JP2021146246A - Device, system and method for water treatment technology - Google Patents

Abstract

To provide a device, a system and a method for properly determining the injection ratio of a flocculant in water purification treatment.SOLUTION: An information processing device includes a learned model, a parameter generation part, and a calculation part. In the device, an explanatory variable of the model includes the injection ratio (EV1) of a flocculant and at least one of water quality data and/or driving operation condition data (EV2) before injecting the flocculant, an objective variable of the model includes at least one of water quality data (OV) after injecting and coagulating the flocculant, the parameter generation part generates a plurality of EV1, the model outputs OV from the EV2 and the generated EV1, the calculation part determines the injection ratio of the flocculant on the basis of the generated EV1 and the outputted OV, and the determination of the injection ratio of the flocculant is made on the basis of the fact that the absolute value of a change ratio of the OV to the EV1 is equal to or less than a prescribed reference value or less than the prescribed reference value.SELECTED DRAWING: Figure 2

Description

The present invention relates to devices, systems and methods of water treatment technology. More specifically, the present invention relates to an information processing device, a system, and a method for optimizing the injection rate of a flocculant used in water treatment.

From the viewpoint of environmental impact, the role of water purification equipment is important. In the water purification apparatus, a flocculant is injected into the water to be treated, whereby suspended solids are aggregated and flocs are formed. The flocs are solid-liquid separated and then filtered, which purifies the water to be treated.

Patent Document 1 discloses a system for controlling the injection rate of a flocculant. More specifically, when determining the optimum coagulant injection rate Kx for the water quality item x, the coagulant injection rate at which dX / dKx = 0 is extracted, and this is used as the optimum coagulant injection rate for the water quality item x. Is disclosed.

Patent Document 2 discloses a method of manufacturing a database for learning data used when constructing a model for acquiring an arbitrary predicted value in water treatment or sludge treatment by using a machine learning algorithm.

Patent Document 3 discloses a method of predicting drug injection data at the next time point from sensing data and drug injection data by using a decision tree learning algorithm.

Patent Document 4 discloses a control method and a control device for a water treatment plant that determines an optimum coagulant injection rate, and for the purpose of reducing the number of repetitions in a jar test regardless of the skill level of an operator or the like. It discloses that a control method and a control device which can be performed are provided.

Japanese Unexamined Patent Publication No. 2012-170848 JP-A-2018-158284 JP-A-2017-1230888 Japanese Unexamined Patent Publication No. 2019-181318

In water purification treatment, instead of blindly injecting a flocculant into raw water, it is common to carry out a test called a jar test to determine an appropriate injection rate of the flocculant. However, skill and experience are required to carry out this jar test and judge the result. Therefore, there is a problem that there is a shortage of veteran engineers and the transfer of technology is interrupted. Furthermore, there are large individual differences in the implementation of this jar test and the judgment of the results.

In addition, in the event of heavy guerrilla rains, which have been seen in recent years, sudden changes in water quality may occur, and the jar test may not be carried out in time. Although some of the above-mentioned Patent Documents 1 to 4 disclose that the injection rate of the flocculant is determined quickly by utilizing information technology, there is room for improvement from the viewpoint of determining an appropriate injection rate. there were.

Therefore, it is an object of the present invention to provide an apparatus, system, and method for more appropriately determining the injection rate of a flocculant.

