H _OffBe revised as H respectively _On-Δ H, H _Off-Δ H.

As the concrete value of Δ H, for example patent documentation 5 (2000-No. 328642 communiques of Japan's patent disclosure) is recorded and narrated like that, and existent method is the pump sump water level H with current point in time _pOpen water level H with predetermined pump _OnDifference as Δ H, i.e. Δ H=H _On-H _p

[several 2]

$\mathrm{\ΔH}=f(H_p,H_l,...,H_n,\frac{{\mathrm{\ΔH}}_p}{\mathrm{\Δt}},\frac{\mathrm{\Δ}{H}_l}{\mathrm{\Δt}},...,\frac{{\mathrm{\ΔH}}_n}{\mathrm{\Δt}},...)$ Formula (7)

The method that has is defined as it with pump sump water level, upper reaches sewer pipe water level and their the time rate of change mapping f as variable as top formula (7).

The method that has is shifted the break-make water level when the break-make water level of many pumps of decision as shown in Figure 9.

To open, turn-off the water level modifier different with the pump that patent documentation 5 is recorded and narrated, and present embodiment is characterised in that pump break-make water level is revised the water level that uses the sewer pipe upper reaches.

Return Fig. 7, judge in its step 24 whether the index that expression sewer pipe upper reaches water level raises surpasses threshold value, and step 25 only reduces the break-make water level of this pump when surpassing threshold value.

In the step 26, judge whether the index that expression sewer pipe upper reaches water level raises is lower than threshold value, and only when being lower than threshold value, the raise break-make water level of this pump of step 25.If sewer pipe upper reaches water level rising index is in bound, step 27 is just carried out the break-make of pump according to predefined water level.Then, in the step 28,, export rain-water drainage pump make-and-break signal from apparatus for controlling pump at the time point that pump sump water level Hp reaches the water level of setting.

Figure 10 illustrates about the level measuring value of the index that raises as the expression water level and the table of level rise velocity.Can be predetermined the relation of level measuring value and level rise velocity like this, be used to judge whether water level and level rise velocity surpass threshold value.

Figure 11 is the ideograph of the dirty waterlevel data sequence data of the water level time series data at sewer pipe upper reaches and pump field periphery.Here, time difference that upper reaches and dirty water level can be raise was used as in the delivery time of sewer pipe toward dirty rainwater.That is, in the moment that shifts to an earlier date the delivery time from the moment of observing sewer pipe upper reaches water level raise, dirty water level should raise really.

In sum, utilize the rain-water drainage support method of present embodiment, can carry out following item.At first, by with above-mentioned influx projecting body 40,41 couples of amount of precipitation time series data matrix X as k * n matrix carry out Linear Mapping in Linear Mapping portion, it is transformed into (the Linear Mapping data matrix Y of matrix of m＜n) as k * m, thereby can reduce the input variable number of the influx forecast model of influx prediction section 42, the influx prediction that can make this influx prediction section 42 is easily.

Linear Mapping portion 41 usefulness are rain time sequence data matrix X in the past _hVariance-covariance matrix S _hThe performance matrix P that forms of the inherent vector unit line linearity mapping of going forward side by side, obtain Linear Mapping data matrix Y, thereby to make the unit of the Linear Mapping data matrix Y that Linear Mapping obtains be mutual incoherent independent time sequence data.

Again, in the inherent vector that makes the unit that constitutes the performance matrix, the accumulation contribution rate a that calculates according to this inherent vector is greater than preset threshold, and the mode of the information that can have with the amount of precipitation time series data matrix X that does not lose before the Linear Mapping as far as possible obtains Linear Mapping data matrix Y.

Linear Mapping portion 41 usefulness are to past amount of precipitation time series data matrix X _hThe process analyzed of principal component in the performance matrix P that forms of the load matrix that obtains carry out Linear Mapping, obtain Linear Mapping data matrix Y, thereby the mode of the information that can have with the amount of precipitation time series data matrix X that does not lose before the Linear Mapping as far as possible obtains Linear Mapping data matrix Y, and can also make the unit of the Linear Mapping data matrix Y that obtains simultaneously is mutual incoherent linear independent time sequence data.

Model evaluation portion 44 bases are carried out Linear Mapping and are transformed into the past amount of precipitation time series data and the past influx time series data of few variable data from multivariate data, set up the influx forecast model, thereby the quantity of past amount of precipitation time series data that is input to model evaluation portion 44 is few, can carry out the foundation of influx forecast model easily, set up the influx forecast model owing to independently pass by the amount of precipitation time series data again, can obtain the high influx forecast model of precision according to mutual incoherent linearity.

