Monitoring and control system, monitoring and control method, and monitoring and control program

Abstract

To obtain a monitoring control system that can easily operate a plant stably.SOLUTION: A monitoring control system 1 includes: a data acquisition unit 31 that acquires water quality impact conditions that are information on conditions that affect a water quality of water to be treated when chlorine is injected into the water to be treated in a water treatment plant; an inference device 30 that infers a chlorine injection amount from the water quality impact conditions acquired by the data acquisition unit 31 using a trained model for inferring the chlorine injection amount which indicates the chlorine injection amount when the water quality impact conditions are met; and an output device 35 that outputs the chlorine injection amount inferred by the inference device 30.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a monitoring and control system, a monitoring and control method, and a monitoring and control program for a water treatment plant. [Background technology]

[0002] In water treatment plants, it is desirable to easily perform stable water quality management. However, in plants such as water treatment plants, when the plant conditions change, it becomes difficult to perform stable treatment.

[0003] For example, the operation monitoring device described in Patent Document 1 generates and displays on a display unit one of a diagram, graph, or chart based on measurement data at the installation location or measurement location of a selected measurement device from among measurement data from multiple measurement devices in water treatment. This operation monitoring device generates one of a diagram, graph, or chart from the measurement data by performing a simulation using pre-registered calculation formulas and parameters. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2019-10613 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the technology of Patent Document 1 merely displays the measurement data on a display unit, and does not allow accurate control variables for the plant to be inferred, which makes it difficult to operate the plant stably.

[0006] The present disclosure has been made in view of the above, and aims to provide a monitoring and control system that can easily operate a plant stably. [Means for solving the problem]

[0007] To solve the above-mentioned problems and achieve the object, the monitoring and control system of the present disclosure includes a first data acquisition device that acquires water quality impact conditions, which are information on conditions that affect the water quality of water being treated when chlorine is injected into the water treatment plant. The monitoring and control system of the present disclosure also includes an inference device that infers the amount of chlorine to be injected from the water quality impact conditions acquired by the first data acquisition device using a trained model for inferring the amount of chlorine to be injected when the water quality impact conditions are satisfied, and an output device that outputs the amount of chlorine to be injected inferred by the inference device. The water quality impact conditions include a target residual chlorine concentration, which is a target value for the concentration of residual chlorine contained in tap water provided from a water treatment plant, the amount of water to be treated flowing into the treatment tank of the water treatment plant, and a chlorine demand, which is the amount of chlorine to be injected into the water to be treated until residual chlorine begins to be detected in the water to be treated when chlorine is injected into the water to be treated. [Effects of the Invention]

[0008] The monitoring and control system according to the present disclosure has the effect of easily operating a plant stably. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram illustrating a configuration of a monitoring and control system according to an embodiment. [Figure 2] FIG. 1 is a diagram illustrating a configuration of a learning device included in a monitoring and control system according to an embodiment. [Figure 3] FIG. 1 is a diagram for explaining a neural network used by a learning device according to an embodiment. [Figure 4] 1 is a flowchart showing a procedure of a learning process executed by a learning device according to an embodiment; [Figure 5] FIG. 1 is a diagram illustrating a configuration of an inference device included in a monitoring and control system according to an embodiment. [Figure 6] 1 is a flowchart showing a procedure of an inference process executed by an inference device according to an embodiment; [Figure 7] FIG. 1 is a diagram showing an example of the configuration of a processing circuit provided in a monitoring and control system according to an embodiment when the processing circuit is realized by a processor and a memory. [Figure 8]FIG. 1 is a diagram showing an example of a processing circuit when the processing circuit provided in the water treatment control support system according to the embodiment is configured with dedicated hardware. DETAILED DESCRIPTION OF THE INVENTION

[0010] A monitoring and control system, a monitoring and control method, and a monitoring and control program according to embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0011] Embodiment 1 is a diagram showing the configuration of a monitoring and control system according to an embodiment. The monitoring and control system 1 is a system that monitors the operation of a water treatment plant used in a water purification plant. The monitoring and control system 1 calculates the amount of chlorine to be injected so that the chlorine concentration (target residual chlorine concentration, described later) of tap water provided to ordinary households by the water treatment plant (water treatment facility) falls within a standard range, and provides the calculated amount to an operator who operates the water treatment plant.

[0012] The monitoring and control system 1 uses artificial intelligence (AI) to calculate the amount of chlorine to be injected into raw water, which is the water to be treated. Note that sodium hypochlorite is injected as chlorine into the water to be treated.

[0013] The monitoring and control system 1 calculates the amount of chlorine to be injected into the water to be treated (hereinafter referred to as the amount of chlorine to be injected) based on the amount of water to be treated flowing into the treatment tank of the water treatment plant (hereinafter referred to as the amount of inflow water) and the target value of the concentration of residual chlorine contained in the tap water that the water treatment plant provides to ordinary households (hereinafter referred to as the target residual chlorine concentration). The target residual chlorine concentration is determined by law, etc., and chlorine is injected so that the target residual chlorine concentration falls within the standard range determined by this law, etc.

