JPH01251211A - Controller for water distribution pipe network - Google Patents

Abstract

PURPOSE:To set hydraulic pressure of a specific spot to within an allowable range by determining a hydraulic pressure target value schedule so that hydraulic pressure of each spot in a distributing water pipe network goes to uniform, based on a water distribution flow rate predicted value, correcting this hydraulic target value schedule and executing a hydraulic pressure control in the distributing water pipe network. CONSTITUTION:A water level H of a distributing reservoir 1, a water distribution flow rate Q from the reservoir 1, and hydraulic pressure (h) of each spot in a distributing water pipe network are detected by a water level detector 5, a water distribution flow rate detector 6, pressure reducing valve secondary pressure detectors 7a, 7b... and in-distributing water pipe network flow rate detectors 9a, 9b..., respectively, and supplied to an operation controller 13 through an input device 12. By the controller 13, a hydraulic pressure target value of each pressure reducing valve 4a, 4b... is calculated by an arithmetic operation, and outputted to pressure reducing valve controllers 10a, 10b... of the pressure reducing valves 4a, 4b... through an output device 14. Subsequently, the controllers 10a, 10b adjust the valve opening of the pressure reducing valves 4a, 4b... in accordance with this pressure reducing valve hydraulic pressure target value, and control the pressure reducing valve secondary pressure, by which hydraulic pressure in a distributing water pipe network 3 can be made correct.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to a water distribution pipe network control device in a water distribution facility such as a water supply facility, for example.

(Conventional technology) In general, in water supply facilities, the water pressure within the water distribution pipe network changes due to flow rate fluctuations within the water distribution pipe network, but water pressure control always maintains the upper and lower limits of the water pressure within the water distribution pipe network, regardless of flow rate fluctuations. Must be kept within range.

In order to solve this problem, the applicant of this invention disclosed the invention previously (Japanese Patent Application No. 61-308925).
Predicting future water distribution flow from past water distribution flow data,
Based on this predicted water flow rate, a water pressure target value schedule that satisfies the upper and lower allowable limits of water pressure within the water distribution pipe network and makes the water pressure within the water distribution pipe network uniform is determined, and the water pressure target value schedule is determined according to this target value schedule. We proposed a water distribution pipe network control device that controls water pressure by operating water distribution pumps, pressure reducing valves, etc.

(Problem to be solved by the invention) However, in the water distribution pipe network control device described above,
The knowledge to correct the water pressure target value is created based on the current plant characteristics (piping system, etc.), so if the plant characteristics change due to piping maintenance or changes to the water distribution system, the knowledge to correct the water pressure target value is created based on the plant characteristics. It no longer matches,
Problems have arisen in that control performance deteriorates and pressure fluctuations are sometimes not suppressed.

This invention was made to solve the technical problems remaining in the prior invention, and it is possible to perform stable water pressure management of the water distribution pipe network with always appropriate water pressure even if the plant characteristics change. The purpose is to provide a water distribution pipe network control device that can

[Structure of the Invention] (Means for Solving the Problems) A water distribution pipe network control device of the present invention includes a water distribution flow rate prediction means for predicting a future water distribution flow rate from past water distribution flow rate data in a water distribution pipe network, and this means. A water pressure calculation means that determines a target value schedule for the pressure in the water distribution pipe network that makes the water pressure at each point in the water distribution pipe network uniform based on the predicted water flow rate predicted by A database of actual process values from installed pressure gauges and flow meters, a knowledge base in which the relationship between the operation of a control point and the pressure change at each point for this operation is constructed as knowledge, said database and this knowledge base. means for inferring a correction amount for the pressure target value based on the data from the inference means, and adjusting the water pressure at each point so as to match the water pressure target value after correction of the water distribution pipe network pressure determined by the inference means. water pressure control means, and the knowledge base is constructed based on field know-how of water distribution systems, water distribution pipe network operation plans, etc.

(Function) The water distribution pipe network control device of the present invention predicts future water distribution flow rate based on past actual water distribution flow rate data. Then, based on this predicted value of water distribution flow rate, a water pressure target value schedule is determined so that the water pressure at each point in the water distribution pipe network is uniform, and based on this water pressure target value schedule, the water pressure in the water distribution pipe network, The correction amount for the water pressure target value at a certain point in time is inferred from the knowledge base of the relationship between process data such as flow rate and the pressure in the control point and the water distribution pipe network, and the correction amount is calculated to match the corrected water pressure target value. Water pressure is controlled by controlling the water pressure control means at each point, and the water pressure at each point within the water distribution pipe network is optimized.