Rain drainage supporting system and rain-water drainage support method according to present embodiment, flow into water quality prediction mechanism 50 by using, go into water quality and flow into the water yield according to the either party of Using Radar and Rain Gauge Data 11 and ground rainfall gauge 12 or both sides' amount of precipitation predicted flows, thereby can predict the generation of just gushing.Promptly, drive support device 56 can be differentiated first gushing according to the inflow water yield of prediction, and sends the running instruction, makes rainwater plash 32 accumulate the water that this just gushes, and become setting when following at amount of precipitation, utilize the pump 37 of rainwater plash 32 to make the water that accumulates return sewage pump well 31.For example, produce when just gushing, open and flow into the lock (not shown) of canal 21, make the water beginning of just gushing accumulate in rainwater plash 32 earlier to rainwater plash 32.Flowing into the time point that the water yield fully reduces, the time point that can not measure for example at the rainfall measurement of Using Radar and Rain Gauge Data 11 or ground rainfall gauge 12, the first water burst that can use rainwater plash pump 37 to stop is sent sewage pump well 31 back to from rainwater plash 32.Deliver to unshowned water treatment facilities down with foul water pump 36 from sewage pump well 31, be discharged into rivers and creeks 34 after the following water treatment (advanced processes, conventional actived sludge, oxidation channel process), can carry out the running little the environment harmful effect.

Rain drainage control system and rain-water drainage control method according to present embodiment, by using rainwater pump controlling organization 60, rainwater is down flow to the pump field needs the sewer pipe water level in place of time of a few minutes as index, revise the break-make water level of pump, for example make pump early open-minded, raise thereby can produce the time point inhibition sand trap rainwater pump sump water level that sharply flows into sand trap rainwater pump sump in reality.Otherwise or it is open-minded that the rainwater pump is postponed, thereby can reduce the rainwater pump displacement, and the water pollution in territory, intake area is reduced.

And, except that influx predicted value etc. might produce the index of mistake, also the sewer pipe upper reaches of actual measurement and the water level of sand trap rainwater pump sump are taken as the index of rainwater pump break-make, thereby reduce the danger of rainwater pump running instruction mistake.

Drainage system against rain of the present invention is not limited to aforesaid way, can add various conversion.

Utilize Figure 12 that the conversion example of drainage system against rain of the present invention is described below.Among Figure 12, to the embodiment identical part mark identical label extremely shown in Figure 11, detailed with Fig. 1.

As shown in figure 12, the drainage system against rain of this conversion example, its difference only is among the influx projecting body 40p, between amount of precipitation measurement section 10 and Linear Mapping portion 41 amount of precipitation prediction section 15 is set, and others are identical to embodiment shown in Figure 11 with Fig. 1 in fact.

Below, with Figure 12 this influx testing result 40p is described.

As shown in figure 12, amount of precipitation prediction section 15 connects amount of precipitation measurement section 10, sends n the regional time series amount of precipitation data that amount of precipitation measurement section 10 is measured.

And amount of precipitation prediction section 15 is according to the amount of precipitation time series data of this measurement, and prediction is each regional time series amount of precipitation in the future.

Then, the amount of precipitation data in a plurality of places in the future that amount of precipitation prediction section 15 is predicted are delivered to the Linear Mapping portion 41 that connects this amount of precipitation prediction section 15.Then, in this Linear Mapping portion 41, the matrix Z of in the future prediction amount of precipitation time series data is carried out Linear Mapping, multivariate data is transformed into few variable data, and obtains Linear Mapping data matrix Y.

In sum, according to this conversion example influx projecting body 40p, predict that amount of precipitation time series data matrix Z carries out Linear Mapping 41 pairs of future as k * n matrix in Linear Mapping portion, be transformed into (the Linear Mapping data matrix Y of m＜n) as k * m, thereby can reduce the input variable number of the influx forecast model of influx prediction section 42, the influx prediction that can make this influx prediction section 42 is easily.

Then, with Figure 13 another conversion example of the present invention is described.

As shown in figure 13, only in flowing into the water quality prediction 50p of mechanism, the composition of transacter 52p is different with transacter shown in Figure 5 52 for this conversion example drainage system against rain, and others are identical to embodiment shown in Figure 11 with Fig. 1 in fact.

Below, with Figure 13 this inflow water quality prediction 50p of mechanism is described.

The drainage system against rain of this conversion example, embodiment use shown in Figure 5 is located at 20 interior water quality analysers 25, flow meter 26, fluviograph 27 on the sewer, by transacter 52p in accordance with regulations the cycle collect their survey data 25a, 26a, 27a, store the data of this collection into data storage device 53.Transacter 52p and data storage device 53 are also obtained weather information from unshowned Meteorological Information System and are stored.