As a result of diligent studies by the inventor, we decided to focus on the relationship between the injection rate of the flocculant and the water quality data. More specifically, paying attention to the rate of change of both, I came up with the idea of adopting an injection rate of a flocculant such that the absolute value of the rate of change is lower than a predetermined reference value. In this case, even if the value is slightly different from the optimum value of the coagulant injection rate, an effect comparable to that of the optimum value can be obtained. The present invention has been completed based on the above findings and includes the following inventions in one aspect.
(Invention 1)
An information processing device for optimizing the injection rate of coagulant.
The information processing device includes a trained model, a parameter generation unit, an information acquisition unit, and a calculation unit.
The trained model is configured to receive a given explanatory variable and output a given objective variable.
The predetermined explanatory variable includes the injection rate of the flocculant (EV1) and at least one water quality data and / or operation operation condition data (EV2) before the injection of the flocculant.
The predetermined objective variable includes at least one water quality data (OV) after injection / aggregation of the flocculant.
The information acquisition unit is configured to acquire EV2 of the water treatment plant.
The parameter generation unit is configured to generate a plurality of EV1s.
The trained model is configured to output an OV from the acquired EV2 and the generated EV1.
The calculation unit is configured to determine the injection rate of the flocculant based on the generated EV1 and the output OV.
The determination of the injection rate of the flocculant is determined based on the absolute value of the rate of change of the OV with respect to EV1 being less than or equal to a predetermined reference value or less than a predetermined reference value.
Device.
(Invention 2)
The apparatus of the present invention, wherein the determination of the injection rate of the flocculant is further determined based on the following conditions.
The determined coagulant injection rate is lower than EV1 when the OV is optimal.
(Invention 3)
The device of invention 1 or 2,
The device further comprises a group determination unit.
The group determination unit is configured to perform cluster analysis of the EV2 to determine a group.
The trained model is provided for each group.
The group determination unit selects the trained model based on the group determination result, and causes the selected trained model to output an OV.
Device.
(Invention 4)
It is a system for optimizing the injection rate of coagulant.
The system includes the information processing apparatus according to any one of Inventions 1 to 3 and a database.
In the database, the EV1 and the EV2 and the OV are stored in association with each other.
The information processing device is configured to acquire the EV1, the EV2, and the OV as learning data from the database, and construct a trained model.
system.
(Invention 5)
The system of invention 4
In the database, meteorological information is further stored in association with the EV1, the EV2, and the OV.
The meteorological information is stored over a range of at least one year.
system.
(Invention 6)
It is a method to optimize the injection rate of the coagulant.
The method is a method utilizing an information processing device including a trained model.
The trained model receives a predetermined explanatory variable, outputs a predetermined objective variable, and outputs a predetermined objective variable.
The predetermined explanatory variable includes the injection rate of the flocculant (EV1) and at least one water quality data and / or operation operation condition data (EV2) before the injection of the flocculant.
The predetermined objective variable includes at least one water quality data (OV) after injection / aggregation of the flocculant.
The trained model is constructed using a database in which EV1, EV2, and OV are associated and stored.
The method is
The process in which the information processing device acquires EV2 of the water treatment plant, and
A process in which the information processing device generates a plurality of EV1s,
A step in which the information processing device outputs an OV from the acquired EV2 and the generated EV1.
A step in which the information processing apparatus determines the injection rate of the flocculant based on the generated EV1 and the output OV.
Including
The step of determining the injection rate of the flocculant is determined based on the fact that the absolute value of the rate of change of the OV with respect to EV1 is equal to or less than a predetermined reference value or less than a predetermined reference value.
Method.

In one aspect, in the present invention, the determination of the coagulant injection rate is determined based on at least the following conditions.
-The absolute value of the rate of change of OV with respect to EV1 is less than or equal to the predetermined reference value or less than the predetermined reference value.

As a result, even if the value is slightly different from the optimum value of the coagulant injection rate, an effect comparable to that of the optimum value can be obtained.

In a further aspect, the determination of the coagulant injection rate is further determined based on the following conditions.
-The determined coagulant injection rate is lower than EV1 when OV is optimal.

As a result, the injection cost of the flocculant can be suppressed, and nevertheless, an effect comparable to that of the optimum value can be obtained.

The water purification system in one embodiment is shown. The information processing apparatus in one embodiment is shown. In one embodiment, the relationship between the coagulant injection rate (horizontal axis) and turbidity (vertical axis) is shown. The triangles in the graph conceptually indicate the rate of change. In one embodiment, the predicted value by the trained model and the measured value when the flocculant is injected are shown. The circle symbol indicates the measured value, and the × symbol indicates the predicted value.

Hereinafter, specific embodiments for carrying out the present invention will be described. The following description is for facilitating the understanding of the present invention. That is, it is not intended to limit the scope of the present invention.

