DATA PROCESSING APPARATUS, PROGRAM, AND DATA PROCESSING METHOD

Abstract

To provide a data processing device, a program, and a data processing method related to a water treatment facility.SOLUTION: The data processing device includes: a facility-related data acquisition part for acquiring facility-related data related to a water treatment facility; a learning model storing part for storing multiple types of learning models with the facility-related data as input and control-related data related to control of the water treatment facility as output; an environment identification part for identifying an environment of the water treatment facility based on the facility-related data; a model selecting part for selecting one learning model corresponding to the environment of the water treatment facility identified by the environment identifying part from among the multiple types of learning models stored in the learning model storing part; and a control-related data outputting part for inputting the facility-related data before a reference time to the one learning model selected by the model selecting part, and outputting the control-related data after the reference time.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a data processing device, a program, and a data processing method. [Background technology]

[0002] Patent Document 1 describes a sewage treatment control device that automatically sets control parameters suitable for a sewage treatment plant. [Prior art document] [Patent documents] [Patent Document 1] JP 2001-047080 A Summary of the Invention

[0003] According to an embodiment of the present invention, there is provided a data processing device. The data processing device may include a facility-related data acquisition unit that acquires facility-related data related to a water treatment facility. The data processing device may include a learning model storage unit that stores a plurality of types of learning models that receive facility-related data as input and output control-related data related to control of the water treatment facility. The data processing device may include an environment identification unit that identifies an environment of the water treatment facility based on the facility-related data. The data processing device may include a model selection unit that selects one learning model corresponding to the environment of the water treatment facility identified by the environment identification unit from the plurality of types of learning models stored in the learning model storage unit. The data processing device may include a control-related data output unit that inputs facility-related data from a reference time before to one learning model selected by the model selection unit and outputs control-related data from the reference time onward.

[0004] The environment identification unit may identify the environment of the water treatment facility from a plurality of environment candidates for each external factor that affects the processing capacity of the water treatment. The environment identification unit may identify the environment of the water treatment facility from environment candidates including two or more of a normal state, a highly turbid raw water state, a rainy state, a cold state, and a high temperature state. The data processing device may further include a facility-related data storage unit that stores the facility-related data of the water treatment facility in the environment identified by the environment identification unit in association with the environment. The data processing device may further include a learning model generation unit that generates the learning model for each environment of the water treatment facility using the facility-related data corresponding to the environment and stores the learning model in the learning model storage unit.

[0005] The data processing device may further include a learning model update unit that generates a new learning model using the facility-related data corresponding to the environment in response to determining that it is time to perform an update process to update the learning model corresponding to the environment after the one learning model corresponding to the environment of the water treatment facility is selected by the model selection unit, and when the new learning model is selected based on a comparison result between the new learning model and the one learning model, associates the environment with the new learning model and stores it in the learning model storage unit instead of the one learning model. The learning model update unit may generate the new learning model using the facility-related data corresponding to the environment other than a predetermined period in the past based on the update timing, and compare the new learning model with the one learning model using the facility-related data for a predetermined period in the past based on the update timing. The learning model update unit may generate a plurality of different new learning models using the facility-related data corresponding to the environment, and when selecting one of the plurality of different new learning models based on a comparison result between the plurality of different new learning models and the one learning model, may store the selected new learning model in the learning model storage unit in association with the environment instead of the one learning model. The learning model update unit may generate the plurality of different new learning models with different parameters of a learning method. The learning model update unit may generate the plurality of different new learning models by changing the type of facility-related data used among the plurality of types of facility-related data stored in the facility-related data storage unit. The learning model update unit may generate the plurality of different new learning models using the facility-related data corresponding to different periods among the facility-related data stored in the facility-related data storage unit. The learning model update unit may generate the plurality of different new learning models with different degrees of learning.

