JP7375616B2 - Operation variable explanatory variable selection device for water treatment facilities - Google Patents

Description

The present invention relates to an operation control variable explanatory variable selection device for water treatment facilities.

Conventionally, water quality control in water treatment facilities has been carried out by operators with experienced skills, taking into account the various conditions of the water treatment facility. There is a need for automation using artificial intelligence.
When automating water quality control using artificial intelligence, it is necessary to construct a predictive model for operating values in water treatment facilities and select explanatory variables in the predictive model.

Patent Document 1, which is an example of the prior art, discloses that operating variables that have a large correlation with an analysis value are selected by principal component analysis, and the relationship between at least one operating variable and the analysis value among the selected operating variables is calculated using a neural network. Based on a genetic algorithm, the combination of operating variables that minimizes the estimation error between the estimated value given as the output of the neural network and the analytical value obtained from actual measurements is determined from the operating variables other than the replaced operating variables that are replaced by the network. A technique for estimating an analysis value by searching has been disclosed.

Japanese Patent Application Publication No. 2001-273474

However, in the above-mentioned conventional technology, the estimated value and the actually measured analytical value are one-to-one, and changes over time are not taken into account, making it difficult to apply to systems that change over time, such as water treatment facilities. There was a problem that.

The present invention has been made in view of the above, and an object of the present invention is to provide a technique for selecting explanatory variables of a system that changes over time.

The present invention, which solves the above-mentioned problems and achieves the purpose, includes a data storage unit that stores trend data and water quality data, and uses a genetic algorithm to determine explanatory variables from the trend data and water quality data stored in the data storage unit. an explanatory variable selection and explanatory variable data acquisition unit that acquires data on combinations of generated explanatory variables, and operation operations necessary for creating an estimation model from the trend data and water quality data accumulated in the data storage unit; A driving operation amount data acquisition unit that acquires data on the quantity setting value, explanatory variable data acquired by the explanatory variable data acquisition unit, and driving operation amount data acquired by the driving operation amount data acquisition unit. an estimation model creation unit that creates an estimation model using the estimated model creation unit; an estimation model parameter storage unit that stores the parameters of the estimation model created in the estimation model creation unit; a driving operation amount estimation unit that outputs an estimated driving operation amount for a predetermined period using the data and parameters of the estimated model stored in the estimated model parameter storage unit; An operating amount explanatory variable selection device for a water treatment facility, comprising: an estimation accuracy calculation unit that calculates an estimation accuracy that is the accuracy of the estimated operating amount according to an error with the estimated operating amount.

In the operation variable explanatory variable selection device for a water treatment facility according to the present invention having the above configuration, the timing of a change over time in the estimated value of the operation input during the predetermined period is adjusted based on the timing of a change over time in the actual operation input. It is preferable to further include an alignment application section.

In the operation variable explanatory variable selection device for a water treatment facility according to the present invention having the above-mentioned configuration, the estimation accuracy calculation unit calculates the time-varying coefficient that increases with time and becomes larger as the value approaches the current time. It is preferable to calculate the estimation accuracy by multiplying by .

According to the present invention, explanatory variables of a system that change over time can be selected.

FIG. 1 is a diagram showing the configuration of a water treatment facility to which the operation amount explanatory variable selection device for a water treatment facility according to the first embodiment can be applied. FIG. 2 is a block diagram showing the configuration of the operation amount explanatory variable selection device for a water treatment facility according to the first embodiment. FIG. 3 is a flowchart showing a processing procedure for acquiring explanatory variable data using a genetic algorithm, which is used in the explanatory variable data acquisition unit shown in FIG. FIG. 4 is an explanatory diagram of a method for selecting explanatory variables using a genetic algorithm in the explanatory variable data acquisition unit shown in FIG. 2. FIG. 5 is a block diagram showing the configuration of an operation variable explanatory variable selection device for a water treatment facility according to the second embodiment. FIG. 6 is a diagram showing temporal changes in driving operation values, with the horizontal axis representing time t and the vertical axis representing driving operation values, in the second embodiment. FIG. 7 is a block diagram showing the configuration of an operation amount explanatory variable selection device for a water treatment facility according to the third embodiment.

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be described with reference to the accompanying drawings.
However, the present invention is not limited to the description of the embodiments below.