Prediction unit 54 is also predicted water quality, the flow in the sewer pipe 20 in the sewer pipe 20 respectively and flow into sand or the deposit quantity that flows into canal 21 in sewer pipe 20 except that the inflow water quality and influx of the above-mentioned inflow canal 21 of prediction.

Promptly, prediction unit 54 flow into the water quality on the set-point that flows into canal 21 preceding water quality analysers 25 in sewer pipe 20 according to above-mentioned modular form (4) with the water quality in the water quality in the present sewer pipe 20, the some past sewer pipes 20, some amount of precipitations in the past, some power value predictions of amount of precipitations in the past.

Prediction unit 54 flow into the flow on the set-point that flows into canal 21 preceding flow meters 26 in sewer pipe 20 according to above-mentioned modular form (5) with the flow in the flow in the present sewer pipe 20, the some past sewer pipes 20, some amount of precipitations in the past, some power value predictions of amount of precipitations in the past.

And then, prediction unit 54 is obtained fine water level waterpipe 20 in down, is obtained sand or deposit height in the sewer pipe 20 from data storage device 53, and according to flow and the water quality prediction flow and the water quality that flow into the first water burst of inflow canal 21 of rainy day in sewer pipe 20.

In the present embodiment, use is located at water quality analyser 25, flow meter 26, the fluviograph 27 in the sewer pipe 20, cycle collects their survey data 25a, 26a, 27a and is stored in accordance with regulations, thereby either party or both sides from Using Radar and Rain Gauge Data 11 and ground rainfall gauge 12 obtain amount of precipitation, and that can predict described measurement mechanism in the sewer pipe 20 is provided with locational flow and water quality.That is, can predict and flow into influx and the water quality that flows into before the canal 21.Therefore, can be more reliable and exactly the accumulating of rainwater plash 32, the break-make of rainwater pump 35, the break-make of foul water pump 36 are supported and controlled.

Again, according to continuous day fine fate of the computings such as weather forecast of Using Radar and Rain Gauge Data data 52a, ground rainfall gauge data 52b or Meteorological Information System, measure the water level 27a in the continuous sky sewer pipe 20 on fine date, the sand of hoarding in the prediction sewer pipe 20 and the height of deposit.Then, flow into the inflow water yield that flows into canal 21 and flow into water quality according to water quality 25a in the sewer pipe 20 of rainy day and flow 26a prediction.Fine consecutive hours, usually exist and adhere to the deposit long-pending trend of windrow of going forward side by side in the sewer pipe 20, rainy day then this deposit flow into and flow in the canal 21, if thereby the accumulating amount of obtaining like that in advance mentioned above, just can accurately predict just the inflow water yield and the inflow water quality of water burst.

Also can calculate BOD-SS load (handling the inflow load capacity of the organic dirty material BOD of each biological weight of place) according to flowing into water quality.

Claims (5)

1. rain drainage control system according to the flow decision rainwater pump running quantity of rainwater inflow pump field, and changes control rainwater pump according to pump field middle water level and turns on and off, and it is characterized in that, comprises

The prediction unit of the influx of prediction rainwater inflow pump field,

Calculate to discharge the required rainwater pump running quantity Calculation device of the water of rainwater influx of this prediction unit prediction,

The checkout gear of the upper reaches water level of the sewer pipe of testing pump field and

According to described rainwater pump running quantity and the present pump sump water level that the upper reaches water level of the detected sewer pipe of described checkout gear, described calculation element calculate, revise the predetermined water level and the predetermined modifier that turn-offs at least one side in the water level opened of rainwater pump.

2. the rain drainage control system described in claim 1 is characterized in that,

When the water level that should predetermined go up flow point surpassed predetermined threshold, this modifier is revised the rainwater pump predeterminedly opened water level and the predetermined at least one side who turn-offs in the water level.

3. the rain drainage control system described in claim 1 is characterized in that,

When should the predetermined water level of going up flow point being lower than predetermined threshold, this modifier is revised the rainwater pump predeterminedly opens water level and the predetermined at least one side who turn-offs in the water level.

4. the rain drainage control system described in claim 1 is characterized in that,

Predetermined water level and the predetermined SEA LEVEL VARIATION speed that goes up flow point that goes up flow point all is used as the predetermined index that goes up the water level of flow point.

5. a rain-water drainage control method is characterized in that, comprises following steps:

According to the step of the rainwater inflow pump field influx decision rainwater pump running quantity of prediction,

The step that turns on and off according to pump field SEA LEVEL VARIATION control rainwater pump and

According to the upper reaches water level of the sewer pipe of pump field, the described rainwater pump running quantity that is determined and pump sump water level now, revise opening water level and turn-offing water level of rainwater pump, make the rainwater pump open the step of constant time lag.

Images