In this case, in the present invention, the knowledge base is constructed based on field know-how of water distribution systems, water distribution pipe network operation plans, etc., so even if plant characteristics change, optimal water pressure control is always performed.

(Example) Hereinafter, one embodiment of the present invention will be explained in detail based on the drawings.

FIG. 1 shows an embodiment of the present invention, and FIG. 2 is a flowchart explaining the operation. This embodiment relates to a water distribution pipe network for water supply, and water is distributed from a water distribution reservoir 1 through a water distribution pipe 2 to a water distribution pipe network 3. A pressure reducing valve 4a is installed at a suitable location within the water distribution pipe network 3 as a means for controlling water distribution pressure.
, 4b, . . . are installed.

The water distribution reservoir 1 is provided with a water level detector 5 for detecting its water level, and the water distribution pipe 2 is provided with a water distribution flow rate detector 6.
is provided.

A pressure reducing valve 2 is provided on the secondary side of the pressure reducing valves 4a, 4b,...
Next, pressure detectors 7a, 7b, .
, 8b, ..., and flow rate detectors 9a, 9b in the water distribution pipe network
,... are provided.

Each of the pressure reducing valves 4a, 4b, . . . is provided with a pressure reducing valve control device 10a, 10b, . . . as pressure reducing valve control means for adjusting the valve opening degree. The pressure reducing valve control devices 10a, 10b, . . . are controlled, and the pressure reducing valve 4
The control system for adjusting the valve opening degrees of a, 4b, ... is as follows:
It is composed of a parameter setting device 11, an input device 12, an arithmetic and control device 13, and an output device 14.

The parameter setting device 11 sets a water pressure upper limit value δ, a water pressure lower limit value δ2, and a general fixed value h r @ f for water pressure within the water distribution pipe network.
and parameters such as initial values of the coefficients aO to a of the autoregressive model, and input them to the arithmetic and control unit 1 via the input device 12.
Enter 3.

The input device 12 includes a water distribution reservoir water level detector 5, a water distribution flow rate detector 6, and a pressure reducing valve secondary pressure detector 7a.

7b,..., water pressure detectors in the water pipe network 8a, 8b,...
, and various process values H, Q, h, q detected by the water distribution pipe network flow rate detectors 9a, 9b, . is stored as.

The arithmetic and control unit 13 includes a water distribution flow rate prediction means 13a and a water pressure calculation means 13b.The water distribution flow rate prediction means 13a calculates a predicted value of water distribution flow rate, and the water pressure calculation means 13b determines a water pressure target value in the water pipe. Therefore, for example, each of these parameters is calculated by a pre-built-in program configured by a process control computer, and the pressure reducing valves 4a, 4b are outputted via the output device 14.

. . to the pressure reducing valve control devices 10a, 10b, . . . .

This arithmetic and control device 13 further includes an inference means 1.3c, a database 13d, and a knowledge base 13e, and extracts the water pressure target value at the current time from the determined target value schedule every fixed period τ, and extracts the water pressure target value at the current time from the determined target value schedule, 7b, ・-:
8a, 8b, ...: 9a, 9b.

A database 13d that stores process values detected from ... and which pressure changes when the pressure reducing valves 4a, 4b, ... of which control points are operated in order to correct the water pressure target value. Infer a corrected water pressure target value based on information from the knowledge base 13e regarding the knowledge,
The corrected control value is output to the output device 14.

In particular, the knowledge base 13e in this embodiment is constructed based on the water distribution system, water distribution pipe network operation plan, etc. by the process shown in FIG. 3, as will be described later.

The operation of the water distribution pipe network control device having the above configuration will be described next.

Water is distributed from a water distribution reservoir 1 provided at a high altitude to a water distribution pipe network 3 via water distribution pipes 2.

In this water distribution operation, the water level H1 of the water distribution reservoir 1 is the water distribution flow rate Q from the water distribution reservoir 1 through the water distribution pipe 2, the water distribution pipe Mi43
The water pressure at each point within is measured by a water level detector 5, a water distribution flow rate detector 6, a pressure reducing valve secondary pressure detector 7a, 7b, . . .
Water pressure detectors 8a and 8b in the water distribution pipe network.

..., and flow rate detectors 9a, 9b, ... in the water distribution pipe network.
and is applied to the arithmetic and control unit 13 via the input device 12. In addition, the parameter setting device 11 sends necessary parameters, upper and lower limit values δ1 and δ2, and the coefficient of the autoregressive model to the arithmetic and control device 13.
a m and a general fixed value hrat for the water pressure within the water distribution pipe network are given via the input device 12.

