JPH0968170A - Sewage pump control device in sewage treatment plant - Google Patents

Abstract

PROBLEM TO BE SOLVED: To transport a fixed amount of sewage water, unsteadily flowing in, to a waste water disposal facility, by estimating an inflow amount of sewage water flowing in for a fixed time to a sewage storage facility in a sewage treatment plant, and determining a transport amount of sewage water in accordance with a predetermined rule based on a present and future sewage water inflow condition. SOLUTION: A water level meter 18 in an inflow conduit 11 of sewage water flowing in from a sewage pipe conduit 21, water level meter 20 in a pump well 14, flow meter 16 behind a sewage pump 15 and a gate opening meter 19 in a dust removing inflow gate 12 are respectively arranged, and these measuring data are stored in a storage device 51. Here in an arithmetic device 52, an inflow amount during 1 control period is converted in that per 1 hour, so as to calculate an actual inflow amount. A weight coefficient of neural network determined in a learning device 31 is reflected in an inflow amount estimating device 32. Further, membership and arithmetic rule determined by a change device 41 are reflected to a target delivery amount arithmetic device 42. In accordance with a target water level or the like set in a setting device 53, a target delivery amount is determined.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control system for a sewage pump in a sewage treatment plant, and in particular, sewage that unsteadily flows into a sewage storage facility is treated as much as possible regardless of the inflow amount. The present invention relates to a sewage pump controller adapted to be transported to a facility.

[0002]

2. Description of the Related Art A sewage treatment plant is provided with a sewage storage facility for storing sewage collected from a sewage treatment area through a sewer pipe, a sewage treatment facility, and a sewage pump for transporting sewage from the sewage storage facility to the sewage treatment facility. Have. In addition, a sewage inflow culvert, a dust removal water channel, and a pump well are usually provided in the wastewater storage facility.

Sewage flows from a pipe to an inflow pipe in a wastewater storage facility and reaches a pump well. After that, it is transported to a sewage treatment facility by a sewage pump, treated, and discharged into a river.

When the sewage pump is automatically operated, usually, either one of the discharge amount constant control for keeping the discharge amount of the sewage pump constant and the water level constant control for keeping the water level of the pump well constant. Is adopted.

In the constant discharge amount control, the water level in the pump well changes due to the change in the amount of sewage flowing into the sewage treatment plant.

On the other hand, in the constant water level control, the water level in the pump well is kept constant, but the discharge amount of the pump fluctuates.

From the sewage treatment facility side, there is a demand for keeping the amount of sewage flowing into the treatment facility as constant as possible and performing stable treatment. However, since the fluctuation of the water level in the pump well is ignored, there is a problem that abnormal water level fluctuations tend to cause inundation of the wastewater storage facility and pump failure.

Therefore, as an alternative to the constant discharge amount control or the constant water level control, the invention of operating the pump by observing the tendency of the change of the present water level and the past water level and / or the inflow to the pump well. An invention has been found in which a pump is operated by storing a flow rate pattern and creating a discharge rate pattern such that the discharge rate does not change as much as possible, and the former is disclosed in JP-A-3-184101 and JP-A-4.
-191482, and the latter is disclosed in JP-A-3-115786.

[0009]

In the invention described in Japanese Patent Laid-Open No. 3-184101, the pump well water level when the pump is started and the pump well water level when the pump is stopped are determined based on the tendency of the water level of the pump well. Control the discharge amount of the sewage pump. This invention has no idea to control the discharge amount of the sewage pump to be constant.

Japanese Unexamined Patent Publication Nos. 4-191482 and 2-115
In all of the inventions described in Japanese Patent No. 786, the discharge amount of the sewage pump is aimed to be equalized. However, JP-A-3-115
The invention described in Japanese Patent No. 786 is not necessarily a satisfactory control method because it cannot deal with the case where the stored inflow flow rate pattern and the actual inflow flow rate are different.

According to the present invention, the sewage that flows in unsteadily is
An object of the present invention is to provide a sewage pump control device capable of transporting a certain amount to a sewage treatment facility as much as possible regardless of the inflow amount.