1. 1. System Configuration (1) Configuration of the entire system In one embodiment, the present invention relates to a system for optimizing the injection rate of a flocculant. In one example, the plants to which the system is introduced are water treatment plants (for example, water purification plants, wastewater treatment plants, sewage treatment plants, urine treatment (sludge regeneration) plants, leachate treatment plants, seawater desalination plants, etc. It may also include a sludge treatment plant). FIG. 1 shows an outline of the water purification system according to the embodiment. The system includes at least an information processing device and a database. In addition, the system may further include a landing well, a flocculant injection device, a floc forming pond, a settling basin, a rapid sand filter, a measuring instrument, and the like. Raw water is purified through a landing well, a floc forming pond, a settling basin and a rapid sand filter.

The information processing device may have a general hardware configuration, and may include a processor, a memory (eg, RAM, etc.), a storage medium (eg, HDD, SSD, etc.), and a communication module. The information processing device may function as a server device or a client device.

The information processing device and the database may constitute one piece of hardware. Alternatively, the information processing apparatus or database may be distributed to a plurality of pieces of hardware. The software used when constructing the database is not particularly limited, and software known in the art can be used (Oracle (registered trademark), MySQL (registered trademark), SQL Server (registered trademark), etc.).

The landing well plays a role in adjusting the amount of raw water taken in and / or the water level. The flocculant injection device plays a role of injecting the flocculant into the floc forming pond. The floc forming pond plays a role of forming flocs and agglomerating turbidity after injecting a flocculant into raw water. The settling basin serves to settle the agglomerated turbidity. The rapid sand filter serves to filter the supernatant after precipitation. The measuring instrument plays a role of measuring the characteristics of water at each location in the equipment. Although FIG. 1 conceptually represents one measuring instrument, the system can include a plurality of and / or a plurality of types of measuring instruments. Typically, the same type of measuring instrument (for example, a device for measuring turbidity before injection and a device for measuring turbidity after agglutination) can be provided before and after injection of the agglutinating agent. .. The types of measuring instruments include explanatory variables described later (for example, temperature, turbidity, transparency, chromaticity, pH, alkalinity, number of fine particles, coagulant injection rate, pH adjusting chemical injection rate, and TOC (total organic carbon). A type of measuring instrument corresponding to carbon (total organic carbon), COD (Chemical Oxygen Demand), ultraviolet absorbance, organic matter fraction, etc.) can be provided.

The flocculant injection device and the measuring instrument may be connected through a network so as to be able to communicate with each other with the information processing device and the database described above.

(2) Information Processing Device FIG. 2 shows an information processing device according to an embodiment. The information processing device includes a trained model, a parameter generation unit, an information acquisition unit, and a calculation unit. Further, although not clearly shown in FIG. 2, a group determination unit may be provided.

The trained model is typically formed by applying a machine learning technique, and a predetermined objective variable can be output by inputting a predetermined explanatory variable. The specific technique of machine learning is not particularly limited, but is limited to the SVR method (support vector regression method), PLS method (partial least squares method: PartialLeast Squares), neural network method (for example, Deep Learning method), random forest method, or The decision tree method or the like is preferable. More preferably, it is a neural network method.

The number of neurons in each layer and the number of layers in the neural network method are not particularly limited and can be appropriately determined. For example, the number of neurons in each layer may be 1 to 20 (for example, 2 to 6), and the number of layers may be 1 or more (for example, 1 to 10).

The parameter generation unit generates parameters (specifically, the injection rate of the flocculant) used as a part of the explanatory variables of the trained model. For example, the parameter generator can generate values for the injection rates of the plurality of flocculants. The generated value is passed to the trained model together with the data from the measuring instrument, and the values of a plurality of predetermined objective variables corresponding to the generated values are output (for example, the parameter generator determines the injection rate of the candidate flocculant). When 6 are generated, 6 values of the objective variable are output correspondingly).