The arithmetic and control device 13 calculates water pressure target values for each of the pressure reducing valves 4a, 4b, .
, . . . pressure reducing valve control device 10a.

10b, . . .

Then, in each of the pressure reducing valve control devices 10a, 10b, . . . , the pressure reducing valves 4a, 4b, . . .
, . . . and control the secondary pressure of the pressure reducing valve, the water pressure in the water distribution pipe network 3 is optimized.

Next, the arithmetic function in the arithmetic and control device 13 described above is
This will be explained with reference to the flowchart shown in the figure.

The arithmetic control device 13 receives setting parameters δ1, δ21 Flrer, ao~a from the parameter setting device 11.
Input m and store it in the database 13d.

Further, if there is a change in the parameter setting value in the parameter setting device 11, that value is also taken in and stored in the database 13d (step 100).

Subsequently, the arithmetic and control unit 13 performs demand forecasting processing (
Step 110). This demand forecasting process is as follows.

First, the water distribution flow rate Q input from the water distribution flow rate detector 6 is integrated over a certain period of time, for example, every hour, and is integrated at each time period k (
The water distribution flow JtQ(k> for each k=1 to 24) is accumulated.Then, the past few days, that is, the data for each day from j days ago (n-j> to yesterday (n-1>) are stored as Q- J (k >
, Qa-J+1 (k), =-.

...Q. −1(k), the predicted value Q of the water distribution flow rate Q for the time period k on the day n. (k) is calculated using the autoregressive model shown in equation (1) below.

Q-(k)=ao Qa-J (k) 10a I Ql
l-J+1 (k) 10-...+a□Q,-1(
k> (1) Here, the coefficients ao to am are parameters of an autoregressive model, and are determined by the iterative method of least squares or the like.

In this way, all time slots k (k=1~
24), the predicted water distribution flow rate value Q1(k) is calculated.

The predicted water distribution flow rate value Q for all time zones k on a specific day obtained in this way. (k>), a water distribution pipe network operation planning process is performed to determine a water pressure target value schedule P(k) for each time zone k (step 120).

In this process, the water pressure target value schedule P at a certain time is
(k), the water pressure h+ (k
> is the preset upper limit value of water pressure in the water pipe network δ1,
A water pressure target value schedule P (k) for each time period is created so that the water pressure within the water distribution pipe network becomes uniform while maintaining the lower limit value δ2.
Determine.

In this case, the problem can be formulated as follows, for example.

Objective function Σ(h+ (k) hr-r)2-minimum...(2
) Constraint condition δt > h+ (k) > δ2 (i=1~m) ... (3) f(h (k
), Q, (k>, r) = 0 (4) where, h + (k): Water pressure at point i during time period hrmf: Water pressure lower limit fixed value in the water distribution pipe network δ1:
Upper limit value of water pressure in the water distribution pipe network δ2: Lower limit value of water pressure in the water distribution pipe network f (h (k), Q. (k>, r) = 0: Water distribution pipe network equation h(k): Each point i in the time period For the water pressure ht(k>, h (k)”' (ht (k), ht(k
), ..., h, (k)). (k): Predicted water distribution flow rate value r: HaZen-W
The objective function of the above equation (2) is a coefficient representing the characteristics of the water distribution pipe network derived from the equation of illiamS.
This means that the sum of squares of the residual difference between the water pressure hl at each point i and the water pressure uniform general constant value h r *f in the water distribution pipe network 3 is minimized in order to equalize the water pressure.

In addition, here, the water pressure average constant value hr, t in the water distribution pipe network 3 refers to whether the water pressure at a high altitude point is operated at the lower limit of water pressure, or
Alternatively, it is an evaluation value used to decide whether to operate the water pressure at a low-elevation point at the upper limit, or whether to operate the water pressure at all points at the center of the range of upper and lower water pressure limits.

Moreover, the constraint condition of the above equation (3) means that the water pressure at all points is maintained within the upper and lower limit value ranges.

Furthermore, the constraint condition of equation (4) is the water pressure at each point.

This means that satisfies the pipe network equation, which is the basic equation of the water distribution pipe network 3. In other words, the flow rate balance equation at each point i, the water head balance equation between points, and the conditions that the total outflow in the water distribution pipe network 3 is equal to the water distribution amount predicted value, etc. are satisfied.

Then, such a problem is solved based on the objective function and constraint (F) based on nonlinear programming. Then, by solving this problem, the calculated water pressure hI at all points i
Among (k), pressure reducing valves 4a and 4b.

