JP2023147019A - Chemical injection control system, chemical injection control method, and automatic coagulation treatment apparatus - Google Patents

Abstract

To provide a chemical injection control system, a chemical injection control method, and an automatic coagulation treatment system capable of appropriately controlling an injection rate of chemicals injected into raw water even when a sudden change in properties of raw water occurs.SOLUTION: A chemical injection control system 100 includes: an information acquisition unit 11 that acquires information having water quality data of raw water and operation condition data of a water treatment plant that treats the raw water; a prediction unit 12 which predicts an injection rate of the chemical to be injected into the raw water using a machine learning algorithm with a learned model that acquires explanatory variables including water quality data and operation condition data and outputs an objective variable including an optimal injection rate of a chemical to be injected into the raw water; a confirmation unit 13 which performs a confirmation test by injecting the chemical into the raw water based on the prediction result of the injection rate of the chemical and confirms whether the prediction result is appropriate by comparing the water quality data of the treated water obtained by the confirmation test with a reference value; and a chemical injection control unit 14 which, based on the confirmation result of the confirmation unit 13, controls the injection rate of chemicals to be injected into raw water.SELECTED DRAWING: Figure 1

Description

The present invention relates to a chemical injection control system, a chemical injection control method, and an automatic aggregation processing apparatus.

From the perspective of environmental impact, the role of water purification equipment is important. In a water purification treatment apparatus, a flocculant is injected into the water to be treated, whereby suspended substances are flocculated and flocs are formed. The flocs are subjected to solid-liquid separation and then filtered, thereby purifying the water to be treated.

In water purification treatment, rather than blindly injecting a flocculant into raw water, it is common to conduct a test called a jar test to determine the appropriate injection rate of the flocculant. However, skill and experience are required to conduct this jar test and judge the results. Furthermore, there are large individual differences in the implementation of jar tests and the judgment of results. Therefore, if there is a shortage of veteran engineers, there is a problem that the transmission of technology will be disrupted.

In recent years, the field of machine learning has made remarkable progress, and research has been conducted on controlling the injection of flocculants using machine learning. Machine learning generally uses computer algorithms to automatically analyze patterns and rules in input data from past data. This machine learning allows anyone to quickly determine the injection rate of the coagulant without requiring the experience of a veteran engineer.

For example, Japanese Patent Laid-Open No. 2017-123088 (Patent Document 1) discloses a method of predicting drug injection data at the next point in time from sensing data and drug injection data using a decision tree learning algorithm. There is. In addition, JP 2021-146246 A (Patent Document 2) discloses that a trained model is used to predict the water quality after treatment in a coagulation treatment test from water quality data or driving operation condition data, and to inject a flocculant. A method for predicting the flocculant injection rate from the absolute value of the rate and the rate of change in water quality after treatment is disclosed.

Japanese Patent Application Publication No. 2017-123088 Japanese Patent Application Publication No. 2021-146246

In general, predictions using artificial intelligence (AI) such as machine learning can maintain relatively high precision in controlling the flocculant injection rate within the data range used for learning. However, the prediction accuracy when calculating data outside the range, such as during abnormal times, tends to decrease. On the other hand, in recent years, unprecedented abnormal situations such as torrential rains have occurred. When the water quality of inflowing raw water changes rapidly due to such an abnormal situation, it is not possible to deal with it simply by predicting the flocculant injection rate based on machine learning that analyzes patterns and rules, as described in Patent Documents 1 and 2. It can be difficult.

In view of the above problems, the present invention provides a chemical injection control system, a chemical injection control method, and an automatic aggregation process that can appropriately control the injection rate of chemicals injected into raw water even if sudden changes in the properties of raw water occur. Provide equipment.