(Embodiment 1)
FIG. 1 is a diagram showing the configuration of a water treatment facility 200 to which a water treatment facility operation amount explanatory variable selection device 100 according to the present embodiment can be applied.
The water treatment facility 200 shown in FIG. 1 includes a first sedimentation tank 1, a reaction tank 2, a final sedimentation tank 3, a blower 4, an air volume adjustment valve 5, an aeration device 6, a first pump 7, It includes a second pump 8, a measuring device 9, a gravity concentration tank 10, a mechanical concentration tank 11, a digestion tank 12, a dehydration tank 13, and piping 21, 22, 23, and 24.

The first settling tank 1 is a settling tank into which raw water is introduced.
This raw water is wastewater containing organic matter.
In the first settling tank 1, solid-liquid separation of raw water is performed, and the effluent from the first settling tank 1 is sent to the reaction tank 2 through a pipe 21.
Raw sludge, which is the sludge initially settled in the settling tank 1, is sent to the gravity thickening tank 10 through the pipe 24.

The reaction tank 2 is a tank that contains microorganisms and uses the microorganisms to purify the water flowing out from the primary settling tank 1.
In the reaction tank 2, the microorganisms proliferate by assimilating organic matter contained in the outflow water from the initial settling tank 1, and activated sludge is formed by biological treatment using the microorganisms.
Outflow water from the reaction tank 2 is sent to the final settling tank 3 through a pipe 22.

The final settling tank 3 is a settling tank in which activated sludge contained in the outflow water from the reaction tank 2 is precipitated.
The supernatant of the final settling tank 3 is discharged outside the water treatment facility 200 as treated water.
A part of the sludge settled in the final settling tank 3 is returned to the reaction tank 2 through the piping 23 by the first pump 7 and reused.
The remaining sludge settled in the final settling tank 3 is sent as surplus sludge to the mechanical thickening tank 11 by the second pump 8.

The blower 4 supplies air to the plurality of air diffusers 6.
The air volume adjustment valve 5 is provided in a pipe that passes through each of the plurality of air diffusers 6, and adjusts the air volume by opening and closing.
The plurality of air diffusers 6 are provided at the lower part of the reaction tank 2 and are connected to the piping passed through the air volume adjustment valve 5 to send air whose amount of air is adjusted by the air volume adjustment valve 5 into the reaction tank 2. supply to.
When the amount of air blown to the reaction tank 2 is adjusted in this way, the DO (Dissolved Oxygen) value, which is the amount of dissolved oxygen in the reaction tank 2, is adjusted, and the progress of biological treatment is adjusted.

The first pump 7 is a return sludge pump that sends a portion of the sludge settled in the final settling tank 3 to the reaction tank 2 through the pipe 23.
The second pump 8 is a surplus sludge drawing pump that sends the remaining sludge settled in the final settling tank 3 to the mechanical thickening tank 11 as surplus sludge.

The measuring device 9 is a measuring device that measures each parameter indicating the water quality of the reaction tank 2, and measurement value data, which is the measured parameters, is sent to the operation amount explanatory variable selection device 100.
Here, as each parameter indicating water quality, the DO value, which is the amount of dissolved oxygen, and the MLSS (Mixed Liquor Suspended Solids) value, which is the concentration of suspended solids, can be exemplified.
The operator of the water treatment facility 200 determines the amount of operation to be controlled by referring to each parameter indicating the water quality measured by the measuring device 9.
Here, the manipulated variables to be controlled include the rotation speed of the first pump 7 that adjusts the amount of sludge returned from the final settling tank 3 of the water treatment facility 200, and the extraction of excess sludge from the final settling tank 3 of the water treatment facility 200. For example, the drawing amount per unit time of the second pump 8 to adjust the amount or the drawing time of excess sludge, and the injection rate of the polymer flocculant into the dewatering tank 13 of the water treatment facility 200 can be exemplified.

The gravity thickening tank 10 is a tank that concentrates raw sludge that has initially settled in the settling tank 1.
The mechanical thickening tank 11 is a tank that thickens the excess sludge drawn out from the final settling tank 3 by the second pump 8.
The sludge concentrated in the gravity thickening tank 10 and the mechanical thickening tank 11 is sent to the digestion tank 12.

The digestion tank 12 is a tank in which concentrated sludge is digested.
Here, the digestion process is preferably performed by, for example, an anaerobic digestion process.
In the anaerobic digestion treatment method, organic sludge is decomposed by anaerobic microorganisms.
The sludge decomposed by the digestion process is sent to the dewatering tank 13.

The dewatering tank 13 is a tank that dehydrates the sludge decomposed by the digestion process to reduce the water content of the sludge and reduce its volume.