[0006] The data processing device may further include a control parameter acquisition unit that acquires a control parameter to be set in the water treatment facility, a change parameter generation unit that generates a plurality of change parameters by changing the control parameter, and a parameter selection unit that selects one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the one learning model selected by the model selection unit. The control parameter acquisition unit may acquire the control parameter included in the control-related data output by the control-related data output unit. The parameter selection unit may include a candidate parameter identification unit that identifies a candidate parameter for each of the plurality of candidate parameters, the first parameter being input to the one learning model in which a target variable is a first parameter and output, and a cost estimation unit that estimates a cost of a consumable object to be consumed when each of the plurality of candidate parameters is set in the water treatment facility when the plurality of candidate parameters are identified by the candidate parameter identification unit, and may select one of the plurality of candidate parameters based on the cost. The parameter selection unit may include a cost estimation unit that estimates the cost of a consumption object consumed when the plurality of candidate parameters are set in the water treatment facility, and a candidate parameter identification unit that inputs the plurality of candidate parameters to the one learning model in ascending order of cost until the first parameter output by the one learning model having a first parameter as a target variable satisfies a predetermined condition, and identifies a candidate parameter that satisfies the condition. The cost estimation unit may estimate the cost of electricity consumed when the candidate parameters are set in the water treatment facility. The cost estimation unit may estimate the cost of chemicals consumed when the candidate parameters are set in the water treatment facility. The first parameter may be the quality of water discharged from the water treatment facility, and may satisfy the predetermined condition when the water quality is better than a predetermined threshold. The first parameter may be the moisture content of sludge in the water treatment facility, and may satisfy the predetermined condition when the moisture content is lower than a predetermined threshold.

[0007] According to an embodiment of the present invention, a data processing device is provided. The data processing device may include a facility-related data acquisition unit that acquires facility-related data related to a processing facility. The data processing device may include a learning model storage unit that stores a plurality of types of learning models that receive facility-related data as input and output control-related data related to control of the processing facility. The data processing device may include an environment identification unit that identifies an environment of the processing facility based on the facility-related data. The data processing device may include a model selection unit that selects one learning model corresponding to the environment of the processing facility identified by the environment identification unit from the plurality of types of learning models stored in the learning model storage unit. The data processing device may include a control-related data output unit that inputs facility-related data from a reference time before to one learning model selected by the model selection unit and outputs control-related data from the reference time onward.

[0008] According to an embodiment of the present invention, there is provided a data processing device. The data processing device may include a facility-related data acquisition unit that acquires facility-related data related to a water treatment facility. The data processing device may include a learning model storage unit that stores a learning model in which the facility-related data is input and control-related data related to control of the water treatment facility is output. The data processing device may include a learning model update unit that generates a new learning model using the facility-related data in response to determining that it is time to execute an update process to update the learning model stored in the learning model storage unit, and when the new learning model is selected based on a comparison result between the new learning model and the learning model, stores the new learning model in the learning model storage unit instead of the learning model.

[0009] The learning model update unit may generate the new learning model using the facility-related data other than the facility-related data for a predetermined period in the past based on the update timing, and compare the new learning model with the learning model using the facility-related data for a predetermined period in the past based on the update timing. The learning model update unit may generate a plurality of different new learning models using the facility-related data, and when selecting one of the plurality of different new learning models based on a comparison result between the plurality of different new learning models and the learning model, store the new learning model in the learning model storage unit instead of the learning model.

[0010] According to one embodiment of the present invention, there is provided a data processing device. The data processing device may include a facility-related data acquisition unit that acquires facility-related data related to a water treatment facility. The data processing device may include a learning model storage unit that stores a learning model in which the facility-related data is input and control-related data related to control of the water treatment facility is output. The data processing device may include a control parameter acquisition unit that acquires control parameters to be set in the water treatment facility. The data processing device may include a change parameter generation unit that generates a plurality of change parameters by changing the control parameters. The data processing device may include a parameter selection unit that selects one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the learning model stored in the learning model storage unit.

[0011] The parameter selection unit may include a candidate parameter identification unit that identifies a candidate parameter for which the first parameter, which is input to the learning model in which the objective variable is a first parameter and output, satisfies a predetermined condition for each of the plurality of candidate parameters, and a cost identification unit that identifies, when the plurality of candidate parameters are identified by the candidate parameter identification unit, a cost identification unit that identifies, for each of the plurality of candidate parameters, a cost of a consumable object that will be consumed when set in the water treatment facility, and may select one of the plurality of candidate parameters based on the cost. The parameter selection unit may include a cost identification unit that identifies, for the plurality of candidate parameters, a cost of a consumable object that will be consumed when set in the water treatment facility, and a candidate parameter identification unit that inputs the plurality of candidate parameters into the learning model in order of decreasing cost until the first parameter, which is output by the learning model in which the objective variable is a first parameter, satisfies a predetermined condition, and identifies a candidate parameter that satisfies the condition.