In order to solve the above problems, the present inventors have conducted extensive studies and found that a predetermined process for checking whether or not the prediction result of the drug injection rate using a machine learning algorithm is appropriate. We found that it is useful to provide a

One aspect of the present invention, which was completed based on the above knowledge, includes: an information acquisition unit that acquires information including water quality data of raw water and operation condition data of a water treatment plant that processes the raw water; The injection rate of chemicals to be injected into raw water is calculated using a machine learning algorithm using a trained model that acquires explanatory variables including condition data and outputs objective variables including the optimal injection rate of chemicals to be injected into raw water. Based on the predicted results of the prediction unit and the chemical injection rate, a confirmation test is performed by injecting chemicals into the raw water, and the water quality data of the treated water obtained from the confirmation test is compared with the standard value. This is a chemical injection control system that includes a confirmation section that confirms whether the results are appropriate or not, and an injection control section that controls the injection rate of chemicals to be injected into raw water based on the confirmation results of the confirmation section.

In one embodiment of the chemical injection control system according to the present invention, if the predicted result of the chemical injection rate is appropriate, the confirmation unit adjusts the injection rate based on the predicted result to water treatment by injecting the chemical into raw water. If the predicted result of the chemical injection rate is not appropriate, the injection rate is changed and retested, and based on the retest, the chemical is injected into the raw water. determine the optimal chemical injection rate for water treatment.

In another embodiment of the chemical injection control system according to the present invention, the confirmation unit includes a treatment tank for treating raw water, a chemical injection unit for injecting chemicals into the treatment tank, and a stirrer for stirring the inside of the treatment tank. , a water quality analyzer that analyzes water quality data of treated water obtained in the treatment tank, and a control unit that controls the injection rate of chemicals to be injected into the treatment tank.

In another aspect of the present invention, explanatory variables including water quality data of raw water and operating condition data of a water treatment plant that treats the raw water are acquired, and objective variables including an optimal injection rate of chemicals to be injected into the raw water are determined. A process of predicting the injection rate of chemicals to be injected into raw water using a machine learning algorithm using a trained model to be output, and a confirmation test by injecting chemicals into raw water based on the predicted results of the chemical injection rate. , a process of comparing the water quality data of the treated water obtained from the confirmation test with the standard value to confirm whether the predicted results of the chemical injection rate are appropriate, and based on the confirmation results, This is a chemical injection control method including a step of controlling an injection rate of a chemical to be injected.

In one embodiment of the chemical injection control method according to the present invention, in the step of confirming, if the predicted result of the chemical injection rate is appropriate, the injection rate based on the predicted result is injected into the raw water and The injection rate of chemicals for treatment is determined, and if the predicted chemical injection rate is not appropriate, the injection rate is changed and retested, and the water quality data of the treated water obtained from the retest is This involves repeating the test until it falls within the standard value range, and determining the optimal chemical injection rate for water treatment by injecting chemicals into raw water.

In another embodiment of the chemical injection control method according to the present invention, the step of checking injects a flocculant as a chemical into the raw water to perform a coagulation sedimentation treatment, and as water quality data of the treated water obtained by the coagulation sedimentation treatment, This includes analyzing at least one of turbidity, chromaticity, pH, alkalinity, number of fine particles, TOC, COD, ultraviolet absorbance, and organic matter fraction.

In still another aspect of the present invention, the present invention obtains a prediction result of a coagulant injection rate predicted by a machine learning algorithm using a trained model for predicting an optimal injection rate of a coagulant to be injected into raw water. an information acquisition unit, a treatment tank that conducts a flocculation treatment test including a raw water collection process, a chemical injection process, a stirring process, and a standing process based on the predicted results of the flocculant injection rate and preset test conditions; , a water quality analyzer that analyzes the water quality data of the treated water obtained from the flocculation treatment test, and a confirmation that compares the water quality data of the treated water analyzed by the water quality analyzer with standard values to determine whether the predicted results are appropriate. If the prediction result is appropriate, control the injection rate of the coagulant in the chemical injection equipment for water treatment by injecting the coagulant into raw water based on the injection rate based on the prediction result, and calculate the predicted result. If it is not appropriate, perform a flocculation test with a new flocculant injection rate, repeat the flocculation test until the water quality data of the treated water obtained in the flocculation test meets the standard value, and then determine the optimum The present invention is an automatic agglomeration processing apparatus including an injection control section that controls the injection rate of the flocculant of the chemical injection equipment based on the flocculant injection rate.

According to the present invention, there are provided a chemical injection control system, a chemical injection control method, and an automatic aggregation treatment device that can appropriately control the injection rate of chemicals injected into raw water even if sudden changes in the properties of raw water occur. can.