The information acquisition unit is directly or indirectly communicably connected to the above-mentioned measuring instrument, and can receive the value measured by the measuring instrument. In addition, the information acquisition unit can also acquire the values of parameters related to the operation of the water treatment plant (for example, stirring time, stirring speed, etc.). A part of the received values is used as data for the explanatory variables of the trained model.

The calculation unit can receive the explanatory variables generated by the parameter generation unit. The calculation unit can also receive the objective variable output by the trained model. Based on these, the calculation unit can generate a plot diagram showing the relationship between the explanatory variable and the objective variable. Further, the calculation unit can generate a curve showing the relationship between the explanatory variable and the objective variable by using the plot diagram. Then, the calculation unit can calculate the rate of change of the objective variable with respect to the explanatory variable independently of or depending on the plot diagram.

Further, the calculation unit can determine the injection rate to be adopted. The determination can be made at least on the basis of:
-The absolute value of the rate of change of the objective variable (Objective variable, OV) with respect to the explanatory variable (explanatory variable, EV) is less than or equal to the predetermined reference value or less than the predetermined reference value.

The determination can be further made on the basis of:
-The determined injection rate of the flocculant is lower than the value of the injection rate when the objective variable is optimal.

The determination can be further made on the basis of:
-Whether the value of the objective variable (obvious variable, OV) is equal to or more than a predetermined reference value, exceeds a predetermined reference value, is equal to or less than a predetermined reference value, or is a predetermined reference. Whether it is less than the value.

2. Database formation The most important purpose of the database included in a part of the above system is to store training data for building a trained model. Specifically, the purpose may be to store the result of the jar test as learning data. Of course, the data for learning is not limited to the result of the jar test, and for example, the operation data of the past water treatment plant may be used. Furthermore, a combination of operational data of a plurality of water treatment plants may be used, or a combination of jar test data of a plurality of water treatment plants may be used. Alternatively, a combination of operational data of at least one water treatment plant and jar test data of at least one water treatment plant may be used. For example, as data related to the result of the jar test and operational data, water quality data measured by the measuring instrument, injection rate of coagulant, and the like can be stored. Water quality data measured by the measuring instrument includes, for example, temperature, turbidity, transparency, chromaticity, pH, alkalinity, number of fine particles, TOC (total organic carbon), COD (Chemical Oxygen Demand), ultraviolet absorbance, and the like. Organic fractions can be mentioned. In addition, weather information (for example, weather, temperature, rainfall, typhoon (for example, wind power, central pressure, etc.), etc.), pH adjusting chemical injection rate, and the like may be stored. Water quality data (eg, TOC, UV absorbance, organic matter fractionation, turbidity, and transparency, especially turbidity and transparency) may be stored before and after injection of the flocculant. The format for storing the data is not particularly limited, and typically, the data may be stored in a table format.

In addition to or instead of the above water quality data, operation operation condition data may be stored as learning data. Examples of operating operation condition data include, for example, stirring time, stirring speed, and the like. However, the injection rate of the flocculant is not included in the operating condition data.

Since these data are used as explanatory variables and objective variables when performing learning, they are stored in association with each other (for example, stored in the same row in a table).

The information processing apparatus can acquire the explanatory variables and the objective variables from the database, provide training data to the model to be trained, and can build a trained model. It is preferable to adjust the appropriate amount of data according to the type of training model and the complexity of the model. Further, it is more preferable to store data for at least one year or more regarding weather information (for example, weather, temperature, rainfall, typhoon (for example, wind force, central pressure, etc.), etc.) in consideration of seasonal fluctuations. , This is to be provided as learning data.

The database can be changed as appropriate. For example, new learning data can be added to the database, existing learning data can be updated, and existing learning data can be deleted. Further, after the above-mentioned change is made, the changed learning data may be used for re-learning.

3. 3. Formation of trained model Various techniques as described above can be used for forming a trained model, but here, a neural network will be described as an example. The above-mentioned number of neurons is provided in each layer, and the above-mentioned number of neurons is further provided. In addition, an input unit is provided according to the number of explanatory variables. Explanatory variables include the coagulant injection rate (EV1) and at least one water quality data and / or operating condition data (EV2) prior to coagulant injection. Further, the objective variable includes at least one water quality data (OV) after aggregation occurs by injecting the flocculant (that is, after injection / aggregation of the flocculant) (for example, two or more water quality data). May include a combination of).