[0014] The inflow rate is expressed as the amount of water to be treated per unit time, for example. The target residual chlorine concentration is expressed as parts per million (PPM), for example. The amount of chlorine to be injected is expressed as the weight of chlorine, for example.

[0015] The monitoring and control system 1 includes a data acquisition unit 31, which is a data acquisition device, an inference device 30, an output device 35, a display device 41, a chlorine amount control device 42, a learning device 10, and a trained model storage unit 20. The data acquisition unit 31 is a first data acquisition device. The data acquisition unit 31, the inference device 30, the output device 35, the display device 41, the chlorine amount control device 42, the learning device 10, and the trained model storage unit 20 may be connected via a network or the like, or may be realized as a single computer.

[0016] The data acquisition unit 31 acquires information on conditions that affect water quality (water quality affecting conditions, which will be described later) and sends it to the inference device 30. Examples of the water quality affecting conditions are the inflow water volume and the target residual chlorine concentration.

[0017] The inference device 30 infers the amount of chlorine to be injected from the water quality impact conditions using a trained model that infers the amount of chlorine to be injected corresponding to the water quality impact conditions. That is, the inference device 30 infers the amount of chlorine to be injected by inputting the water quality impact conditions into the trained model.

[0018] The inference device 30 sends the inferred amount of chlorine to the output device 35. The output device 35 outputs the amount of chlorine to be injected to the display device 41 and the chlorine amount control device 42. The display device 41 is a display device such as a liquid crystal monitor that displays the amount of chlorine to be injected.

[0019] The chlorine amount control device 42 is a device that controls the amount of chlorine to be injected. The chlorine amount control device 42 causes the chlorine injector 43 to inject an amount of chlorine corresponding to the amount of chlorine to be injected. The chlorine injector 43 injects chlorine (hypochlorous acid water) into the treatment tank according to instructions from the chlorine amount control device 42.

[0020] The learning device 10 generates a trained model by learning the amount of chlorine to be injected corresponding to the water quality impact conditions. The trained model is a model for inferring an appropriate amount of chlorine to be injected corresponding to the water quality impact conditions. The learning device 10 generates the trained model using the amount of chlorine to be injected, which is information on the amount of chlorine actually injected by the chlorine injector 43.

[0021] The trained model storage unit 20 stores the trained model generated by the learning device 10. The trained model stored in the trained model storage unit 20 is read by the inference device 30 when the inference device 30 infers the amount of chlorine to be injected.

[0022] The learning device 10 and the inference device 30 may be devices separate from the monitoring and control system 1, for example, connected to the monitoring and control system 1 via a network. The learning device 10 and the inference device 30 may also be built into the monitoring and control system 1. The learning device 10 and the inference device 30 may also exist on a cloud server. The learning device 10 and the inference device 30 may also be realized by different computers, or the learning device 10 and the inference device 30 may also be realized by a single computer.

[0023] Furthermore, at least one of the learning device 10, the trained model storage unit 20, the display device 41, the chlorine amount control device 42, and the chlorine injector 43 may be located outside the monitoring and control system 1. In other words, at least one of the learning device 10, the trained model storage unit 20, the display device 41, the chlorine amount control device 42, and the chlorine injector 43 does not have to be a component of the monitoring and control system 1.

[0024] Furthermore, at least one of the data acquisition unit 31 and the output device 35 may be located within the inference device 30.

[0025] 2 is a diagram showing the configuration of a learning device provided in a monitoring and control system according to an embodiment. Learning device 10 has a data acquisition unit 11 and a model generation unit 12. Data acquisition unit 11 acquires water quality impact conditions 51A from outside learning device 10. Data acquisition unit 11 also acquires injection chlorine amount 52A from outside learning device 10. Data acquisition unit 11 is a second data acquisition device.

[0026] The water quality impact conditions 51A acquired by the data acquisition unit 11 include the inflow water volume and the target residual chlorine concentration. The water quality impact conditions 51A acquired by the data acquisition unit 11 may also include information indicating conditions that affect the water quality of the water to be treated, such as the chlorine demand, water temperature, day of the week, and time of day. That is, the water quality impact conditions 51A may include at least one of the chlorine demand, water temperature, day of the week, and time of day. The conditions that affect the water quality of the water to be treated are conditions that affect the chlorine concentration of the water to be treated when chlorine is injected into the water to be treated.

[0027] The chlorine demand is the amount of chlorine injected during water treatment until all oxidized species in the water (water to be treated) are oxidized and residual chlorine begins to be detected. If chlorine is injected into the water to be treated after the chlorine demand has been reached, the residual chlorine will increase. The water temperature is the temperature of the water to be treated in the treatment tank. The day of the week is the day on which chlorine is injected. The time is the time period during which chlorine is injected.