The pipe 21 is arranged between the first settling tank 1 and the reaction tank 2, and is a pipe that sends outflow water from the first settling tank 1 to the reaction tank 2.
The piping 22 is arranged between the reaction tank 2 and the final sedimentation tank 3, and is a pipe that sends outflow water from the reaction tank 2 to the final sedimentation tank 3.
The pipe 23 is arranged between the first pump 7 and the reaction tank 2, and is a pipe that sends a part of the sludge in the final settling tank 3 to the reaction tank 2.
The pipe 24 is arranged between the primary sedimentation tank 1 and the gravity thickening tank 10, and is a pipe that sends raw sludge from the primary settling tank 1 to the gravity thickening tank 10.

In the water treatment facility 200 shown in FIG. 1, the main operation items are the inflow amount, which is the amount of water from the initial settling tank 1 to the reaction tank 2, the air flow rate, which is the amount of air supplied into the reaction tank 2, and the first The returned sludge amount is the amount of sludge returned to the reaction tank 2 by the pump 7, and the excess sludge extraction amount is the amount of sludge pulled out from the final settling tank 3 by the second pump 8.
Settings for each of these operation items differ depending on the processing plant.

The reaction tank 2 can be controlled by, for example, constant air flow control, constant ratio control, and constant DO control.
Here, the constant air volume control is control performed so that the air volume is constant, which is set as the target air volume value.
Further, constant ratio control is control performed so that the amount of air blown is in accordance with the amount of inflow from the initial settling tank 1 to the reaction tank 2, that is, the ratio between the amount of inflow and the amount of air blown is constant.
Further, the constant DO control is control performed so that the DO value of the reaction tank 2 becomes a set target DO value.

In addition, the main operation items are the rotation speed of the first pump 7 for adjusting the amount of returned sludge, the amount of extracted sludge per unit time or the time for extracting excess sludge for adjusting the amount of excess sludge extracted, and the main operation items for adjusting the amount of returned sludge. This is the injection rate of polymer flocculant.
The driving operation amount explanatory variable selection device 100 according to the present embodiment derives the explanatory variables for determining these operation items and the estimation accuracy for determining the explanatory variables.
Then, the operator determines the operation amount setting value, which is the operation amount of the controlled object, based on the explanatory variables derived by the operation amount explanatory variable selection device 100 for the water treatment facility and the estimation accuracy for determining the explanatory variables. .
In this way, by using the driving operation amount explanatory variable selection device 100 to determine the driving operation amount setting value of the controlled object, it becomes possible to use a person who does not have intuition, experience, or know-how as an operator. .

FIG. 2 is a block diagram showing the configuration of the operation amount explanatory variable selection device 100 for a water treatment facility according to the present embodiment.
The operating operation amount explanatory variable selection device 100 for water treatment facilities shown in FIG. , an estimated model parameter storage section 105 , a driving operation amount estimation section 106 , an estimation accuracy calculation section 107 , and an output section 108 .

The data storage unit 101 acquires measurement value data measured by the measuring instrument 9 and stores it as trend data and water quality data.
Examples of trend data include inflow water volume, sludge concentration, sludge flow rate, air flow volume, and temperature.
Examples of water quality data include NH ₄ concentration and NO ₃ concentration.
Note that the data storage unit 101 can be realized by a recording medium such as a semiconductor memory or a magnetic disk.

The explanatory variable selection and explanatory variable data acquisition unit 102 generates (selects) explanatory variables from the trend data and water quality data accumulated in the data accumulation unit 101 using a genetic algorithm, and uses the data of the generated explanatory variable combinations. get.
Note that the explanatory variable selection and explanatory variable data acquisition unit 102 can be realized by combining a processor such as an MPU (Micro-Processing Unit) or a CPU (Central Processing Unit) with a recording medium such as a semiconductor memory or a magnetic disk. can.

The driving operation amount data acquisition unit 103 acquires data of driving operation amount setting values necessary for creating an estimation model from the trend data and water quality data accumulated in the data storage unit 101.
Note that the driving operation amount data acquisition unit 103 can be realized by combining a processor such as an MPU or a CPU, and a recording medium such as a semiconductor memory or a magnetic disk.