[0012] According to one embodiment of the present invention, there is provided a program for causing a computer to function as the data processing device.

[0013] According to one embodiment of the present invention, there is provided a data processing method executed by a computer. The data processing method may include a facility-related data acquisition step of acquiring facility-related data related to a water treatment facility. The data processing method may include a learning model storage step of storing, in a learning model storage unit, a plurality of types of learning models that receive the facility-related data as an input and output control-related data related to control of the water treatment facility. The data processing method may include an environment identification step of identifying an environment of the water treatment facility based on the facility-related data. The data processing method may include a model selection step of selecting, from the plurality of types of learning models stored in the learning model storage unit, one learning model that corresponds to the environment of the water treatment facility identified in the environment identification step. The data processing method may include a control-related data output step of inputting facility-related data from a reference time to one learning model selected in the model selection step, and outputting control-related data from the reference time onward.

[0014] According to one embodiment of the present invention, there is provided a data processing method executed by a computer. The data processing method may include a facility-related data acquisition step of acquiring facility-related data related to a water treatment facility. The data processing method may include a learning model storage step of storing a learning model in a learning model storage unit, the learning model having the facility-related data as an input and control-related data related to control of the water treatment facility as an output. The data processing method may include a learning model update step of generating a new learning model using the facility-related data in response to determining that it is time to perform an update process to update the learning model stored in the learning model storage unit, and when the new learning model is selected based on a comparison result between the new learning model and the learning model, storing the new learning model in the learning model storage unit instead of the learning model.

[0015] According to one embodiment of the present invention, there is provided a data processing method executed by a computer. The data processing method may include a facility-related data acquisition stage of acquiring facility-related data related to a water treatment facility. The data processing method may include a control parameter acquisition unit of acquiring control parameters to be set for the water treatment facility. The data processing method may include a changed parameter generation unit of generating a plurality of changed parameters by changing the control parameters. The data processing method may include a parameter selection unit of selecting one of a plurality of candidate parameters including the control parameter and a plurality of changed parameters by using a learning model that receives the facility-related data as input and outputs control-related data related to control of the water treatment facility.

[0016] The above summary of the invention does not list all of the necessary features of the present invention. Also, subcombinations of these features may also be inventions. [Brief description of the drawings]

[0017] [Figure 1] An example of a system 10 is shown diagrammatically. [Diagram 2] An example of a water treatment facility 200 is shown diagrammatically. [Diagram 3] An example of a water treatment facility 200 is shown diagrammatically. [Figure 4] 1 illustrates an example of a functional configuration of a data processing device 100. [Diagram 5] 1 illustrates an example of a processing flow by the data processing device 100. [Figure 6] An example of the environment determination process of the water treatment facility 200 by the environment determination unit 122 will be described. [Figure 7] 13 shows an example of a flow of a learning model update process performed by the learning model update unit 130. [Figure 8] 1 illustrates an example of a processing flow by the data processing device 100. [Figure 9] 1 illustrates an example of a processing flow by the data processing device 100. [Figure 10]An example of a hardware configuration of a computer 1200 functioning as the data processing device 100 is shown in schematic form. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0018] The present invention will be described below through embodiments of the invention, but the following embodiments do not limit the scope of the invention according to the claims. Furthermore, not all of the combinations of features described in the embodiments are necessarily essential to the solution of the invention.

[0019] FIG. 1 shows an example of a system 10. The system 10 includes a data processing device 100 that processes facility-related data related to a processing facility 20. The data processing device 100 outputs control-related data related to control of the processing facility 20 based on the facility-related data. The data processing device 100 outputs control-related data that is desirable to be set in the processing facility 20 after a reference time, for example, based on facility-related data older than a reference time. The control-related data may include control parameters. The reference time includes at least a time specified by a user, a current time, etc. The user may be a user and an administrator of the data processing device 100, etc. The user may be a user and an administrator of the processing facility 20, etc.

[0020] The facility-related data includes control-related data set in the treatment facility 20. The facility-related data may include control-related data set manually in the treatment facility 20. The control-related data may be data set in the treatment facility 20 by a manager or the like of the treatment facility 20 in response to changes in the environment or situation of the treatment facility 20, or the like.