1 is a schematic diagram showing an example of a chemical injection control system according to an embodiment of the present invention. It is a schematic diagram showing an example of a test device with which a confirmation part is provided. 1 is a flowchart illustrating an example of a drug injection control method according to an embodiment of the present invention. It is a flowchart which shows an example of a confirmation test.

Embodiments of the present invention will be described below with reference to the drawings. In the description of the drawings below, the same or similar parts are denoted by the same or similar symbols. The embodiments shown below are illustrative of devices and methods for embodying the technical idea of this invention. It is not something specific.

(Chemical injection control system)
As shown in FIG. 1, a chemical injection control system 100 according to an embodiment of the present invention includes a chemical injection control device 10 for controlling the injection rate of chemicals injected into a water treatment plant 20 that processes raw water. Be prepared.

As the water treatment plant 20, although not limited to the following, water purification treatment equipment can be used. The water treatment plant 20 includes, for example, a water receiving well 21, a rapid stirring tank 22, a floc formation tank 23, a settling tank 24, a chemical injection facility 25 for injecting chemicals into raw water, a rapid filtration tank 26, a water purification tank 27, and a water distribution tank 28. Be prepared.

The water landing well 21 adjusts the amount of raw water taken in, the water level, etc. The rapid stirring basin 22 performs rapid stirring to perform a mixing process after injecting chemicals such as a coagulant. The flocculation pond 23 injects a flocculant into raw water, forms flocs, and flocculates turbidity. The chemical injection equipment 25 is a device for injecting chemicals into the rapid stirring pond 22 or the flocculation pond 23.

The sedimentation tank 24 settles the coarse flocs formed in the floc formation tank 23 by sedimentation separation. The rapid filtration tank 26 rapidly filtrates the supernatant water obtained in the settling tank 24. The water purification pond 27 stores purified water after performing a predetermined sterilization treatment on the filtered water. The water distribution reservoir 28 stores purified water supplied from the water purification reservoir 27 and supplies it to each facility such as a home, school, or factory by gravity flow or the like.

Each piece of equipment in the water treatment plant 20 is provided with a measuring device (not shown). Although not limited to the following, in order to measure the same type of water quality before and after chemical injection, it is preferable to have the same type of measuring instrument before and after chemical injection. . The types of measuring instruments include temperature, turbidity, transparency, chromaticity, pH, alkalinity, number of fine particles, flocculant injection rate, pH adjustment chemical injection rate, TOC (total organic carbon), COD (chemical oxygen demand). It can be equipped with various measuring instruments for measuring the amount), ultraviolet absorbance, organic matter fraction, etc.

The chemical injection control device 10 can be configured with a general hardware configuration, and can include a processor, a memory (eg, RAM, etc.), a storage medium (eg, HDD, SSD, etc.), and a communication module. The drug injection control device 10 may function as a server device or a client device. The chemical injection control device 10 may constitute one piece of hardware, or may be distributed over a plurality of pieces of hardware.

The drug injection control device 10 includes an information acquisition section 11, a prediction section 12, a confirmation section 13, and an injection control section 14. The information acquisition unit 11 is directly or indirectly connected to the above-mentioned measuring instrument (not shown) included in the water treatment plant 20 so as to be communicable. The information acquisition unit 11 acquires water quality data of treated water, water quality data of raw water, etc. measured by a measuring device. Further, the information acquisition unit 11 acquires various parameters related to the operation of the water treatment plant 20 that processes raw water, such as operation condition data including the amount of raw water flowing into each facility.

As the chemical to be injection controlled, any chemical can be used as long as it is a variety of chemicals used in the water treatment plant 20. A flocculant is typically used. Examples of the flocculant include polyferric sulfate, ferric chloride, PAC (polyaluminum chloride), and aluminum sulfate. In addition to these, a polymer flocculant, a flocculation aid, an organic coagulant, an inorganic coagulant, or a coagulation improver can also be used, and these can be used alone or in combination.