At least one water quality data before injection of the flocculant includes at least one of temperature, turbidity, transparency, chromaticity, pH, alkalinity, number of fine particles, TOC, COD, ultraviolet absorbance, and organic fraction. Furthermore, at least one water quality data prior to injection of the flocculant may include the pH adjusting chemical injection rate. The operating operation condition data may include stirring time, stirring speed, and the like. Moreover, "before injection of a flocculant" used in this specification is used as a relative meaning, not an absolute meaning. For example, when the flocculant is injected three times, the water quality data between after the first injection / aggregation and before the second injection is the water quality data after the injection / aggregation from the viewpoint of the first injection. From the viewpoint of the second injection, it is the water quality data before the injection. Therefore, the water quality data between each of these injections can also be included in the pre-injection water quality data. The same applies to the water quality data after injection and aggregation.

At least one water quality data after injection / aggregation of the flocculant is acquired for the purpose of confirming the effect of the injection of the flocculant, and thus can include those related to the effect. More specifically, at least one water quality data after injection / aggregation of the flocculant can include TOC, ultraviolet absorbance, turbidity, transparency and the like.

Input these explanatory variables (EV1 and EV2), calculate the difference (error) between the output value and the correct answer value, and adjust various parameters (eg, weighting to each neuron, constant term, etc.). By repeating this, a trained model can be constructed.

A trained model can be constructed by a machine learning method other than a neural network, but a method known in the art may be used (for example, refer to Patent Document 3 for a decision tree). Omit.

The trained model to be constructed is not limited to one. In a preferred embodiment, a plurality of trained models may be constructed.

For example, the information processing apparatus may include a group determination unit, and the group determination unit may perform cluster analysis. For example, at least one water quality data before injection of the flocculant, which is a part of the above explanatory variables, may be cluster-analyzed, divided into several groups, and a trained model may be constructed for each group. For example, when the explanatory variables are cluster-analyzed and divided into three groups A, B, and C, the trained model A is constructed using the explanatory variables corresponding to A, and the explanation corresponding to B is used. The trained model B may be constructed by using the variables, and the trained model C may be constructed by using the explanatory variables corresponding to C.

Then, when utilizing the trained model at the operation stage, the group judgment unit determines which group the input explanatory variable belongs to, and makes a judgment such as selecting the corresponding trained model. You may.

The specific method of cluster analysis is not particularly limited, and a method known in the art may be adopted.

4. Types of coagulants The types of coagulants are not particularly limited, and those known in the art can be used. For example, examples of the flocculant include ferric polysulfate, ferric chloride, PAC (polyaluminum chloride) or a sulfate band. Further, examples of the coagulant also include a polymer coagulant (polymer), a coagulation aid, an organic coagulant, an inorganic coagulant, a coagulation improver, and the like, and these may be used alone or in combination. You can also. In the water purification treatment, PAC can be typically used.

5. Method for determining the injection rate of coagulant By constructing the database and the trained model as described above, the preparation for operation in the water treatment plant using the coagulant is completed.

In one embodiment, the present invention relates to a method for optimizing the coagulant injection rate (particularly at the operational stage, relating to optimizing the coagulant injection rate into raw water). The method comprises at least the following steps.
(1) Step of acquiring at least one water quality data and / or operation operation condition data (EV2) before injection of coagulant in a water treatment plant (2) Step of generating a plurality of parameters of coagulant injection rate (EV1) (3) A step of outputting at least one water quality data (OV) after injection / aggregation of the coagulant from the acquired EV2 and the generated EV1 (4) Based on the generated EV1 and the output OV. Step to determine the injection rate of coagulant

Regarding the step (1) above, the information processing apparatus (particularly the information acquisition unit) can acquire water quality data and / or operation operation condition data from the measuring instrument. As for the specific items of water quality data, the same water quality data entered when building the trained model is acquired. For example, temperature, turbidity, chromaticity, pH, alkalinity, number of microparticles, TOC (total organic carbon), COD, UV absorbance, and organic fractions were used in building the trained model. If so, even in the operation stage, from the measuring instrument, the temperature, turbidity, chromaticity, pH, alkalinity, number of fine particles, TOC (total organic carbon, total organic carbon), COD, ultraviolet absorbance, etc. And obtain the organic matter fraction.