[0028] Data acquiring unit 11 acquires, for example, the amount of inflow water from a water meter and acquires the amount of chlorine to be injected 52A from chlorine amount control device 42. Chlorine amount control device 42 accepts the amount of chlorine to be injected 52A input by the operator. Chlorine amount control device 42 sends a command to chlorinator 43 to inject chlorine in the amount of chlorine to be injected 52A input by the operator.

[0029] When the inference device 30 is not used, the amount of chlorine to be injected 52A input by the operator to the chlorine amount control device 42 is, for example, an amount of chlorine to be injected determined by the operator based on experience. For example, the operator determines the amount of chlorine to be injected 52A based on the influent water flow rate and the target residual chlorine concentration so that the residual chlorine concentration measured in the water to be treated (measured chlorine concentration value) falls within a reference range. The operator may also determine the amount of chlorine to be injected 52A with reference to at least one of the chlorine demand, water temperature, day of the week, and time of day.

[0030] When the learning device 10 generates the trained model 60, an injection amount of chlorine 52A, which indicates the amount of chlorine actually injected and determined by the operator, is input to the data acquisition unit 11. Furthermore, when the learning device 10 generates the trained model 60, a water quality impact condition 51A, which is information referenced when determining the injection amount of chlorine 52A, is input to the data acquisition unit 11. The injection amount of chlorine 52A and the water quality impact condition 51A are input to the data acquisition unit 11 when the measured concentration of chlorine remaining in the water to be treated when the operator actually injected chlorine is within the reference range. That is, the injection amount of chlorine 52A input to the data acquisition unit 11 is the correct injection amount of chlorine for the water quality impact condition 51A, in which the measured concentration is within the reference range. The data acquisition unit 11 sends the acquired injection amount of chlorine 52A and the water quality impact condition 51A to the model generation unit 12.

[0031] The model generation unit 12 learns an appropriate chlorine injection amount 52A corresponding to the water quality impacting condition 51A based on training data created based on a combination of the water quality impacting condition 51A and the chlorine injection amount 52A sent from the data acquisition unit 11. In other words, the model generation unit 12 learns an appropriate chlorine injection amount 52A when the water quality impacting condition 51A is satisfied based on training data created based on a combination of the water quality impacting condition 51A and the chlorine injection amount 52A. That is, the model generation unit 12 generates a trained model 60 that infers an appropriate chlorine injection amount 52A from the water quality impacting condition 51A that affects water quality, such as the inflow water volume, target residual chlorine concentration, chlorine demand, water temperature, day of the week, and time of day. Here, the training data is data in which the water quality impacting condition 51A and the chlorine injection amount 52A are associated with each other.

[0032] The model generation unit 12 can use known algorithms such as supervised learning, unsupervised learning, reinforcement learning, etc. As an example, a case where a neural network is applied to the learning algorithm used by the model generation unit 12 will be described.

[0033] The model generation unit 12 learns an appropriate amount of chlorine to be injected 52A corresponding to the water quality impact condition 51A, for example, by so-called supervised learning according to a neural network model. Here, supervised learning refers to a technique in which data sets (learning data) of inputs and results (labels) are provided to the learning device 10, and the learning device 10 learns the features contained in the learning data and infers the results from the inputs.

[0034] A neural network consists of an input layer consisting of multiple neurons, an intermediate layer (hidden layer) consisting of multiple neurons, and an output layer consisting of multiple neurons. The intermediate layer may be one layer or two or more layers.

[0035] 3 is a diagram illustrating a neural network used by a learning device according to an embodiment. For example, in a three-layer neural network as shown in FIG. 3, when multiple inputs are input to the input layer (X1 to X3), the values ​​are multiplied by weights W1 (w11 to w16) and input to the intermediate layer (Y1 to Y2). The results are then further multiplied by weights W2 (w21 to w26) and output from the output layer (Z1 to Z3). This output result varies depending on the values ​​of weights W1 and W2.

[0036] 2 uses supervised learning to learn the amount of chlorine to be injected 52A corresponding to the water quality impacting condition 51A, in accordance with learning data created based on a combination of water quality impacting conditions 51A that affect water quality, such as inflow water volume, target residual chlorine concentration, chlorine demand, water temperature, day of the week, and time of day, and the amount of chlorine to be injected 52A, all of which are acquired by the data acquiring unit 11. In other words, the neural network used by the learning device 10 of FIG. 2 uses supervised learning to learn the amount of chlorine to be injected 52A corresponding to the water quality impacting condition 51A, in accordance with the water quality impacting conditions 51A and the amount of chlorine to be injected 52A, which are created based on a combination of a first input and a second input (correct answer) acquired by the data acquiring unit 11.

[0037] That is, the neural network learns by inputting the water quality influence condition 51A as the first input and adjusting the weights W1 and W2 so that the result output from the output layer approaches the second input (correct answer).