[0012]

Means for Solving the Problems The present invention relates to plant data for storing time series data of the water level of an inflow channel, the water level of a pump well, the gate opening of a dust removal channel, and the discharge amount of a wastewater pump in a wastewater storage facility. A storage unit, and an actual inflow amount calculation unit that calculates the amount of wastewater that has flowed into the sewage treatment plant from a certain point in the past to the present based on the pump well water level and the discharge amount of the wastewater pump stored in the plant data storage unit. And, based on the time-series data stored in the plant data storage means, the actual inflow amount, and the rate of change of the actual inflow amount, predict the inflow amount of the sewage flowing into the sewage storage facility within a certain period of time ahead of the present Sewage inflow predicting means, and a pump well water level deviation is calculated from the pump well water level and a predetermined target water level of the pump well, and the predicted inflow amount obtained by the sewage inflow predicting means and the pump Seeking a pumping amount of deviation from the discharge amount,
By these, a target discharge amount calculation means for creating a target discharge amount calculation rule for keeping the discharge level as constant as possible while suppressing the water level of the pump well and determining the discharge amount of the wastewater pump according to the rule, and the target discharge amount calculation means A sewage treatment system, comprising: a sewage pump operation control means for controlling at least one of the number of operating sewage pumps, the number of revolutions, and the opening of a discharge valve so that the discharge amount determined by the amount calculating means is obtained. It is located in the sewage pump controller of the plant.

The weight coefficient of the neural network is determined by the back-propagation method using the time-series data stored in the plant data storage means and the actual inflow amount and the change rate of the actual inflow amount obtained by the actual inflow amount calculating means. Equipped with learning means for inputting the time series data stored in the plant data storage means, the actual inflow amount and the change rate of the actual inflow amount obtained by the actual inflow amount calculating means, and based on the weighting factor, It is further desirable to include the wastewater inflow predicting means for predicting the inflow quantity of wastewater flowing into the wastewater storage facility within a certain time period ahead.

Further, in the fuzzy inference, using the pump well water level and the pump discharge amount stored in the plant data storage means, the predetermined target water level of the pump well and the estimated inflow amount obtained by the wastewater inflow amount estimation means, The pump well water level stored in the plant data storage means is provided with a membership function for calculating the discharge amount to be as constant as possible while suppressing the fluctuation of the pump water level within a certain range, and means for changing the inference rule. And the pump discharge amount, the predetermined target water level of the pump well and the predicted inflow amount obtained by the wastewater inflow predicting means are input, and the target discharge amount of the wastewater pump is determined by fuzzy reasoning according to the membership function and the inference rule. It is also desirable to do.

In the present invention, when the water level of the inflow channel and the water level of the pump well in the wastewater storage facility are configured to exhibit the same behavior, either one of the inflow channel water level and the pump well water level is It may be input to the learning means or the wastewater inflow prediction means.

[0016]

The present invention predicts the inflow amount of sewage that will flow into the sewage treatment facility of the sewage treatment plant in the future for a certain period of time, and further, based on the present and future sewage inflow conditions, the target transport amount of sewage as much as possible. Determine the amount of sewage transported according to a predetermined rule so as not to change.

[0017]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a block diagram of a sewage pump controller according to the present invention.

The sewage treatment plant runs from its treatment area to the sewer pipe 2
Wastewater such as domestic wastewater and industrial wastewater collected through 1 is treated and discharged into rivers. Sewage that has flowed into the sewage treatment plant first enters the inflow port 11. The inflow channel 11 is connected to the dust removal water channel 13 and the pump well 14, and the wastewater flows from the inflow channel 11 into the pump well 14 and is then sent to the treatment facility 17 by the wastewater pump 15 for treatment.

A water level gauge 18 is provided in the inflow conduit 11 and a pump well 1
4, a water level meter 20 is installed, and a flow meter 16 is installed behind the sewage pump 15. The inflow channel water level, the pump well water level, and the sewage pump discharge amount are measured. The discharge amount of the dirty water pump is output as a discharge amount per hour, for example. In addition, the opening degree of the dust removal inflow gate 12 is measured by the gate opening degree meter 19 and is zero.
To 100 [%].

These measurement data are stored in the plant data storage device 51 as time series data.