Regarding the step (2) above, the parameter generation unit generates a plurality of numerical values of the injection rate of the coagulant. When generating, a preset numerical value may be generated. Alternatively, a numerical value may be randomly generated. Alternatively, a specific value (for example, 0.1 mg / L, 1 mg / L, 10 mg / L, etc.) is centered on the standard flocculant injection rate calculated by substituting the water turbidity to be treated into an existing formula. It may be set in increments. Alternatively, a standard injection rate is determined for each range of water turbidity to be treated, and a specific value (for example, 0.1 mg / L, 1 mg / L, 10 mg / L, etc.) is set around the standard injection rate. You may set with. Alternatively, the range of the injection rate may be determined according to the water turbidity to be treated, and may be set in increments of predetermined values (for example, 0.1 mg / L, 1 mg / L, 10 mg / L, etc.). Later, for the purpose of comparing the predicted values at each injection rate, a plurality of candidate values for the injection rate of the flocculant are created. For example, in the case of a certain water turbidity to be treated, candidate values for injection rates of 11 flocculants are created every 2 mg / L in the range of 15 mg / L to 35 mg / L.

Further, the parameter generation may be divided into several stages. For example, the range of the injection rate is arbitrarily determined in a wide range (Example 10 to 200 mg / L), and EV1 is set in a predetermined step (Example 10 mg / L). After predicting, EV1 is set in a narrower range (eg 30 to 50 mg / L) including the optimum injection rate (zero rate of change) in finer increments (eg 2 mg / L), predicted, and described above. It is also possible to adopt the injection rate according to the rule of the rate of change of.

Regarding the step (3) above, the trained model receives the injection rate of the coagulant generated by the parameter generator and the water quality data and / or the operation operation condition data acquired from the measuring instrument as inputs, and injects and coagulates the coagulant. The predicted value of the subsequent water quality data (for example, the predicted value of turbidity) is output. For example, when the parameter generation unit generates five numerical values of the coagulant injection rate, the predicted values of the water quality data after the injection and coagulation of the five coagulants are output correspondingly.

Regarding the step (4) above, the calculation unit calculates the rate of change of the water quality data after the coagulant injection / aggregation with respect to the coagulant injection rate. For example, when the water quality data is turbidity, it can be calculated by the following formula.
Treatment water turbidity change rate _i = (Prediction treatment water turbidity _{i + 1} -Prediction treatment water turbidity _i ) / (Coagulant injection rate _{i + 1} (mg / L) -Coagulant injection rate _i (mg / L) )
It is also possible to create and display a curve showing the relationship between the water quality data after agglutination (for example, turbidity) and the injection rate of the agglutinant.

Next, the calculation unit calculates the rate of change of the treated water turbidity change rate by the method described above, and aggregates such that the absolute value of the rate of change is less than or less than a predetermined reference value. Determine the range of agent infusion rates. The predetermined reference value regarding the rate of change is not specifically limited, and can be appropriately set in consideration of the purpose of water treatment and the like. For example, in the example of FIG. 3, the range of the coagulant injection rate is determined so that the height of the triangle is lower than a predetermined value. In this case, t2 to t4 are expressed as a range of the injection rate of the flocculant so as to be below the predetermined reference value or below the predetermined reference value.

Preferably, in addition to the above conditions, the calculator determines a range of coagulant injection rates such that the objective variable is lower than the optimum injection rate value. Explaining with the example of FIG. 3, the range of t2 to t3 is expressed as the range of the injection rate of the target flocculant.