[0038] In this way, the neural network learns by inputting conditions that affect water quality, such as inflow water volume, target residual chlorine concentration, chlorine demand, water temperature, day of the week, and time of day, into the input layer and adjusting weights W1 and W2 so that the results output from the output layer approach chlorine dosage 52A. By learning the correspondence between water quality impacting conditions 51A and chlorine dosage 52A, the neural network generates trained model 60 that can output an appropriate chlorine dosage 52A when water quality impacting conditions 51A are input. In this way, learning device 10 learns trained model 60 that can output the correct chlorine dosage 52A when water quality impacting conditions 51A are input.

[0039] By performing the above-described learning, the model generation unit 12 generates and outputs a trained model 60. The trained model storage unit 20 stores the trained model 60 output from the model generation unit 12.

[0040] Next, a processing procedure of the process in which the learning device 10 learns the trained model 60 will be described with reference to Fig. 4. Fig. 4 is a flowchart showing the processing procedure of the learning process executed by the learning device according to the embodiment.

[0041] The data acquisition unit 11 acquires learning data to be used for learning (step S1). Specifically, the data acquisition unit 11 acquires water quality influence conditions 51A that affect water quality, such as inflow water volume, target residual chlorine concentration, chlorine demand, water temperature, day of the week, time of day, etc., and an injection chlorine amount 52A.

[0042] The data acquisition unit 11 simultaneously acquires the chlorine injection amount 52A, inflow water flow rate, target residual chlorine concentration, chlorine demand, water temperature, day of the week, time period, etc. However, the chlorine injection amount 52A, inflow water flow rate, target residual chlorine concentration, chlorine demand, water temperature, day of the week, time period, etc. may be input in association with each other. Therefore, the data acquisition unit 11 may acquire the chlorine injection amount 52A, inflow water flow rate, target residual chlorine concentration, chlorine demand, water temperature, day of the week, time period, etc. at different times. The data acquisition unit 11 sends the water quality influence conditions 51A and the chlorine injection amount 52A to the model generation unit 12.

[0043] Model generation unit 12 executes a learning process using water quality influence conditions 51A and injection chlorine amount 52A (step S2). Specifically, model generation unit 12 learns injection chlorine amount 52A corresponding to water quality influence conditions 51A by so-called supervised learning in accordance with learning data created based on combinations of water quality influence conditions 51A that affect water quality, such as inflow water volume, target residual chlorine concentration, chlorine demand, water temperature, day of the week, and time of day, acquired by data acquisition unit 11, and injection chlorine amount 52A, and generates trained model 60.

[0044] After generating the trained model 60, the model generation unit 12 outputs the trained model 60 to the trained model storage unit 20 (step S3). The trained model storage unit 20 stores the trained model 60 generated by the model generation unit 12.

[0045] 5 is a diagram showing the configuration of an inference device provided in the monitoring and control system according to the embodiment. The inference device 30 includes a data acquisition unit 31 and an inference unit 32. Note that the data acquisition unit 31 may be disposed outside the inference device 30, as described in FIG. 1.

[0046] The data acquisition unit 31 acquires water quality impact conditions 51B from outside the inference device 30. The water quality impact conditions 51B are the same information as the water quality impact conditions 51A. That is, the water quality impact conditions 51B are information indicating conditions that affect water quality, such as the inflow water volume, target residual chlorine concentration, chlorine demand, water temperature, day of the week, and time of day.

[0047] The data acquisition unit 31 acquires the water quality impact conditions 51B in the same manner as the data acquisition unit 11. That is, the data acquisition unit 31 acquires the inflow water volume from a water meter, for example. The data acquisition unit 11 sends the acquired water quality impact conditions 51B to the inference unit 32.

[0048] The inference unit 32 receives the water quality impact conditions 51B sent from the data acquisition unit 31. The inference unit 32 also reads out the learned model 60 from the learned model storage unit 20. The inference unit 32 uses the learned model 60 to infer an injection amount of chlorine 52B corresponding to the water quality impact conditions 51B. That is, by inputting the water quality impact conditions 51B that affect water quality, such as the inflow water volume, target residual chlorine concentration, chlorine demand, water temperature, day of the week, and time of day acquired by the data acquisition unit 31, into the learned model 60, the inference unit 32 can output an appropriate injection amount of chlorine 52B inferred from the water quality impact conditions 51B that affect water quality, such as the inflow water volume, target residual chlorine concentration, chlorine demand, water temperature, day of the week, and time of day.

[0049] It is sufficient that the water quality influence conditions 51B acquired by the data acquisition unit 31 include at least the inflow water volume and the target residual chlorine concentration. In other words, it is sufficient that the water quality influence conditions 51B used by the inference unit 32 when inferring the injection chlorine amount 52B include at least the inflow water volume and the target residual chlorine concentration.