The estimation model creation unit 104 creates an estimation model using the explanatory variable data acquired by the explanatory variable selection and explanatory variable data acquisition unit 102 and the driving operation amount setting value data acquired by the driving operation amount data acquisition unit 103. Create.
Here, the method for creating the estimation model is not particularly limited, but for example, a neural network or random forest can be used.
Note that the estimation model creation unit 104 can be realized by combining a processor such as an MPU or a CPU, and a recording medium such as a semiconductor memory or a magnetic disk.

The estimated model parameter storage unit 105 stores parameters of the estimated model created by the estimated model creation unit 104.
Note that the estimated model parameter storage unit 105 can be realized by a recording medium such as a semiconductor memory or a magnetic disk.

The driving operation amount estimating unit 106 uses the explanatory variable combination data acquired by the explanatory variable selection and explanatory variable data acquisition unit 102 and the parameters of the estimated model stored in the estimated model parameter storage unit 105. Outputs the estimated driving operation amount for a predetermined period.
Note that the driving operation amount estimating unit 106 can be realized by combining a processor such as an MPU or a CPU, and a recording medium such as a semiconductor memory or a magnetic disk.

The estimation accuracy calculation unit 107 uses the driving operation amount set value from the driving operation amount data acquisition unit 103 and the driving operation amount estimated value from the driving operation amount estimation unit 106 to calculate the accuracy of the driving operation amount estimated value. Calculate some estimation accuracy.
Here, the estimation accuracy is defined as, for example, calculating the root mean squared error (RMSE) between the estimated driving operation amount and the actual driving operation amount setting value operated by the operator during a predetermined period. It is evaluated by
Note that the estimation accuracy calculation unit 107 can be realized by combining a processor such as an MPU or a CPU and a recording medium such as a semiconductor memory or a magnetic disk.

The output unit 108 outputs the explanatory variables from the explanatory variable selection and explanatory variable data acquisition unit 102 and the estimation accuracy from the estimation accuracy calculation unit 107.
One or both of the outputted explanatory variables and estimation accuracy is outputted to, for example, an operating device of a water treatment facility, and contributes to the control of the operating device.

FIG. 3 is a flowchart showing a processing procedure for acquiring explanatory variable data using a genetic algorithm, which is used in the explanatory variable selection and explanatory variable data acquisition unit 102 shown in FIG.
The explanatory variable selection and explanatory variable data acquisition unit 102 first generates an initial group of combinations of explanatory variables (S1).
Note that the number of generations at this time is 1, and the number of ending generations is determined.
Next, the explanatory variable selection and explanatory variable data acquisition unit 102 estimates the driving operation value using the initial group of explanatory variable combinations generated in S1, and evaluates the estimation accuracy by calculating the degree of fitness. (S2).
Next, the explanatory variable selection and explanatory variable data acquisition unit 102 selects explanatory variables based on the evaluation results in S2 so that those with high estimation accuracy remain in the next generation (S3).
Next, the explanatory variable selection and explanatory variable data acquisition unit 102 generates a new combination of explanatory variables by crossover or mutation (S4).
Next, the explanatory variable selection and explanatory variable data acquisition unit 102 determines the current number of generations (S5), and if the current number of generations has not reached the predetermined number of ending generations (S5:N), , the explanatory variable selection and explanatory variable data acquisition unit 102 increases the number of generations (S6), returns to S2, and performs the process again starting from fitness evaluation for the increased number of generations.
Alternatively, if the current number of generations has reached the predetermined end generation number (S5: Y), the explanatory variable selection and explanatory variable data acquisition unit 102 ends the process.
According to the flowchart shown in FIG. 3, since the explanatory variables used are changed by crossover or mutation, explanatory variables with high estimation accuracy are selected as generations pass.

FIG. 4 is an explanatory diagram of a method for selecting explanatory variables using a genetic algorithm in the explanatory variable selection and explanatory variable data acquisition unit 102 shown in FIG. 2.
As shown in FIG. 4, when the number of explanatory variables is N (N is a natural number), the number of elements of the genetic algorithm is N, and 0 or 1 is input to the state of each element.
When the state of each element is 0, the explanatory variable of that element is not used, and when the state of each element is 1, the explanatory variable of that element is used.The combination of explanatory variables used for estimation. is determined.
Therefore, in the example shown in FIG. 4, the NH ₄ concentration and pH value are not used as explanatory variables, and the water temperature and MLSS value are used as explanatory variables.