In the prior art (for example, Patent Document 1), when analyzing a curve, a value such that the rate of change is simply zero is extracted (K13 in FIG. 12 of Patent Document 1) (drawing attached to the present specification). 3), a portion below a predetermined threshold was extracted (K14 in FIG. 12 of Patent Document 1) (t1, t5 in FIG. 3 of the drawing attached to the present specification).

In this regard, by determining t2 to t4 as the range of the coagulant injection rate (for example, the range in which t2 <= coagulant injection rate <t3, t3 <coagulant injection rate <= t4). It can be expected to obtain turbidity comparable to that of t3, which is the optimum value. In particular, by determining t2 to t3, which is lower than the optimum value t3, as the range of the coagulant injection rate, it can be expected that the cost of the coagulant is reduced as compared with the optimum value t3. Here, since turbidity is adopted as the objective variable, the optimum value is the lowest value of turbidity. However, the optimum value may be the highest value or the lowest value, depending on the type of the objective variable. For example, TOC, ultraviolet absorbance, turbidity, and the like can be mentioned as examples of objective variables for which the minimum value is optimal. On the contrary, as an example of the objective variable in which the maximum value is optimal, transparency and the like can be mentioned.

Since it is extremely unlikely that a plurality of locations where the rate of change will be zero will occur, it is not necessary to substantially consider a problem such as a locally optimal solution. That is, it is not necessary to separately determine which of the regions of the injection rates of the plurality of coagulants at which the rate of change is zero is to be adopted.

Although it is not essential, in addition to the above, the calculation unit determines whether the value of the water quality data (for example, turbidity) after injection and aggregation of the coagulant is below the predetermined reference value or less than the predetermined reference value. The range of coagulant injection rates may be determined based further on whether or not this is the case. When the water quality data after injecting and aggregating the coagulant is transparent, the calculation unit sets the range of the coagulant injection rate based on whether it exceeds the predetermined reference value or exceeds the predetermined reference value. You may decide. For example, in the range of t2 to t4 where the rate of change is equal to or less than a predetermined value, there are a portion where the value of the water quality data (turbidity) after injection and aggregation of the coagulant exceeds a predetermined reference value and a portion where the value is lower than the predetermined reference value. I have something to do. In such a case, the injection rate may be determined based on the relationship between the value of the water quality data (turbidity) after injection and aggregation of the coagulant and the predetermined reference value.

After determining the coagulant injection rate, the information processing device can transmit, for example, a signal specifying the injection rate to the coagulant injection device. Then, the flocculant injection device can inject the flocculant so as to have a designated injection rate. As a result, the injection rate can be determined quickly, and it is possible to quickly respond to sudden changes in water quality.

Further, regarding the determination of the injection rate of the flocculant, the pinpoint injection rate may be output, or the numerical range of the candidate injection rate may be output. When outputting the pinpoint injection rate, first narrow down the numerical range by the above method, and then select the value with the lowest coagulant injection cost in the numerical range according to a predetermined rule (for example, in the numerical range. Etc.), pinpoint numerical values may be output.

6. Example First, a trained model using a neural network was constructed. The turbidity (degree) of the treated water of the jar test at the water purification plant was adopted as the objective variable. Further, as explanatory variables, PAC injection rate (mg / L), raw water temperature (° C.), turbidity (degree), pH, and alkalinity (mg / L) were adopted. The neural network is provided with a corresponding number of input parts. The number of neurons was two, and the number of layers of neurons in each layer was four in the first layer and two in the second layer.

Here, a trained model was constructed using one year's worth of jar test data as training data.

After constructing the trained model, we prepared the data of the raw water quality instrument to make the trained model predict.

Six candidates for the injection rate of the flocculant were generated. The candidates for the injection rate of the flocculant were 15, 20, 25, 30, 35, and 40 mg / L as random candidate values.

Data on explanatory variables (ie, water temperature, turbidity, pH, and alkalinity) were extracted from the raw water quality meter data. These extracted data and the injection rate of the candidate flocculant were input to the trained model to obtain the predicted turbidity value. The turbidity after injecting the flocculant was also measured at the injection rate of the candidate flocculant. These results are shown in FIG.