[0050] Furthermore, in the present embodiment, the case has been described in which the inference device 30 outputs the appropriate amount of chlorine to be injected 52B using the trained model 60 trained by the model generation unit 12 of the monitoring and control system 1, but the inference device 30 may acquire the trained model 60 from an external source such as another monitoring and control system 1. In this case, the inference device 30 outputs the appropriate amount of chlorine to be injected 52B based on the trained model 60 acquired from the other monitoring and control system 1 or the like.

[0051] Next, a processing procedure of the process in which the inference device 30 infers the amount of injected chlorine 52B using the trained model 60 will be described with reference to Fig. 6. Fig. 6 is a flowchart showing the processing procedure of the inference process executed by the inference device according to the embodiment.

[0052] The data acquisition unit 31 acquires inference data used to infer the amount of chlorine to be injected 52B (step S11). Specifically, the data acquisition unit 31 acquires water quality influence conditions 51B that affect water quality, such as the inflow water volume, target residual chlorine concentration, chlorine demand, water temperature, day of the week, and time of day. The data acquisition unit 31 sends the water quality influence conditions 51B to the inference unit 32. The inference unit 32 acquires the water quality influence conditions 51B from the data acquisition unit 31 and acquires the trained model 60 from the trained model storage unit 20.

[0053] The inference unit 32 inputs water quality influence conditions 51B that affect water quality, such as inflow water volume, target residual chlorine concentration, chlorine demand, water temperature, day of the week, and time of day, into the trained model 60 (step S12), and obtains an appropriate injection amount of chlorine 52B.

[0054] The inference unit 32 outputs data inferred using the trained model 60 and the water quality impact condition 51B (step S13). Specifically, the inference unit 32 outputs the appropriate amount of chlorine to be injected 52B obtained by the trained model 60 to the display device 41 and the chlorine amount control device 42.

[0055] Display device 41 displays injection amount 52B corresponding to water quality impact condition 51B (step S14), which allows the operator to refer to injection amount 52B corresponding to water quality impact condition 51B.

[0056] The monitoring and control system 1 can provide operational support that is close to the actual conditions of water treatment by displaying the amount of chlorine to be injected 52B on the display device 41. This allows the monitoring and control system 1 to provide efficient education and training to operators with little work experience, enabling efficient transfer of skills.

[0057] The operator refers to the chlorine injection amount 52B corresponding to the water quality impact condition 51B and inputs an injection instruction specifying the amount of chlorine to be injected into the water to be treated (hereinafter referred to as a chlorine injection instruction) to the chlorine amount control device 42. The operator inputs the chlorine injection instruction corresponding to the chlorine injection amount 52B to the chlorine amount control device 42, for example, by inputting an injection permission for the chlorine injection amount 52B sent from the inference device 30.

[0058] The chlorine amount control device 42 generates a chlorine injection command corresponding to the chlorine injection instruction and outputs it to the chlorine injector 43. Note that the chlorine amount control device 42 may generate a chlorine injection command corresponding to the injection chlorine amount 52B and output it to the chlorine injector 43 without accepting the chlorine injection instruction.

[0059] Furthermore, when an instruction to correct the amount of chlorine is input from the operator, the chlorine amount control device 42 may correct the amount of chlorine to be injected 52B. That is, the chlorine amount control device 42 may generate an injection command in which the amount of chlorine to be injected 52B inferred by the inference unit 32 is corrected, and output the corrected injection command to the chlorine injector 43.

[0060] In addition, when an operator inputs a chlorine injection instruction determined by his / her own judgment without adopting the injection chlorine amount 52B, the chlorine amount control device 42 may output an injection command corresponding to the chlorine injection instruction input by the operator to the chlorine injector 43.

[0061] Chlorine amount control device 42 may input the amount of chlorine to be injected corresponding to the injection command output to chlorinator 43 and the water quality impact condition corresponding to this amount of chlorine to learning device 10 as injection chlorine amount 52A and water quality impact condition 51A. In other words, chlorine amount control device 42 may input the amount of chlorine to be injected corresponding to the injection command output to chlorinator 43 and the water quality impact condition corresponding to this amount of chlorine to learning device 10 as learning data.

[0062] The concentration of residual chlorine in the water to be treated when chlorine injector 43 injects chlorine corresponding to the injection command into the water to be treated (hereinafter referred to as the measured residual chlorine concentration) is measured by a chlorine concentration meter, which is a device for measuring chlorine concentration. After it is confirmed that the measured residual chlorine concentration measured by the chlorine concentration meter is within the standard range, chlorine amount control device 42 inputs the amount of chlorine to be injected corresponding to the injection command output to chlorine injector 43 and the water quality impact conditions corresponding to this amount of chlorine to learning device 10 as learning data. In other words, chlorine amount control device 42 inputs the amount of chlorine to be injected that is confirmed to be the correct answer to the water quality impact conditions and the water quality impact conditions corresponding to this amount of chlorine to learning device 10 as learning data.