The water pressure at points corresponding to the positions of the pressure reducing valve secondary pressure detectors 7a, 7b, ... is determined by the pressure reducing valve secondary pressure target value P(k
).

In this way, by determining the pressure reducing valve pressure target value P (k), (k = 1 to 24) for each time period on a specific day n, a pressure reducing valve pressure target value schedule for one day is created. can.

Next, the water pressure target value at the current time is extracted from the target value schedule determined for each constant control period τ, and is extracted from the database 13d of process values detected from each detector and the knowledge base 13e for correcting the water pressure target value. The process of inferring the corrected water pressure target value will be explained.

The arithmetic and control unit 13 periodically performs the following operations.

First, the process values H, Q, h, and Q that change from moment to moment are inputted from each of the detectors 5 to 9 and stored in the database 13d (step 130).

Next, the inference means 13c determines the water pressure target value P(
A corrected water pressure target value is inferred from the process values, which are the general water flow rate Q, the flow rate in the water distribution pipe network q, and the water level in the water distribution reservoir H1 and the pressure in the water distribution pipe network (step 140).

This reasoning means 13c determines a corrected water pressure target value according to a knowledge base 13e constructed based on the degree of influence between the pressure reducing valve secondary pressure and the pressure within the water distribution pipe network. knowledge base 1
As the contents of 3e, for example, the following rules are stored.

Next, the procedure for constructing this knowledge base 13e will be explained with reference to the flowchart of FIG.

In step 301, a water distribution pipe network operation plan is input from an input device. The following plans are input as operation plans.

■ Water pressure within the water distribution network must be maintained within the upper and lower water pressure limits.

■ If the water pressure in the water distribution network falls below the lower water pressure limit,
Increase water pressure target value.

■ If the water pressure in the water distribution pipe network exceeds the water pressure upper limit, lower the water pressure target value.

In step 302, plant characteristics are input from the input device. The following data is input as plant characteristic data.

■ Piping system data (information on which pipes are connected to which pipes) ■ Piping data (data on pipe length, diameter, flow velocity coefficient) ■ Data on control points ■ Data steps on water pressure measurement points in the water distribution pipe network 30
In step 3, sensitivity analysis is performed based on the input data.

This sensitivity analysis uses the well-known pipe network calculation method as a tool for investigating the characteristics of hydraulic flow within the water distribution pipe network.

The sensitivity analysis method uses a pipe network calculation method to analyze how the water pressure at each measurement point changes when the operation amount of the control point (pressure reducing valve) is changed in several steps.

In step 304, an influence table of control points and measurement points is created based on the results analyzed in step 303.

For example, ■ The secondary pressure of the pressure reducing valve 4a is the pressure inside the pipe network 8a, 8b,
Affects 8d.

(2) The secondary pressure of the pressure reducing valve 4b affects the pipe network internal pressures 8c and 8d.

In step 305, knowledge regarding control is incorporated into the knowledge base based on the water distribution pipe network operation plan.

Then, the knowledge constructed as described above, the knowledge base 13
The knowledge regarding the control of e is incorporated as m 13 e −1, and the rules are, for example, as follows.

Rule 300: IF "Pressure reducing valve 4a is pressure reducing control effective THEN'" Calculate the maximum pressure pmax a and minimum pressure Pm1n a among the pressures 8a, 8b, and 8c in the water distribution pipe network affected by the secondary pressure of the pressure reducing valve 4a. "Rule 310 IF" Pressure reducing valve 4b is on pressure reducing control "THEN" Calculate the maximum pressure pmax b and minimum pressure Pm1n b among the pressures 8c and 8d in the water distribution pipe network affected by the secondary pressure of the pressure reducing valve 4b Parule 400: I F ”Pmax a <δ1katsuPm1n a>
δ2 PA T HE N"Since the control is normal, the water pressure target value P(
k) is output to the pressure reducing valves 4a, 4b, . . . without correcting the target value P. ,"Rule 5
00: I F ”PmaX a > δ1 or Pm1n a δ2
"THEN" correction amount = δl -PmaX a "Rule 510: IF" δl -pmaxa≧δ2-Pmin a"
THEN" correction amount = δ2-Pmirt a" Rule 6
00: IF "correction amount≠0'.

THEN'"p,,, ==Water pressure target value P, (k) + correction amount".

This knowledge will not change unless the operational strategy changes.

On the other hand, knowledge about the plant in knowledge base 13e "R13
For example, the following rule is used for e-2.