In addition to coagulants, pH adjusters, alkaline agents, methanol and organic substances added in nitrification and denitrification treatments, surfactants, chlorine and hypochlorous acid used in sterilization treatments, etc. are also subject to injection control. It may also be used as a drug.

The prediction unit 12 acquires predetermined explanatory variables that include at least water quality data of the raw water and operational condition data of the water treatment plant 20 that processes the raw water, and obtains predetermined explanatory variables that include the optimal injection rate of chemicals to be injected into the raw water. A machine learning algorithm using a trained model that outputs objective variables is used to predict the injection rate of chemicals to be injected into raw water.

Machine learning algorithms include, but are not limited to, SVR method (support vector regression method), PLS method (Partial Least Squares method), neural network method (for example, Deep Learning method), random forest method, or decision tree method, The LSTM method (Long Short-Term Memory) or the like can be used. Among these, it is preferable to use the LSTM method as the machine learning algorithm in this embodiment.

As a trained model, for example, a data set in which water quality data of raw water, operation condition data of the water treatment plant 20 that processes the raw water, and optimal injection rate of chemicals to be injected into the raw water are associated is used as the learning data. , using a predetermined model built to predict the drug injection rate using the LSTM method as a machine learning algorithm.

For example, in the case of a water treatment plant, the water quality data included in the training data used to construct the trained model includes temperature, turbidity, transparency, chromaticity, pH, alkalinity, number of particles, TOC, COD, Examples include ultraviolet absorbance and organic matter fraction. In addition, it may also include, for example, meteorological information such as weather, temperature, rainfall, wind power of a typhoon, central pressure, etc., chemical injection rate during pH adjustment, and the like. The water quality data may include data before and after the injection of chemicals. If a chemical is injected multiple times, information on water quality data before and after each injection may be included. Operational condition data for each facility includes the size of the treatment tank, type of chemical, rated value of chemical injection rate, amount of inflow water, stirring time, stirring speed, pH condition, standing time, filtration speed, etc. Various information related to driving operations can be mentioned.

The "chemical injection rate" in this embodiment means, for example, the ratio of the chemical injection rate (mg) to the amount of inflow water (L), and is typically expressed by the following formula.
Injection rate (mg/L) = Chemical injection amount (mg) / Inflow water amount (L)
Control of the injection rate of the medicine depends on the method of the medicine injection equipment 25, but is performed, for example, by controlling the valves, pumps, etc. included in the medicine injection equipment 25.

The confirmation unit 13 performs a confirmation test by injecting chemicals into the raw water based on the prediction result of the injection rate of chemicals by the prediction unit 12, and converts the water quality data of the treated water obtained from the confirmation test to a predetermined reference value. By comparing the prediction result with The reference value can be set as appropriate by the operator. For example, in the case of a water treatment plant, the standard value may be a water quality standard value for tap water.

As a result of the confirmation test, if the confirmation unit 13 determines that the prediction result of the chemical injection rate by the prediction unit 12 is appropriate, the confirmation unit 13 applies the injection rate based on the prediction result predicted by the prediction unit 12 to the raw water. It is determined as the injection rate of chemicals for water treatment. On the other hand, if the predicted result of the injection rate of the chemical is not appropriate as a result of the confirmation test, the confirmation unit 13 changes the injection rate and performs a retest, and injects the chemical into the raw water based on the retest. determine the optimal chemical injection rate for water treatment. Then, the injection control section 14 controls the injection rate of the chemical that the chemical injection equipment 25 injects into the raw water based on the confirmation result of the confirmation section 13 .

The confirmation unit 13 may include a simple testing device as shown in FIG. 2 in order to perform a confirmation test by injecting chemicals into raw water based on the prediction result of the injection rate of chemicals by the prediction unit 12. . For example, the confirmation unit 13 includes a treatment tank 1 that processes raw water, a chemical injection unit 3 that injects chemicals into the treatment tank 1, an agitator 2 that stirs the inside of the treatment tank 1, and the treatment obtained in the treatment tank 1. It includes a water quality analyzer 5 that analyzes water quality data, and a control unit 6 that controls the injection rate of chemicals to be injected into the treatment tank 1. The confirmation unit 13 may further include a pH injection unit 4 that injects a pH adjuster into the processing tank 1.