[0063] The amount of chlorine to be injected corresponding to the injection command output by the chlorine amount control device 42 to the chlorine injector 43 may be the amount of chlorine to be injected 52B received from the inference device 30, or may be an amount of chlorine to be injected corrected by the operator. That is, the chlorine amount control device 42 may input the amount of chlorine to be injected 52B received from the inference device 30 to the learning device 10 as the amount of chlorine to be injected 52A, or may input the amount of chlorine to be injected corrected by the operator for the amount of chlorine to be injected 52B received from the inference device 30 to the learning device 10 as the amount of chlorine to be injected 52A.

[0064] Furthermore, the amount of chlorine to be injected corresponding to the injection command output by the chlorine amount control device 42 to the chlorine injector 43 may be an amount of chlorine to be injected that is independently designated by the operator. That is, the chlorine amount control device 42 may input the amount of chlorine to be injected that is independently designated by the operator as a chlorine injection instruction to the learning device 10 as the amount of chlorine to be injected 52A.

[0065] The water quality impact condition that the chlorine amount control device 42 inputs to the learning device 10 is water quality impact condition 51B. Note that the inference device 30 may input water quality impact condition 51B to the learning device 10. The learning device 10 executes a learning process using the learning data input from the chlorine amount control device 42.

[0066] The learning data acquired from the chlorine amount control device 42 may be stored in the monitoring control system 1 or in a device external to the monitoring control system 1. In this case, the learning device 10 executes the learning process using the stored learning data.

[0067] Furthermore, learning device 10 may perform learning processing in real time while chlorine amount control processing is being performed by chlorine amount control device 42. In this case, when chlorine amount control device 42 outputs an injection command to chlorine injector 43, it inputs to learning device 10 the amount of chlorine to be injected corresponding to this injection command and the water quality impact condition corresponding to this amount of chlorine to be injected.

[0068] Furthermore, if the measured residual chlorine concentration measured by the chlorine concentration measuring device is within a reference range, the learning device 10 may include the residual chlorine concentration measured by the chlorine concentration measuring device in the water quality impact conditions 51A. In this case, the data acquisition unit 11 acquires the water quality impact conditions 51A including the measured residual chlorine concentration and the amount of chlorine to be injected 52A as learning data. Then, the model generation unit 12 generates a trained model 60 for calculating the appropriate amount of chlorine to be injected 52A for the water quality impact conditions 51A including the measured residual chlorine concentration. The inference device 30 uses the trained model 60 and the water quality impact conditions 51B including the measured residual chlorine concentration to calculate the amount of chlorine to be injected 52B corresponding to the water quality impact conditions 51B.

[0069] The chlorine amount control device 42 may also perform feedback control using the measured residual chlorine concentration measured by the chlorine concentration measuring device. That is, the chlorine amount control device 42 may adjust the amount of chlorine to be injected so that the measured residual chlorine concentration measured by the chlorine concentration measuring device approaches the target residual chlorine concentration.

[0070] There is a method for calculating the amount of chlorine to be injected using a specific formula. The parameter constants used in this formula are pre-registered. However, the quality of the water to be treated and the amount of inflow water vary depending on weather conditions, the day of the week, the time of day, the water temperature, the air temperature, and other factors. Furthermore, the quality of the water to be treated varies depending on the time lag between the injection of chlorine and the reaction. This makes it difficult to select the formula and parameters, and it is somewhat dependent on the operator's experience. For example, if an accurate formula or parameters are not selected, the calculated amount of chlorine to be injected may differ from the actual water quality, making it impossible to determine whether the operational support is functioning correctly.

[0071] On the other hand, in this embodiment, learning device 10 generates trained model 60 based on water quality impact conditions 51A and chlorine injection amount 52A, and inference device 30 calculates appropriate chlorine injection amount 52B based on trained model 60 and water quality impact conditions 51B. Display device 41 then displays appropriate chlorine injection amount 52B. This allows an operator to determine a chlorine injection command by referring to the appropriate chlorine injection amount 52B displayed on display device 41, and monitoring and control system 1 can adjust the chlorine concentration of the water to be treated according to the determined injection command. Therefore, monitoring and control system 1 can support stable water quality management.

[0072] In this embodiment, an example has been described in which the model generation unit 12 uses a supervised learning algorithm as a learning algorithm, but the learning algorithm used by the model generation unit 12 is not limited to a supervised learning algorithm. The model generation unit 12 can also apply a reinforcement learning algorithm, an unsupervised learning algorithm, a semi-supervised learning algorithm, etc., in addition to the supervised learning algorithm.