The actual inflow amount calculation device 52 first calculates the amount of sewage actually flowing into the treatment plant from the pump well water level and the pump discharge amount according to the following equation within a certain fixed time (control cycle).

Qin = ΣPout + ΔV Qin: Inflow amount of sewage during one control period ΣPout: Discharge amount of sewage pump during one control period ΔV: Change amount of storage amount during one control period Here, the storage amount is as shown in FIG. Represents the amount of sewage stored in pipes and storage facilities in the treatment plant. In addition, the storage volume is obtained from the "storage volume table" that represents the relationship between the well water level and the storage volume that was created in advance from the civil engineering drawings. This storage volume table shows the pump well water level and storage volume as shown in Fig. 3. It is represented in a one-to-one relationship.

The inflow amount obtained during one control cycle obtained here is further converted into an actual inflow amount per hour.

The learning device 31 determines the weighting coefficient of the neural network. The weighting factor is determined by the back propagation method based on various plant data accumulated in the plant data storage device 51. The determined weight coefficient is immediately reflected in the inflow amount prediction device 32.

The inflow amount predicting device 32 predicts the future inflow amount of sewage from various plant data stored in the plant data storage device 51. For the prediction, a multiple regression model that calculates the future inflow as a linear sum of the past inflow, the water level, and the pump discharge, and the fluctuation pattern of the inflow of sewage are stored, and future values are stored based on that memory. Although a prediction method or the like can be used, a neural network as shown in FIG. 4 is used in this embodiment.

The neural network shown in FIG. 4 has the pump well water level (if the water level of the pump well and the water level of the inflow culvert are the same behavior, the inflow basin water level may be used) as an input, the dust removal inflow gate opening, and the sewage. Given the time-series data of the plant such as the pump discharge amount, the actual inflow amount, the actual inflow amount change rate, etc., the predicted inflow amount (per hour) is output. Here, the actual inflow rate change rate means the difference between the current inflow rate and the actual inflow rate one control cycle before.

The membership rule changing device 41 can create and modify membership functions and calculation rules for fuzzy inference used for target discharge amount calculation.
At that time, the past data stored in the plant data storage device can be used to create / correct. The membership function and calculation rule determined by the membership rule changing device 41 are immediately applied to the target discharge amount calculation device 4
Reflected in 2.

In the pump well target water level setting device 53, the operator sets the pump well target water level as a control target, and the value is stored.

In the target discharge amount calculation device 42, the target water level of the pump well set by the pump well target water level setting device 53,
Information that is calculated from information such as the predicted inflow amount obtained by the inflow amount prediction device 32, the actual inflow amount stored in the plant data storage device 51, the pump well water level, and the sewage pump discharge amount (for example, future pump well water level) Is input and the target discharge amount is determined.

In this embodiment, fuzzy inference is used to determine the target discharge amount, and the pump well water level deviation and the pumped water amount deviation are used as the inputs of the fuzzy inference. The pump well water level deviation is the difference between the pump well water level and the pump well target water level at that time, and the pumping water quantity deviation is the difference between the predicted inflow amount and the current pump discharge amount.

Pump well water level deviation = Pump well water level-Pump well target water level Pumped water quantity deviation = Predicted inflow amount-Pump discharge amount Table 1 shows an example of a target discharge amount calculation rule by fuzzy inference used in this embodiment.

[0033]

[Table 1]

According to the rule of Table 1, the water level deviation is large in the plus and the pumping rate deviation is not large, or when it is in the plus side, the target discharge amount is increased, and the pump well water level deviation is large in the minus and the pumping quantity deviation is not large. Alternatively, when it is negative and large, the target discharge amount is reduced. That is, the target discharge rate is changed only when the current situation is not so good (the pump well water level deviation is large) and the situation is unlikely to improve in the future, and is constant at other times.

Here, ambiguous expressions such as “plus large”, “not large”, “increase”, “decrease” are expressed by membership functions as shown in FIGS. 5 to 7.

In FIG. 5, the pump well water level deviation is ± 0.
The evaluation of “not large” at 5 [m] is high, which means that this fluctuation range is within an allowable range. On the other hand, ± 1
When the value is [m] or more, the evaluation of “large in plus” or “large in minus” is high, and this fluctuation range is not desirable.