The control unit 6 is connected to the stirrer 2 , the inflow valve 7 , the drain valve 8 , the chemical injection pump 31 , the pH adjustment pump 41 , and the sampling pump 51 . The control unit 6 is configured to output predetermined control signals to the stirrer 2, inflow valve 7, drain valve 8, chemical injection pump 31, pH adjustment pump 41, and sampling pump 51 to control the operation of each device. ing. Details of the operation of the confirmation test by the confirmation unit 13 will be described later.

According to the chemical injection control system 100 according to the embodiment of the present invention, the confirmation unit 13 performs a confirmation test on the prediction result of the chemical injection rate using machine learning, and performs water treatment by injecting the chemical into raw water. The optimal chemical injection rate for this purpose can be determined. As a result, even if the prediction results obtained by machine learning are not appropriate due to the effects of sudden weather changes, the prediction results can be easily tested and confirmed using the testing device shown in Figure 2, and the appropriate chemical injection rate can be determined. By determining this, it is possible to always inject chemicals into raw water at an appropriate injection rate. As a result, it is possible to provide a chemical injection control system 100 that can appropriately control the injection rate of chemicals injected into raw water and suppress deterioration in the quality of treated water even if a sudden change in the properties of raw water occurs.

(Chemical injection control method)
The drug injection control method according to the embodiment of the present invention can be controlled using the drug injection control system 100 shown in FIG. 1 according to the flowchart shown in FIG. 3. That is, the chemical injection control method according to the embodiment of the present invention includes a step S1 of acquiring water quality data of raw water and operation condition data of the water treatment plant 20 that processes the raw water, and a step S1 of acquiring water quality data of raw water and operation condition data of the water treatment plant 20 that processes the raw water. of chemicals to be injected into raw water using a machine learning algorithm using a trained model that obtains predetermined explanatory variables including at least data and outputs predetermined objective variables including the optimal injection rate of chemicals to be injected into raw water. Step S2 of predicting the injection rate; and based on the prediction result of the chemical injection rate, a confirmation test is performed by injecting the chemical into the raw water, and the water quality data of the treated water obtained from the confirmation test is compared with the standard value. , a step of confirming whether the predicted result of the chemical injection rate is appropriate (S3 to S7), and a step S8 of controlling the injection rate of the chemical to be injected into the raw water based on the confirmation result.

In step S1, the information acquisition unit 11 in FIG. 1 acquires water quality data and operation condition data of raw water to be treated in the water treatment plant 20. In step S2, the prediction unit 12 uses the same water quality data and driving operation condition data as the water quality data and driving operation condition data input when constructing the learned model from the water quality data and driving operation condition data acquired by the information acquisition unit 11. Obtain condition data and input it as an explanatory variable into the trained model to obtain a predicted value for the optimal injection rate of the drug. In step S3, the prediction unit 12 transmits the predicted value of the drug injection rate predicted by the learned model to the confirmation unit 13.

In step S4 of FIG. 3, the confirmation unit 13 of FIG. 1 receives the output of the predicted value of the chemical injection rate from the prediction unit 12, and performs a confirmation test by injecting the chemical into the raw water using the testing device of FIG. implement. Details of the confirmation test in step S4 will be described later. In step S5, the confirmation unit 13 analyzes the water quality data of the treated water obtained by the confirmation test using the water quality analyzer 5 included in the confirmation unit 13.

The water quality analyzer 5 used in step S5 includes temperature, turbidity, transparency, chromaticity, pH, alkalinity, number of particles, TOC, COD, ultraviolet absorbance, organic matter fraction, solid content (SS), MLSS, and pH. Various measuring devices are used to obtain water quality data such as:

In step S6, the confirmation unit 13 confirms whether the predicted result of the chemical injection rate is appropriate by comparing the water quality data analyzed by the water quality analyzer 5 with a predetermined reference value. When the water quality data satisfies the standard value, that is, when the predicted result of the chemical injection rate is appropriate, the confirmation unit 13 injects the chemical into the raw water to perform water treatment using the injection rate based on the predicted result. The chemical injection rate is determined as the chemical injection rate for carrying out the process, and the process proceeds to step S8. On the other hand, if the water quality data does not meet the standard value, that is, if the prediction result of the chemical injection rate is not appropriate, the process proceeds to step S7.