[0073] Furthermore, the model generation unit 12 may learn an appropriate amount of chlorine to be injected 52A corresponding to the water quality impact condition 51A in accordance with learning data created by multiple monitoring and control systems 1. Furthermore, the model generation unit 12 may perform learning using learning data acquired from multiple monitoring and control systems 1 used in the same area, or may perform learning using learning data acquired from multiple monitoring and control systems 1 operating independently in different areas.

[0074] In addition, a monitoring and control system 1 that collects learning data may be added to or removed from the targets during the process. Furthermore, a learning device 10 that has learned an appropriate amount of chlorine to be injected 52A corresponding to a water quality impact condition 51A for a certain monitoring and control system 1 may be applied to another monitoring and control system 1, and the appropriate amount of chlorine to be injected 52A corresponding to the water quality impact condition 51A for the other monitoring and control system 1 may be re-learned and updated.

[0075] Deep learning, which learns to extract features themselves, can also be used as the learning algorithm of the model generation unit 12. The model generation unit 12 may also perform machine learning according to other known methods, such as genetic programming, functional logic programming, or support vector machines.

[0076] Next, a description will be given of the hardware configuration of the monitoring and control system 1. Here, a description will be given of the hardware configuration of the monitoring and control system 1 when the monitoring and control system 1 includes a data acquisition unit 31, an inference device 30, and an output device 35.

[0077] The monitoring and control system 1 is realized by a processing circuit. This processing circuit may be a processor and memory that executes a program stored in memory, or may be dedicated hardware. The processing circuit is also called a control circuit.

[0078] FIG. 7 is a diagram illustrating an example of the configuration of a processing circuit provided in a monitoring and control system according to an embodiment, where the processing circuit is implemented by a processor and a memory. The processing circuit 90 illustrated in FIG. 7 is a control circuit and includes a processor 91 and a memory 92. When the processing circuit 90 is configured with the processor 91 and the memory 92, each function of the processing circuit 90 is implemented by software, firmware, or a combination of software and firmware. The software or firmware is written as a program and stored in the memory 92. The processor 91 reads and executes the program stored in the memory 92 to implement each function of the processing circuit 90. That is, the processing circuit 90 includes the memory 92 for storing a monitoring and control program that results in the processing of the monitoring and control system 1. This monitoring and control program can also be considered a program that causes the monitoring and control system 1 to execute each function implemented by the processing circuit 90. This monitoring and control program may be provided by a storage medium on which the program is stored, or by other means such as a communication medium.

[0079] The monitoring and control program executed by the monitoring and control system 1 has a modular configuration including a data acquisition unit 31, an inference device 30, and an output device 35, which are loaded onto a main memory device and generated on the main memory device. Note that, when the monitoring and control system 1 is equipped with a learning device 10 and a chlorine amount control device 42, the monitoring and control program has a modular configuration including the learning device 10 and the chlorine amount control device 42, which are loaded onto a main memory device and generated on the main memory device.

[0080] Here, the processor 91 is, for example, a CPU (Central Processing Unit), a processing device, an arithmetic device, a microprocessor, a microcomputer, or a DSP (Digital Signal Processor), etc. Furthermore, the memory 92 is, for example, a non-volatile or volatile semiconductor memory such as a RAM (Random Access Memory), a ROM (Read Only Memory), a flash memory, an EPROM (Erasable Programmable ROM), or an EEPROM (registered trademark) (Electrically EPROM), a magnetic disk, a flexible disk, an optical disk, a compact disk, a minidisk, or a DVD (Digital Versatile Disc).

[0081] Fig. 8 is a diagram showing an example of a processing circuit provided in a water treatment control support system according to an embodiment, configured with dedicated hardware. The processing circuit 93 shown in Fig. 8 corresponds to, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a combination thereof.

[0082] The processing circuits 90 and 93 may be partially implemented by dedicated hardware and partially implemented by software or firmware. In this way, the processing circuits 90 and 93 can realize the above-described functions by dedicated hardware, software, firmware, or a combination of these.

[0083] The monitoring and control system 1 may be realized by one processing circuit or by multiple processing circuits. When the learning device 10 is realized by one processing circuit, the learning device 10 has the same hardware configuration as the monitoring and control system 1 described in Figures 7 and 8. When the chlorine amount control device 42 is realized by one processing circuit, the chlorine amount control device 42 has the same hardware configuration as the monitoring and control system 1 described in Figures 7 and 8.

[0084] As described above, according to the embodiment, monitoring and control system 1 uses trained model 60 to infer amount of chlorine to be injected 52B from water quality impact conditions 51B acquired by data acquisition unit 31, making it possible to provide an operator with an appropriate amount of chlorine to be injected 52B that corresponds to water quality impact conditions 51B. This allows the operator to determine the amount of chlorine to be injected into the water to be treated by referring to the provided amount of chlorine to be injected 52B, making it possible for monitoring and control system 1 to operate the water treatment plant stably.