[0021] The facility-related data may include data indicative of the environment of the treatment facility 20. For example, the facility-related data may include data indicative of the climate of the area of ​​the treatment facility 20. The facility-related data may include data indicative of the weather of the area of ​​the treatment facility 20. The facility-related data may include data indicative of the temperature of the area of ​​the treatment facility 20. The facility-related data may include data indicative of the humidity of the area of ​​the treatment facility 20. The facility-related data may include data indicative of the solar radiation conditions of the area of ​​the treatment facility 20.

[0022] The facility-related data may include data indicating the status of the treatment facility 20. The facility-related data may include data indicating the status of each piece of equipment in the treatment facility 20.

[0023] The data processing device 100 may collect facility-related data of the processing facility 20 from the processing facility 20. The data processing device 100 collects the facility-related data of the processing facility 20, for example, from a management device that is disposed in the processing facility 20 and manages the processing facility 20. The data processing device 100 may collect the facility-related data of the processing facility 20 from a sensor 30 that is disposed in the processing facility 20. The data processing device 100 may collect the facility-related data of the processing facility 20 from an information providing server 40. The information providing server 40 may be, for example, a server that is disposed on the Internet and provides climate data of various locations, and the data processing device 100 may receive data related to the climate of the processing facility 20 from the information providing server 40.

[0024] The data processing device 100 and the management device of the processing facility 20 may be directly connected. For example, the data processing device 100 and the management device of the processing facility 20 are connected by wire. For example, the data processing device 100 and the management device of the processing facility 20 are connected wirelessly. Examples of wireless communication methods used by the data processing device 100 and the processing facility 20 include, but are not limited to, Bluetooth (registered trademark), ZigBee (registered trademark), Wi-Fi (registered trademark) (Wireless Fidelity), EnOcean (registered trademark), and Wi-SUN (registered trademark). Similarly, the data processing device 100 and the sensor 30 may be directly connected. The data processing device 100 and the management device of the processing facility 20 may be connected via a network. Examples of networks used by the data processing device 100 and the management device of the processing facility 20 include a LAN (Local Area Network), a WAN (Wide Area Network), and the Internet. Similarly, the data processing device 100 and the sensor 30 may be connected via a network. Note that the data processing device 100 and the management device of the processing facility 20 may be integrated.

[0025] The data processing device 100 generates a learning model that uses, for example, facility-related data as input and control-related data as output. The data processing device 100 generates a learning model using, for example, control-related data set in the treatment facility 20 by an administrator or the like of the treatment facility 20 as training data. This learning model makes it possible to predict the control-related data that the administrator or the like of the treatment facility 20 will set in accordance with the environment and circumstances of the treatment facility 20. This makes it possible to reduce the burden on the administrator or the like of the treatment facility 20 and to realize an environment in which the treatment facility 20 can be appropriately controlled even when the administrator or the like of the treatment facility 20 is absent.

[0026] The data processing device 100 may generate multiple types of learning models. For example, the data processing device 100 generates and stores a learning model for each environment of the treatment facility 20. When outputting control-related data, the data processing device 100 identifies the environment of the treatment facility 20, inputs facility-related data from before a reference time into one learning model corresponding to the identified environment, and outputs control-related data from the reference time onward. This can improve the prediction accuracy of predicting appropriate control-related data compared to a case where the same learning model is used in any environment.

[0027] The data processing device 100 identifies, for example, the climate of the area of ​​the processing facility 20 as the environment of the processing facility 20. The data processing device 100 identifies, for example, events occurring in the processing facility 20 as the environment of the processing facility 20. The data processing device 100 identifies, for example, weekdays, Saturdays, Sundays, holidays, etc. as the environment of the processing facility 20. For example, when an abnormality occurs in equipment of the processing facility 20, the data processing device 100 identifies an equipment abnormality as the environment of the processing facility 20.

[0028] The treatment facility 20 is, for example, a water treatment facility. The treatment facility 20 may be a sewage treatment plant. The treatment facility 20 may be a water purification plant. Furthermore, the treatment facility 20 is, for example, a steel plant. These are merely examples, and the treatment facility 20 may be a treatment facility that treats any object. For example, the treatment facility 20 may be a facility in which control parameters need to be adjusted depending on the situation. For example, the treatment facility 20 may be a facility in which control parameters are changed but there is a time lag before the results appear.