The pumping amount deviation of 6, 0 [m ³ / hr] has a "not greater" in ± 600 [m ³ / hr] mainly, the difference in the predicted inflow and sewage pump discharge quantity The smaller the value, the higher the evaluation.

Next, the process until the target discharge amount is actually obtained will be described in detail. As an example, FIGS. 8 to 9 show the target discharge rate calculation process in the case of the pump well water level deviation: -1.2 [m] and the pumped water quantity deviation: -500 [m ³ / hr].

First, when the pump well water level deviation and the pumped water quantity deviation are obtained, the respective suitability is obtained based on the membership function. In the case of the example, the pump well water level is -1.2
Since it is [m], the degree of “negative and large” is 1.
It becomes 0 (Fig. 8). On the other hand, the deviation of the amount of pumped water is -500 [m ³
/ Hr] spans both "minus and large" and "not large". In this case, the suitability is calculated for both of the two. It is 0.83 for "minus large" and 0.17 for "not large" (Fig. 9).

In the rule of this example, there are two conditions, that is, the pump well water level deviation and the pumped water quantity deviation, and therefore the degree of conformity as the two combined conditions is obtained. The conformity of the condition part shall be the smaller value of the conformity of pump well water level deviation and pumped water quantity deviation.

Condition part conformity = min (pump well water level deviation, pumped water amount deviation) Incidentally, in the calculation of the conformity of the pumped water amount deviation, the pumped water amount deviation spanned two memberships, so Calculate for two ways.

When the deviation of the pumped water amount is large and is large, the conformity of the conditional part = min (1.0, 0.83) = 0.83 When the deviation of the pumped water amount is not large, the conformity of the conditional part = min (1. 0,0.17) = 0.17.

According to the rules of Table 1, the conclusion corresponding to the condition part is extracted. In the case of this example, the pump well water level deviation is “negative and large”, and the pumping water amount deviation is “negative and large” and “not large”, so the conclusion is “to reduce the target discharge amount”. The membership function of the extracted conclusion part is divided by the goodness of fit of the condition part, and the trapezoid is cut out (hatched part in FIGS. 10 and 11).

Next, the cut trapezoids are overlapped with each other under all conditions (FIG. 12). The center of gravity is calculated for the superposed figures, and the amount of pumped water corresponding to the center of gravity becomes the target discharge amount deviation (this operation is called fuzzy).
In the case of this example, it becomes −395 [m ³ / hr].

This target discharge amount deviation is added to the current target discharge amount to obtain a new target discharge amount.

The sewage pump control device 54 controls the sewage pump 15 according to the target discharge amount calculated by the target discharge amount calculation device 42.
The number of pumps, the number of rotations, the discharge valve opening, etc. are determined, and the pump discharge rate is adjusted to the target discharge rate.

[0047]

The present invention predicts the future inflow amount of sewage, and based on the present / future situation, the target discharge amount of the sewage pump is set according to a rule created in advance so as not to change the target discharge amount as much as possible. decide. Therefore, regardless of the inflow of sewage, a certain amount of sewage can be transported to the treatment facility as much as possible, and stable treatment can be performed in the treatment facility.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a sewage pump control device.

FIG. 2 is a conceptual diagram of the actual inflow amount.

FIG. 3 is a graph showing the relationship between the storage amount and the pump well water level.

FIG. 4 is a configuration diagram of an inflow prediction neural network.

FIG. 5 is a diagram showing a membership function (pump well water level deviation).

FIG. 6 is a diagram showing a membership function (pumping amount deviation).

FIG. 7 is a diagram showing a membership function (target discharge amount).

FIG. 8 is a diagram showing the calculation of the fitness for each condition (pump well water level deviation).

FIG. 9 is a diagram showing the calculation of the fitness for each condition (pumping amount deviation).

FIG. 10 is a diagram showing the calculation of the condition part conformance (when the deviation of the pumping amount is negative and large).

FIG. 11 is a diagram showing the calculation of the degree of conformity of the condition part (when the deviation of the pumping amount is not large).

FIG. 12 is a diagram showing defuzziness.