In conventional water treatment facilities, operators have to collect various data including at least trend data such as inflow water volume, sludge concentration, sludge flow rate, air flow rate, and temperature, and water quality data such as _NH4 concentration and _NO3 concentration. Based on the items, operational setting values such as the amount of sludge drawn out in the initial settling tank and the final settling tank, the DO set value or air blowing rate in the reaction tank, or the flocculant injection rate are determined.
Determination of such operating operation setting values largely depends on the intuition, experience, and know-how of skilled engineers built up over a long period of time, and it takes a long time to train operators with skilled techniques.
On the other hand, with the aging of operators, there is an urgent need to pass on the technology, and a replacement using artificial intelligence (machine learning) is desired.

When reproducing past driving using machine learning, a learning model is constructed by determining combinations of explanatory variables from among these various items.
Furthermore, the learning model for water treatment operation is not constructed by estimation only at a certain point in time, but by estimation over a predetermined period such as several months or years.
Therefore, explanatory variables used for estimation should not be evaluated at a single point in time, but based on evaluation over a predetermined period of time, so that highly accurate results can be derived over the entire predetermined period.
Therefore, the estimation accuracy requires evaluation based on the cumulative error between the driving operation amount set value and the driving operation amount estimated value over a predetermined period.
In this embodiment, the cumulative error between the driving operation amount set value and the driving operation amount estimated value is evaluated as time series data, and explanatory variables are selected.
At this time, the selection of explanatory variables is important because the accuracy of estimating driving operation values varies greatly depending on the explanatory variables used to construct the learning model.
However, if the number of explanatory variables is n (n is a natural number) and all combinations of explanatory variables are tried, the number of trials will be 2 ⁿ -1, so efficiency is required in searching for combinations of explanatory variables.
As described in this embodiment, by using a genetic algorithm, it is possible to efficiently select explanatory variables with high accuracy even in a system that changes over time.

(Embodiment 2)
In this embodiment, a configuration in which an alignment application unit 109 is added to the configuration of Embodiment 1 will be described.
FIG. 5 is a block diagram showing the configuration of the operation amount explanatory variable selection device 100a for a water treatment facility according to the present embodiment.
The water treatment facility operating amount explanatory variable selection device 100a shown in FIG. The configuration is such that an alignment application unit 109 is added to the configuration, and the other configurations are the same as in the first embodiment, so the description thereof will be used.

Similarly to the first embodiment, the driving operation amount estimating unit 106 uses the data of the explanatory variable combinations generated by the explanatory variable selection and explanatory variable data acquisition unit 102 and the estimated model stored in the estimated model parameter storage unit 105. An estimated value of the driving operation amount for a predetermined period is output using the parameter and.
Then, as in the first embodiment, the estimation accuracy calculation unit 107 calculates the estimation accuracy, which is the accuracy of the estimated driving operation amount, and selects an explanatory variable based on the estimation accuracy.
Here, the estimation accuracy is evaluated by calculating the root mean squared error (RMSE) between the estimated value of the estimated driving operation amount and the driving operation amount operated by the operator during a predetermined period. Ru.
Then, explanatory variables with high estimation accuracy are selected.

FIG. 6 is a diagram showing temporal changes in driving operation values, with the horizontal axis representing time t and the vertical axis representing driving operation values, in this embodiment.
FIG. 6(A) is a diagram showing a waveform of a time change in an actual driving operation value, and FIG. 6(B) is a diagram showing a waveform of a time change in an estimated driving operation amount.
Comparing the waveform shown in FIG. 6(A) and the waveform shown in FIG. 6(B), although the shapes are similar, the timing at which the driving operation value changes is different.
When the estimation accuracy calculation unit 107 calculates the root mean squared error (RMSE) between the waveform shown in FIG. 6(A) and the waveform shown in FIG. 6(B), an error occurs and the estimation accuracy decreases. will be invited.
However, since water treatment data is time-series and does not undergo sudden changes, strict accuracy in the time direction of the data is not required.

Therefore, the alignment application unit 109 applies alignment to the waveform of the driving operation amount estimated value, which is the driving operation amount for a predetermined period, estimated by the driving operation amount estimating unit 106.
Specifically, the alignment applying unit 109 expands or contracts the waveform of the estimated driving operation amount so that the timing at which the slope changes matches the waveform of the actual driving operation value.
Here, as the waveform expansion/contraction method, a dynamic time expansion/contraction method can be exemplified.
FIG. 6(C) is a diagram showing the waveform of the time change of the actual driving operation value after alignment is applied, and FIG. 6(D) is a diagram showing the waveform of the time change of the estimated driving operation amount after alignment is applied. It is a diagram.
In this way, by using the alignment application unit 109, even if the waveform of the estimated driving operation amount is similar to the waveform of the actual driving operation amount set value and includes a deviation in the time direction, the By using the estimated values of the waveforms, it is possible to obtain appropriate explanatory variables for water treatment facilities while suppressing a decrease in estimation accuracy.