In step S7, the confirmation unit 13 changes the chemical injection rate to an injection rate different from the predicted value. Although the changing method is not limited to the following, the confirmation unit 13 increases the chemical injection rate when the predicted value of the chemical injection rate exceeds a preset reference value. If the predicted value of the chemical injection rate is less than the reference value, a changed chemical injection rate is determined so as to reduce the chemical injection rate, and the process proceeds to step S4. The increase or decrease in the injection rate at the time of change can be set in advance.

Next, in step S4, a retest is performed by injecting a chemical at the changed chemical injection rate determined in step S7. The confirmation unit 13 determines the optimum chemical injection rate by repeating steps S4 to S7 until the water quality data satisfies the reference value as a result of the retest. If the optimum chemical injection rate has been determined, the process advances to step S8.

In step S8, the injection control unit 14 transmits the optimum chemical injection rate for raw water treatment to the chemical injection equipment 25 of the water treatment plant 20. The chemical injection equipment 25 injects a chemical into the raw water at an optimal chemical injection rate based on the control signal output from the injection control unit 14 .

(About the detailed flow of steps S4 and S5)
The operation flow of the test step S4 and the water quality analysis step S5 using the test device of FIG. 2 included in the confirmation unit 13 will be described below using the flowchart of FIG. 4 and the test device of FIG. 2 as examples.

In the drug injection rate setting step S41 in FIG. 4, when information on the drug injection rate is input to the control unit 6 in FIG. 2 based on the prediction result of the drug injection rate by the prediction unit 12, the control unit 6 in FIG. , the drug injection rate by the drug injection unit 3 is set.

In the raw water collection step S42, the control unit 6 opens the inflow valve 7 connected to the treatment tank 1 to collect a predetermined amount of raw water to be treated by the water treatment plant 20, and causes the raw water to flow into the treatment tank 1. The inflow amount can be appropriately controlled by adjusting the water level using an overflow pipe, controlling using a water level gauge, controlling the inflow speed and opening time of a valve, etc.

In the chemical injection step S43, the control unit 6 controls the chemical injection pump 31 to inject, for example, a flocculant as a chemical into the processing tank 1 so as to achieve the chemical injection rate set in the chemical injection rate setting step S41. . At this time, if it is necessary to adjust the pH within the processing tank 1, the control unit 6 injects the pH adjuster into the processing tank 1 via the pH adjustment pump 41. There are two methods for adjusting the pH: a method of injecting a fixed amount of a pH adjuster, and a method of controlling using a pH measuring device (not shown).

In the rapid stirring step S44, the control unit 6 controls the stirrer 2 to stir the treated water at a preset stirring speed and stirring time. Generally, the stirring speed for rapid stirring is 100 to 150/min, and the stirring time is about 3 minutes.

In the slow stirring step S45, the control unit 6 controls the stirrer 2 to stir the treated water at a preset stirring speed and stirring time. The slow stirring is generally slower than the stirring speed in the rapid stirring step S44, for example, about 30/min, and the stirring time is about 10 minutes.

In the standing step S46, the control unit 6 stops the operation of the stirrer 2, and allows the stirrer 2 to stand still until a preset standing time. The standing time is preferably 5 minutes or more.

In the water quality analysis step S51, the control unit 6 sends the treated water in the upper part of the treatment tank 1 after the standing step S46 to the water quality analyzer 5 equipped with the sampling pump 51, and analyzes the water quality data of the treated water.

In the draining step S52, the control unit 6 opens the drain valve 8 to drain the treated water in the treatment tank 1. Subsequently, in the cleaning step S53, the treatment tank 1 and the water quality analyzer 5 are cleaned with the treated water after the confirmation test, pure water, tap water, or the like.

As described above, according to the drug injection control method according to the embodiment of the present invention, a confirmation test is performed on the predicted result of the drug injection rate using machine learning, and the predicted result is corrected as necessary. , it is possible to determine the optimal chemical injection rate for water treatment by injecting chemicals into raw water. Thereby, even if the prediction results obtained by machine learning are not appropriate due to sudden weather changes, etc., the prediction results can be corrected to an appropriate injection rate. As a result, even if a sudden change in the properties of raw water occurs, it is possible to appropriately control the injection rate of chemicals injected into raw water.