[0085] Furthermore, the learning device 10 performs further learning using the amount of chlorine injected 52A corrected or determined by the operator and the water quality impact condition 51A corresponding to this amount of chlorine injected 52A, thereby making it possible to generate a learned model 60 that can output a more accurate amount of chlorine injected 52B.

[0086] Furthermore, if the measured residual chlorine concentration measured by the chlorine concentration meter is within the standard range, the learning device 10 includes the residual chlorine concentration measured by the chlorine concentration meter in the water quality impact condition 51A, making it possible to generate a trained model 60 that can output a more accurate injection chlorine amount 52B.

[0087] The configurations shown in the above embodiments are merely examples, and may be combined with other known technologies, and parts of the configurations may be omitted or modified without departing from the spirit of the invention. [Explanation of symbols]

[0088] 1 Monitoring and control system, 10 Learning device, 11, 31 Data acquisition unit, 12 Model generation unit, 20 Trained model memory unit, 30 Inference device, 32 Inference unit, 35 Output device, 41 Display device, 42 Chlorine amount control device, 43 Chlorinator, 51A, 51B Water quality impact conditions, 52A, 52B Injected chlorine amount, 60 Trained model, 90, 93 Processing circuit, 91 Processor, 92 Memory.

Claims

1. a first data acquisition device that acquires water quality influence conditions, which are information on conditions that affect the water quality of water to be treated when chlorine is injected into the water to be treated in a water treatment plant; an inference device that infers the amount of chlorine to be injected from the water quality impact condition acquired by the first data acquisition device using a trained model for inferring the amount of chlorine to be injected, which indicates the amount of chlorine to be injected when the water quality impact condition is satisfied; an output device that outputs the amount of chlorine injected that is inferred by the inference device; Equipped with The water quality influence conditions include a target residual chlorine concentration, which is a target value of the concentration of residual chlorine contained in tap water provided from the water treatment plant, the amount of the water to be treated flowing into a treatment tank of the water treatment plant, and a chlorine demand, which is the amount of chlorine to be injected until residual chlorine begins to be detected in the water to be treated when chlorine is injected into the water to be treated. A monitoring and control system characterized by:

2. The water quality impact conditions include at least one of the day of the week when the chlorine is injected and the time of day when the chlorine is injected.

2. The monitoring and control system according to claim 1.

3. Further provided is a learning device that generates the trained model based on the water quality impact conditions and the amount of chlorine injected, The learning device a second data acquisition device that acquires the water quality influence conditions and the amount of chlorine injected as learning data; a model generation unit that generates the trained model using the training data; having 3. The monitoring and control system according to claim 1 or 2.

4. the second data acquisition device, The water quality influence conditions and the amount of chlorine injected when the chlorine is actually injected into the water to be treated are acquired as the learning data.

4. The monitoring and control system according to claim 3.

5. the second data acquisition device, When the amount of chlorine to be injected output by the output device is not applied and the amount of chlorine to be injected is specified by an operator, the water quality influence condition and the amount of chlorine to be injected specified by the operator are acquired as the learning data.

5. The monitoring and control system according to claim 3 or 4.

6. Further provided is a display device that displays the amount of injected chlorine output by the output device, 6. The monitoring and control system according to claim 1, wherein:

7. a data acquisition step in which a data acquisition device acquires water quality influence conditions, which are information on conditions that influence the water quality of water to be treated when chlorine is injected into the water to be treated in a water treatment plant; an inference step in which an inference device infers the amount of chlorine to be injected from the water quality impact condition acquired in the data acquisition step using a trained model for inferring the amount of chlorine to be injected when the water quality impact condition is satisfied; an output step in which an output device outputs the amount of chlorine to be injected inferred by the inference device; Including, The water quality influence conditions include a target residual chlorine concentration, which is a target value of the concentration of residual chlorine contained in tap water provided from the water treatment plant, the amount of the water to be treated flowing into a treatment tank of the water treatment plant, and a chlorine demand, which is the amount of chlorine to be injected until residual chlorine begins to be detected in the water to be treated when chlorine is injected into the water to be treated. A monitoring and control method comprising:

8. a data acquisition step of acquiring water quality influence conditions, which are information on conditions that affect the water quality of the water to be treated when chlorine is injected into the water to be treated in the water treatment plant; an inference step of inferring the amount of chlorine to be injected from the water quality impact conditions acquired in the data acquisition step using a trained model for inferring the amount of chlorine to be injected, which indicates the amount of chlorine to be injected when the water quality impact conditions are satisfied; an output step of outputting the estimated injection amount of chlorine; on the computer, The water quality influence conditions include a target residual chlorine concentration, which is a target value of the concentration of residual chlorine contained in tap water provided from the water treatment plant, the amount of the water to be treated flowing into a treatment tank of the water treatment plant, and a chlorine demand, which is the amount of chlorine to be injected until residual chlorine begins to be detected in the water to be treated when chlorine is injected into the water to be treated. A monitoring and control program comprising:

Images