[Explanation of symbols]

11 ... Inlet Drain, 12 ... Dust Removal Inlet Gate, 13 ... Dust Removal Channel, 14 ... Pump Well, 15 ... Sewage Pump, 16 ... Flow Meter, 17 ... Treatment Facility, 18 ... Water Level Meter, 19 ... Gate Opening Meter, 20 ... Water level gauge, 21 ... Sewer pipe, 31 ... Learning device,
32 ... Inflow amount predicting device, 41 ... Membership rule changing device, 42 ... Target discharge amount computing device, 51 ... Plant data storage device, 52 ... Actual inflow amount computing device, 53 ... Pump well target water level setting device, 54 ... Sewage pump controller.

Front page continuation (72) Inventor Mikio Yoda 52-1 Omika-cho, Hitachi-shi, Ibaraki Hitachi Ltd. Omika factory

Claims (4)

[Claims] 1. A sewage storage facility for storing sewage collected from a sewage treatment area via a sewer pipe, a sewage treatment facility, and a sewage pump for transporting sewage from the sewage treatment facility to the sewage treatment facility, A sewage treatment plant sewage treatment plant control device comprising a sewage inflow channel, a dust removal channel and a pump well in the sewage storage facility, the water level of the inflow channel and the water level of the pump well and the dust removal channel in the sewage storage facility. Based on the pump well water level and the sewage pump discharge amount stored in the plant data storage unit, and the plant data storage unit that stores each time series data of the gate opening degree and the discharge amount of the sewage pump. Actual inflow amount calculation means for calculating the amount of sewage that has flown into the sewage treatment plant from the time to the present, time series data stored in the plant data storage means, the actual inflow amount and the actual inflow amount Based on the rate of change, from the sewage inflow predicting means for predicting the sewage inflow into the sewage storage facility within a certain time ahead of the present, from the pump well water level and the predetermined target water level of the pump well The pump well water level deviation is calculated, and the pumped water quantity deviation is calculated from the predicted inflow quantity obtained by the sewage inflow quantity prediction means and the pump well discharge quantity, and the discharge quantity is held while suppressing the water level of the pump well within a certain range. Target discharge amount calculation rule for making the discharge amount constant as much as possible, and a target discharge amount calculation means for determining the discharge amount of the sewage pump according to the rule, and a sewage pump for achieving the discharge amount determined by the target discharge amount calculation means. And a sewage pump operation control means for controlling at least one of the number of operating units, the number of revolutions, and the opening degree of the discharge valve of the sewage treatment plant. 2. The back propagation method according to claim 1, wherein the time-series data stored in the plant data storage means, the actual inflow quantity and the change rate of the actual inflow quantity obtained by the actual inflow calculation means are used. Learning means for determining the weighting coefficient of the neural network by, the time-series data stored in the plant data storage means and the actual inflow amount and the rate of change of the actual inflow amount obtained by the actual inflow operation means are input, Based on the weighting factor
A sewage pump control device for a sewage treatment plant, comprising: the sewage inflow amount predicting unit that predicts an inflow amount of sewage flowing into the sewage storage facility within a certain time period ahead of the present time. 3. The inflow channel water level and the pump according to claim 1 or 2, when the water level of the inflow channel and the water level of the pump well in the wastewater storage facility are configured to exhibit the same behavior. A sewage pump control device for a sewage treatment plant, wherein either one of the well water levels is input to the learning means or the sewage inflow prediction means. 4. The pump well water level and pump discharge amount stored in the plant data storage means, the target water level of a predetermined pump well, and the predicted inflow obtained by the wastewater inflow amount prediction means according to claim 1 or 2. And means for changing the membership function and inference rule for calculating the discharge amount to be as constant as possible while suppressing the water level fluctuation of the pump well within a certain range in the fuzzy inference, and the plant data storage means The pump well water level and pump discharge amount, the target water level of the predetermined pump well, and the predicted inflow amount obtained by the wastewater inflow predicting means, which are stored in, are input, and the wastewater is fuzzy inferred according to the membership function and inference rules. A sewage pump control device for a sewage treatment plant, comprising the target discharge amount calculation means for determining a target discharge amount of a pump.