In the plot diagram shown in FIG. 4, the rate of change is visually shown, but the injection rate of the flocculant to be adopted may be determined based on the rate of change and the like. For example, as shown by the arrow in FIG. 4, the portion corresponding to the left side of the minimum value (that is, the direction in which the injection rate of the coagulant decreases), which has a low rate of change, is set as the injection rate of the coagulant to be adopted. Can be decided. In this example, 20 mg / L, which is the lowest injection rate among injection points having an absolute value of change of 0.1, more preferably 0.02, is adopted. Such a decision can be made quickly by using an information processing device.

The specific embodiments of the present invention have been described above. The above-described embodiment is merely a specific example of the present invention, and the present invention is not limited to the above-described embodiment. For example, the technical features disclosed in one of the above embodiments can be applied to other embodiments. Further, unless otherwise specified, for a specific method, it is possible to replace some steps with the order of other steps, and an additional step may be added between the two specific steps. The scope of the present invention is defined by the claims.

Claims (6)

An information processing device for optimizing the injection rate of coagulant.
The information processing device includes a trained model, a parameter generation unit, an information acquisition unit, and a calculation unit.
The trained model is configured to receive a given explanatory variable and output a given objective variable.
The predetermined explanatory variable includes the injection rate of the flocculant (EV1) and at least one water quality data and / or operation operation condition data (EV2) before the injection of the flocculant.
The predetermined objective variable includes at least one water quality data (OV) after injection / aggregation of the flocculant.
The information acquisition unit is configured to acquire EV2 of the water treatment plant.
The parameter generation unit is configured to generate a plurality of EV1s.
The trained model is configured to output an OV from the acquired EV2 and the generated EV1.
The calculation unit is configured to determine the injection rate of the flocculant based on the generated EV1 and the output OV.
The determination of the injection rate of the flocculant is determined based on the absolute value of the rate of change of the OV with respect to EV1 being less than or equal to a predetermined reference value or less than a predetermined reference value.
Device. The device according to claim 1, wherein the determination of the injection rate of the flocculant is further determined based on the following conditions.
The determined coagulant injection rate is lower than EV1 when the OV is optimal. The device according to claim 1 or 2.
The device further comprises a group determination unit.
The group determination unit is configured to perform cluster analysis of the EV2 to determine a group.
The trained model is provided for each group.
The group determination unit selects the trained model based on the group determination result, and causes the selected trained model to output an OV.
Device. It is a system for optimizing the injection rate of coagulant.
The system includes the information processing apparatus according to any one of claims 1 to 3 and a database.
In the database, the EV1 and the EV2 and the OV are stored in association with each other.
The information processing device is configured to acquire the EV1, the EV2, and the OV as learning data from the database, and construct a trained model.
system. The system of claim 4
In the database, meteorological information is further stored in association with the EV1, the EV2, and the OV.
The meteorological information is stored over a range of at least one year.
system. It is a method to optimize the injection rate of the coagulant.
The method is a method utilizing an information processing device including a trained model.
The trained model receives a predetermined explanatory variable, outputs a predetermined objective variable, and outputs a predetermined objective variable.
The predetermined explanatory variable includes the injection rate of the flocculant (EV1) and at least one water quality data and / or operation operation condition data (EV2) before the injection of the flocculant.
The predetermined objective variable includes at least one water quality data (OV) after injection / aggregation of the flocculant.
The trained model is constructed using a database in which EV1, EV2, and OV are associated and stored.
The method is
The process in which the information processing device acquires EV2 of the water treatment plant, and
A process in which the information processing device generates a plurality of EV1s,
A step in which the information processing device outputs an OV from the acquired EV2 and the generated EV1.
A step in which the information processing apparatus determines the injection rate of the flocculant based on the generated EV1 and the output OV.
Including
The step of determining the injection rate of the flocculant is determined based on the fact that the absolute value of the rate of change of the OV with respect to EV1 is equal to or less than a predetermined reference value or less than a predetermined reference value.
Method.

Images