Rule 700: IF "Pressure reducing valve 4a is controlled THEN" The water pressure measurement points in the water distribution pipe network that affect the secondary pressure of the pressure reducing valve 4a are 8a, b, -8d. "Rule 710: IF" The pressure reducing valve 4b is controlled by THEN "The water pressure measurement point in the water distribution pipe network that affects the secondary pressure of the pressure reducing valve 4b is 8c.

It is 8d. This knowledge will not change unless there is a change in plant characteristics, such as maintenance (replacement) of water pipes or changes in the water distribution system.

The pressure target value p corrected by inference in this way,
, , is given to the output device 14, which gives a command output to the pressure reducing valve control device 10a for the pressure reducing valve 4a, and delays the control period (step 150).
.

Therefore, each pressure reducing valve control device 10a, 10b.

. . . controls the opening degrees of the pressure reducing valves 4a, 4b, . .

In this way, water pressure control for one day is performed in loop A, and water pressure control is performed while adding corrections to the water pressure target value at regular intervals, and when the water pressure control schedule for one day is completed, steps are performed based on loop B. Return to step 100 and create a water pressure target value schedule for the next day (step 160
).

In this way, in this embodiment, a daily target value schedule for the water pressure in the water distribution pipe network is created for each day, and the water pressure target value schedule is adjusted at regular intervals based on the actual flow rate and water pressure fluctuations of that day. By adding corrections based on the knowledge base, it is possible to control water pressure in line with reality.

It should be noted that the present invention is not limited to the above-mentioned embodiments, and the predicted value of the water distribution flow rate for the water distribution pipe network is not based on the iterative least squares method, but is based on GM that takes into account weather factors, for example.
It is also possible to use the DH method. Furthermore, instead of controlling water pressure within a water distribution pipe network using a pressure reducing valve, the present invention can also be applied to water pressure control within a water distribution pipe network using a water distribution pump. In this case, the relationship between the discharge pressure of the water distribution pump and the water pressure h (k) at each point in the distribution pipe network is made into a knowledge base, and the water pressure at each point operates within the allowable operational range. The discharge pressure may be controlled.

Further, the present invention is not limited to the above embodiments, and various modifications can be made within the scope of the technical idea described in the claims.

[Effects of the Invention] As described above, according to the present invention, the future water distribution flow rate is predicted from the past water distribution flow rate data in the water distribution pipe network, and the water distribution flow rate at each point in the water distribution pipe network is calculated based on the predicted water flow rate value. The water pressure target value schedule is determined so that the water pressure is uniform, the water pressure target value schedule is corrected based on the process values that fluctuate from moment to moment, and the water pressure in the water distribution pipe network is controlled using this corrected target value. , the water pressure at a specific point does not exceed the upper and lower allowable limits.

In addition, since the knowledge base has been constructed based on water distribution systems and water distribution pipe network operation plans, optimal water pressure control can be achieved without deteriorating control performance even when changes in bouland characteristics or operation methods occur. becomes.

[Brief explanation of the drawing]

FIG. 1 is a system diagram of an embodiment of the present invention, FIG. 2 is a flowchart explaining the operation of the embodiment, and FIG. 3 is a flowchart explaining the procedure for constructing knowledge in a knowledge base. 1...Water reservoir 2...Water pipe 3...Water pipe network 4a, 4b,...Pressure reducing valve 5...Water pressure detector 6...Flow rate detector? a,
7b, . . . Pressure reducing valve secondary pressure detector 8a, 8b,
...Water pressure detectors 9a, 9b, ... in the water distribution pipe network
...Flow rate detectors 10a, 10b in the water distribution pipe network, ...
...Reducing valve control device 11...Parameter setting device 12...Input device 13...Arithmetic control device 1
3a...Water distribution flow rate prediction means 13b...Water pressure calculation means 13c...Inference means 13
d...Database 13e...Knowledge base 14.
・・Output device

Claims (1)

[Claims] Water distribution flow rate prediction means for predicting future water distribution flow rate from past water distribution flow rate data within the water distribution pipe network, and distribution flow rate prediction means for predicting the future water distribution flow rate from past water distribution flow rate data within the water distribution pipe network, a water pressure calculation means for determining a target value schedule for the pressure within the water distribution pipe network at which the water pressure at each point is uniform; a database of actual process values from pressure gauges and flow meters installed at each point within the water distribution pipe network; a knowledge base in which a relationship between a control point operation and a pressure change at each point in response to this operation is constructed as knowledge; and means for inferring a correction amount for the pressure target value based on the database and data from the knowledge base. , a water pressure control means for adjusting the water pressure at each point to match the water pressure target value after correction of the water distribution pipe network pressure determined by the inference means, A water distribution pipe network control device characterized in that it is constructed based on field know-how such as operational plans.