(Embodiment 3)
In Embodiment 1, a mode was described in which the estimation accuracy of the estimated value of the operation amount in a predetermined period was evaluated, but when automating the operation operation in an actual water treatment facility, it is preferable to apply the most recently introduced operation method. .
Therefore, in this embodiment, a mode will be described in which emphasis is placed on accuracy of time close to the current time.

FIG. 7 is a block diagram showing the configuration of the operation amount explanatory variable selection device 100b for a water treatment facility according to the present embodiment.
The water treatment facility operation amount explanatory variable selection device 100b shown in FIG. The only difference is that the second embodiment is provided, and the other configurations are the same as those of the first embodiment, so the description thereof will be referred to.

The estimation accuracy calculation unit 107b shown in FIG. 7 increases the estimation accuracy with time in order to increase the influence of estimation errors at times close to the current time, and the value becomes larger as the time approaches the current time. The time change coefficient _at is used.
Here, the time change coefficient a _t can be defined by the following equation (1) using time t and the number T of data to be estimated.

Then, as shown in equation (2) below, the error is multiplied by a time change _coefficient at in order to increase the influence of the estimation error at a time close to the current time.
Here, _yt represents the actual amount of driving operation, and _yt with a hat represents the estimated value of the amount of driving operation.

By using the above equation (2), it becomes possible to strongly reflect the current operating method in the estimation accuracy.

According to this embodiment, it is possible to obtain explanatory variables that strongly reflect the influence of the current operating method.

Note that the configuration of this embodiment and the configuration of Embodiment 2 may be combined.

Further, the present invention is not limited to the above-described embodiments, but also includes various modifications in which components are added, deleted, or converted to the above-described configuration.

1 Initial sedimentation tank 2 Reaction tank 3 Final sedimentation tank 4 Blower 5 Adjustment valve 6 Aeration device 7 First pump 8 Second pump 9 Measuring device 10 Gravity concentration tank 11 Mechanical concentration tank 12 Digestion tank 13 Dehydration tank 21, 22 , 23, 24 Piping 100, 100a, 100b Water treatment facility operation amount explanatory variable selection device 101 Data storage section 102 Explanatory variable selection and explanatory variable data acquisition section 103 Operation amount data acquisition section 104 Estimation model creation section 105 Estimation model Parameter storage unit 106 Operation amount estimation unit 107, 107b Estimation accuracy calculation unit 108 Output unit 109 Alignment application unit 200 Water treatment facility

Claims (3)

A data storage unit that stores trend data and water quality data;
an explanatory variable selection and explanatory variable data acquisition unit that generates explanatory variables using a genetic algorithm from the trend data and water quality data accumulated in the data accumulation unit, and acquires data on combinations of the generated explanatory variables;
a driving operation amount data acquisition unit that acquires data of driving operation amount setting values necessary for creating an estimation model from the trend data and water quality data accumulated in the data storage unit;
an estimation model creation unit that creates an estimation model using explanatory variable data acquired by the explanatory variable data acquisition unit and driving operation amount data acquired by the driving operation amount data acquisition unit;
an estimated model parameter storage unit that stores parameters of the estimated model created in the estimated model creation unit;
An operation that outputs an estimated driving operation amount for a predetermined period using data of a combination of explanatory variables generated by the explanatory variable data acquisition unit and parameters of the estimated model stored in the estimated model parameter storage unit. a manipulated variable estimator;
an estimated accuracy calculation unit that calculates an estimation accuracy that is the accuracy of the estimated driving operation amount according to a cumulative error between the operating amount set value for a predetermined period and the estimated driving operation amount for a predetermined period. A device for selecting explanatory variables for operation amount of treatment facilities. The water treatment facility according to claim 1, further comprising an alignment application unit that adjusts the timing of the temporal change in the estimated value of the operating amount for the predetermined period based on the timing of the temporal change in the actual operating amount. A device for selecting explanatory variables for driving operations. The estimation accuracy calculation unit calculates the estimation accuracy by multiplying the error by a time-varying coefficient that increases with time and becomes larger as the value approaches the current time. The apparatus for selecting an explanatory variable for operation amount of a water treatment facility according to item 2.

Images