Although the present invention has been described by the above-described embodiments, it should not be understood that the statements and drawings that form part of this disclosure limit the present invention. That is, the present disclosure is not limited to the above-described embodiments, and it goes without saying that components can be combined and modified with each other without departing from the spirit of the disclosure.

(Water treatment plant 20)
In the example of FIG. 1, the water treatment plant 20 is a water purification treatment plant. However, the water treatment plant 20 according to the embodiment of the present invention is not limited to the configuration shown in FIG. 1, and can of course be used with various types of equipment that inject predetermined chemicals and perform predetermined water treatment. It is. For example, the water treatment plant 20 may be a water treatment plant, a pure water production plant, a wastewater treatment plant, a sewage treatment plant, a human waste treatment (sludge regeneration) plant, a leachate treatment plant, a seawater desalination plant, a sludge treatment plant, etc. Of course, it can be used in water treatment plants.

(Automatic flocculation processing equipment)
The chemical injection control device 10 in FIG. 1 is configured independently as an automatic coagulation treatment device, and can be connected not only to the water treatment plant 20 described above but also to various equipment including chemical injection equipment 25 that requires chemical injection control. Can be done.

That is, the chemical injection control device 10 according to the embodiment of the present invention performs injection of the flocculant predicted by a machine learning algorithm using a trained model for predicting the optimal injection rate of the flocculant to be injected into raw water. Based on the information acquisition unit 11 that acquires the prediction result of the coagulant injection rate, and the test conditions set in advance, the raw water collection step S42, the chemical injection step S43, the stirring steps S44, S45, and the static A treatment tank 1 that carries out a flocculation treatment test including a coagulation treatment step S46, a water quality analyzer 5 that analyzes water quality data of treated water obtained in the flocculation treatment test, and a water quality analyzer 5 that analyzes water quality data of treated water analyzed by the water quality analyzer 5. A confirmation unit 13 compares it with a reference value and determines whether the predicted result is appropriate, and if the predicted result is appropriate, a coagulant is injected into the raw water based on the injection rate based on the predicted result to perform water treatment. The injection rate of the flocculant in the chemical injection equipment 25 is controlled, and if the predicted result is not appropriate, a flocculation treatment test is carried out with a new flocculant injection rate, and the water quality data of the treated water obtained from the flocculation treatment test is The apparatus includes an injection control unit 14 that controls the injection rate of the flocculant of the chemical injection equipment 25 based on the optimum flocculant injection rate determined as a result of repeating the flocculant treatment test until the aggregation treatment test satisfies the reference value.

According to the chemical injection control device 10 according to the embodiment of the present invention, the prediction result of the chemical injection rate based on machine learning can be verified using a simple test device, so that abnormal weather events such as torrential rain can be avoided. Even if this occurs and there is a risk that the predicted result will deviate significantly, the predicted result can be appropriately corrected and chemicals can be injected into the raw water at an appropriate injection rate. As a result, the amount of flocculant used can be optimized, efficient coagulation-sedimentation treatment can be performed, and purified water with good quality can be obtained.

1... Treatment tank 2... Stirrer 3... Chemical injection part 4... pH injection part 5... Water quality analyzer 6... Control part 7... Inflow valve 8... Drain valve 10... Chemical injection control device 11... Information acquisition part 12... Prediction part 13... Confirmation unit 14... Injection control unit 20... Water treatment plant 21... Water receiving well 22... Rapid stirring pond 23... Floc formation pond 24... Sedimentation basin 25... Chemical injection equipment 26... Rapid filtration basin 27... Water purification basin 28... Water distribution basin 31... Chemical injection pump 41... pH adjustment pump 51... Sampling pump 100... Chemical injection control system