[0029] Fig. 2 shows a schematic diagram of an example of a water treatment facility 200. Fig. 2 shows an example in which the water treatment facility 200 is a sewage treatment plant. The water treatment facility 200 includes a management device 210, a grit tank 211, a primary sedimentation tank 212, a reaction tank 213, a final sedimentation tank 214, a sludge treatment facility 215, and a chlorine contact tank 216.

[0030] The management device 210 manages the treatment in each of the facilities, namely, the grit basin 211, the primary sedimentation basin 212, the reaction tank 213, the final sedimentation tank 214, the sludge treatment facility 215, and the chlorine contact tank 216. The management device 210 manages, for example, control parameters set for the facilities. The management device 210 also manages, for example, the status of the facilities.

[0031] Sewage flows into the settling basin 211 from the outside. In the settling basin 211, garbage, soil, and the like are removed from the sewage.

[0032] Sewage slowly flows into the primary sedimentation tank 212 from the grit basin 211. In the primary sedimentation tank 212, fine sand and the like that could not be removed in the grit basin 211 settles to the bottom. The sludge that settles in the primary sedimentation tank 212 is sent to a sludge treatment facility 215.

[0033] The supernatant water from the primary sedimentation tank 212 flows into the reaction tank 213. In the reaction tank 213, the supernatant water is mixed with activated sludge, and air is pumped in to stir the mixture.

[0034] Activated sludge flows into the final settling tank 214 from the reaction tank 213. In the final settling tank 214, activated sludge flocs are settled and separated into supernatant water (treated water) and sludge. The sludge settled in the final settling tank 214 is returned to the reaction tank 213 or sent to a sludge treatment facility 215.

[0035] The sludge treatment facility 215 treats the sludge sent from the primary sedimentation tank 212 and the final sedimentation tank 214. The sludge treatment facility 215 reduces the water content of the sludge and reduces its volume, for example, by thickening or mechanical dehydration.

[0036] Treated water flows into the chlorine contact tank 216 from the final sedimentation tank 214. In the chlorine contact tank 216, the treated water is disinfected with chlorine and then discharged into a river, the sea, or the like.

[0037] The management device 210 transmits facility-related data of the water treatment facility 200 to the data processing device 100. The management device 210 includes, for example, the amount of water flowing in from outside in the facility-related data. The management device 210 includes, for example, the amount of water in each piece of equipment in the facility-related data.

[0038] The management device 210 may include, for example, the amount of sludge sent from the primary sedimentation tank 212 to the sludge treatment facility 215 in the facility-related data. The management device 210 may include, in the facility-related data, the amount of sludge actually withdrawn from the primary sedimentation tank 212. The management device 210 may include, in the facility-related data, a set value for the amount of sludge withdrawn from the primary sedimentation tank 212.

[0039] The management device 210 may include, for example, the amount of air sent to the activated sludge in the reaction tank 213 in the facility-related data. The management device 210 may include, in the facility-related data, the amount of air actually sent to the reaction tank 213. The management device 210 may include, in the facility-related data, a set value for the amount of air sent to the reaction tank 213.

[0040] The management device 210 includes, for example, dissolved oxygen (DO) in the reaction tank 213 in the facility-related data. The management device 210 may include the dissolved oxygen measured in the reaction tank 213 in the facility-related data. The management device 210 may include the dissolved oxygen set as a target value in the facility-related data.

[0041] The management device 210 may include, for example, the amount of sludge withdrawn from the final settling tank 214 in the facility-related data. The management device 210 may include, in the facility-related data, the amount of sludge actually withdrawn from the final settling tank 214. The management device 210 may include, in the facility-related data, a set value for the amount of sludge withdrawn from the final settling tank 214.

[0042] The management device 210 may include, for example, the amount of returned sludge returned from the final settling tank 214 to the reaction tank 213 in the facility-related data. The management device 210 may include, in the facility-related data, the amount of sludge actually returned from the final settling tank 214 to the reaction tank 213. The management device 210 may include, in the facility-related data, a set value for the amount of returned sludge.