Claims (7)

an information acquisition unit that acquires information including water quality data of raw water and operation condition data of a water treatment plant that processes the raw water;
Using a machine learning algorithm using a trained model that acquires explanatory variables including the water quality data and the operation condition data and outputs objective variables including the optimal injection rate of chemicals to be injected into the raw water, a prediction unit that predicts the injection rate of the chemical to be injected into raw water;
Based on the prediction result of the injection rate of the chemical, the chemical is injected into the raw water and a confirmation test is performed, and the water quality data of the treated water obtained from the confirmation test is compared with the reference value, so that the predicted result is confirmed. a confirmation section that confirms whether the
A chemical injection control system comprising: an injection control section that controls an injection rate of the chemical to be injected into the raw water based on a confirmation result of the confirmation section. The confirmation section
If the predicted result of the injection rate of the chemical is appropriate, determine the injection rate based on the predicted result as the injection rate of the chemical for injecting the chemical into the raw water to perform water treatment,
If the prediction result of the injection rate of the chemical is not appropriate, change the injection rate and retest, and based on the retest, determine the optimum method for injecting the chemical into the raw water for water treatment. The chemical injection control system according to claim 1, further comprising: determining an injection rate of the chemical. The confirmation section
a treatment tank for treating the raw water;
a chemical injection unit that injects the chemical into the processing tank;
a stirrer that stirs the inside of the processing tank;
a water quality analyzer that analyzes water quality data of the treated water obtained in the treatment tank;
The chemical injection control system according to claim 1 or 2, further comprising: a control unit that controls an injection rate of the chemical to be injected into the processing tank. A trained model is used that acquires explanatory variables including water quality data of raw water and operational condition data of a water treatment plant that processes the raw water, and outputs objective variables including the optimal injection rate of chemicals to be injected into the raw water. predicting the injection rate of the chemical to be injected into the raw water using a machine learning algorithm;
Based on the prediction result of the injection rate of the chemical, the chemical is injected into the raw water and a confirmation test is performed, and the water quality data of the treated water obtained from the confirmation test is compared with the standard value. a step of confirming whether the prediction result of the rate is appropriate;
A chemical injection control method comprising: controlling an injection rate of the chemical to be injected into the raw water based on the result of the confirmation. The step of confirming is
If the predicted result of the injection rate of the chemical is appropriate, determine the injection rate based on the predicted result as the injection rate of the chemical for injecting the chemical into the raw water to perform water treatment,
If the predicted result of the injection rate of the chemical is not appropriate, the injection rate is changed and the test is performed again. 5. The chemical injection control method according to claim 4, further comprising repeating retesting to determine an optimum injection rate of the chemical for injecting the chemical into the raw water to perform water treatment. The step of confirming is
Injecting a flocculant as the chemical into the raw water to perform a coagulation sedimentation treatment,
The water quality data of the treated water obtained by the coagulation-sedimentation treatment includes analyzing at least one of turbidity, chromaticity, pH, alkalinity, number of fine particles, TOC, COD, ultraviolet absorbance, and organic matter fraction. The drug injection control method according to claim 4 or 5. an information acquisition unit that acquires a prediction result of the injection rate of the flocculant predicted by a machine learning algorithm using a learned model for predicting the optimal injection rate of the flocculant to be injected into raw water;
A treatment tank for conducting a flocculation treatment test including a raw water collection step, a chemical injection step, a stirring step, and a standing step based on the predicted result of the injection rate of the flocculant and preset test conditions;
a water quality analyzer that analyzes water quality data of the treated water obtained in the flocculation treatment test;
a confirmation unit that compares the water quality data of the treated water analyzed by the water quality analyzer with a reference value and determines whether the predicted result is appropriate;
If the prediction result is appropriate, the injection rate of the flocculant of the chemical injection equipment for injecting the flocculant into the raw water to perform water treatment is controlled based on the injection rate based on the prediction result, and the injection rate of the flocculant is controlled based on the injection rate based on the prediction result. If the result is not appropriate, carry out the flocculation treatment test with a new flocculant injection rate, and repeat the flocculation treatment test until the water quality data of the treated water obtained in the flocculation treatment test satisfies the standard value. An automatic flocculation processing apparatus comprising: an injection control unit that controls the injection rate of the flocculant of the chemical injection equipment based on the determined optimal flocculant injection rate.