[0043] The management device 210 may include, for example, the amount of excess sludge disposed of at the sludge treatment facility 215 in the facility-related data. The management device 210 may include, in the facility-related data, the amount of sludge actually disposed of from the sludge treatment facility 215. The management device 210 may include, in the facility-related data, a set value for the amount of excess sludge disposed of at the sludge treatment facility 215.

[0044] The management device 210 may include, for example, the water quality of the water discharged from the chlorine contact tank 216 to a river, the sea, or the like in the facility-related data. The management device 210 may include the water quality measured at the time of discharge in the facility-related data.

[0045] The management device 210 may include, in the facility-related data, data indicating the climate of the area of ​​the water treatment facility 200. The management device 210 includes, in the facility-related data, data indicating the climate of the area of ​​the water treatment facility 200, which is manually input by, for example, an administrator of the water treatment facility 200.

[0046] The management device 210 may include information on events at the water treatment facility 200 in the facility-related data. The management device 210, for example, includes information on construction work being carried out at the water treatment facility 200 in the facility-related data. The management device 210, for example, includes information on construction work planned for the water treatment facility 200 in the facility-related data. For example, when an abnormality or failure occurs in equipment at the water treatment facility 200, the management device 210 includes information indicating the equipment in which the abnormality or failure occurs in the facility-related data. For example, when a disaster such as an earthquake occurs at the water treatment facility 200, the management device 210 includes information indicating the disaster that has occurred in the facility-related data.

[0047] The sensor 30 transmits facility-related data of the water treatment facility 200 to the data processing device 100. The sensor 30 may include a rain gauge, and the facility-related data may include the amount of rainfall in the area of ​​the water treatment facility 200. The sensor 30 may include a thermometer, and the facility-related data may include the air temperature in the area of ​​the water treatment facility 200. The sensor 30 may include an actinometer, and the facility-related data may include the amount of solar radiation in the area of ​​the water treatment facility 200. The sensor 30 may include a hygrometer, and the facility-related data may include the humidity in the area of ​​the water treatment facility 200.

[0048] The data processing device 100, for example, takes facility-related data up to a reference time as input, and generates a learning model that outputs the amount of air to be sent to the activated sludge in the reaction tank 213, which should be set at an arbitrary time, such as one hour, from the reference time. The data processing device 100, for example, takes facility-related data up to a reference time as input, and generates a learning model that outputs the amount of sludge to be returned from the final settling tank 214 to the reaction tank 213, which should be set at an arbitrary time from the reference time. The data processing device 100, for example, takes facility-related data up to a reference time as input, and generates a learning model that outputs the amount of excess sludge to be disposed of in the sludge treatment facility 215, which should be set at an arbitrary time from the reference time. The data processing device 100, for example, takes facility-related data up to a reference time as input, and generates a learning model that outputs the amount of sewage inflow into the gritification tank 211, which should be set at an arbitrary time from the reference time. The data processing device 100, for example, takes facility-related data up to a reference time as input, and generates a learning model that outputs the target value of dissolved oxygen in the reaction tank 213, which should be set at an arbitrary time from the reference time.

[0049] It should be noted that the configuration of water treatment facility 200 shown in FIG. 2 is just one example, and water treatment facility 200 may include facilities other than those shown in FIG. 2, or may not include some of the facilities shown in FIG.

[0050] Fig. 3 is a schematic diagram of an example of a water treatment facility 200. Fig. 3 illustrates a case where the water treatment facility 200 is a water purification plant. The water treatment facility 200 includes a management device 220, a receiving tank 221, a mixing tank 222, a flocculation tank 223, a sedimentation tank 224, a filtration tank 225, a chlorine mixing tank 226, a distribution tank 227, a pumping station 228, and a sludge treatment facility 229.

[0051] The management device 220 manages the treatment in each of the facilities, namely, the receiving tank 221, the mixing tank 222, the flocculation tank 223, the sedimentation tank 224, the filtration tank 225, the chlorine mixing tank 226, the distribution tank 227, the pumping station 228, and the sludge treatment facility 229. The management device 220 manages, for example, control parameters set for the facilities. The management device 220 also manages, for example, the status of the facilities.

[0052] Raw water sent to the water treatment facility 200 flows into the receiving tank 221. In the receiving tank 221, garbage, soil, and the like are removed from the raw water.

[0053] Raw water flows into the mixing basin 222 from the receiving basin 221. In the mixing basin 222, a flocculant is added to the raw water.

[0054] Raw water to which a coagulant has been added flows from the mixing basin 222 into the flocculation basin 223. In the flocculation basin 223, the coagulant combines with suspended solids in the raw water to form flocs.

[0055] Flocs flow into the settling tank 224 from the flocculation tank 223. In the settling tank 224, the flocs and fine sand are settled and separated from the supernatant water. The settled sludge is sent to a sludge treatment facility 229.

[0056] The supernatant water flows into the filtration basin 225 from the settling basin 224. In the filtration basin 225, the supernatant water is filtered to remove fine suspended matter that could not be removed by coagulation and sedimentation.

[0057] The water filtered in the filtration basin 225 flows into the chlorine mixing basin 226. In the chlorine mixing basin 226, chlorine is injected into the inflowing water.

[0058] Water into which chlorine has been injected flows from the chlorine mixing tank 226 into the distributing tank 227. Purified water accumulates in the distributing tank 227. A pump station 228 sends the water from the distributing tank 227 to a water supply station or the like.

[0059] The sludge treatment facility 229 treats the sludge sent from the settling tank 224. The sludge treatment facility 229 reduces the water content of the sludge and reduces its volume, for example, by thickening or mechanical dehydration.

[0060] The management device 220 transmits facility-related data of the water treatment facility 200 to the data processing device 100. The management device 220 includes, for example, the amount of raw water flowing in from outside in the facility-related data. The management device 220 includes, for example, the amount of water in each piece of equipment in the facility-related data.

[0061] The management device 220 may include in the facility-related data, for example, the amount of flocculant added to the raw water in the mixing basin 222. The management device 220 may include in the facility-related data the amount of flocculant actually added to the mixing basin 222. The management device 220 may include in the facility-related data the set value of the flocculant to be added to the mixing basin 222.

[0062] The management device 220 may, for example, include in the facility-related data the amount of sludge sent from the settling tank 224 to the sludge treatment facility 229. The management device 220 may include in the facility-related data the amount of sludge actually removed from the settling tank 224. The management device 220 may include in the facility-related data a set value for the amount of sludge to be removed from the settling tank 224.

[0063] The management device 220 may include, for example, the amount of chlorine injected into the chlorine mixing pond 226 in the facility-related data. The management device 220 may include, in the facility-related data, the amount of chlorine actually injected into the chlorine mixing pond 226. The management device 220 may include, in the facility-related data, the set value of chlorine to be injected into the chlorine mixing pond 226.

[0064] The management device 220 may include, for example, the amount of excess sludge disposed of at the sludge treatment facility 229 in the facility-related data. The management device 220 may include, in the facility-related data, the amount of sludge actually disposed of from the sludge treatment facility 229. The management device 220 may include, in the facility-related data, a set value for the amount of excess sludge disposed of at the sludge treatment facility 229.

[0065] The management device 220 may include, for example, the water quality of the water discharged from the pump station 228 in the facility-related data. The management device 220 may include the water quality measured in the water reservoir 227, the pump station 228, etc. in the facility-related data.

[0066] The management device 220 may include, in the facility-related data, data indicating the climate of the area of ​​the water treatment facility 200. The management device 220 includes, in the facility-related data, data indicating the climate of the area of ​​the water treatment facility 200, which is manually input by, for example, an administrator of the water treatment facility 200.

[0067] The management device 220 may include information on events at the water treatment facility 200 in the facility-related data. The management device 220, for example, includes information on construction work being carried out at the water treatment facility 200 in the facility-related data. The management device 220, for example, includes information on construction work planned for the water treatment facility 200 in the facility-related data. For example, when an abnormality or failure occurs in equipment at the water treatment facility 200, the management device 220 includes information indicating the equipment in which the abnormality or failure occurs in the facility-related data. For example, when a disaster such as an earthquake occurs at the water treatment facility 200, the management device 220 includes information indicating the disaster that has occurred in the facility-related data.

[0068] The sensor 30 transmits facility-related data of the water treatment facility 200 to the data processing device 100. The sensor 30 may include a rain gauge, and the facility-related data may include the amount of rainfall in the area of ​​the water treatment facility 200. The sensor 30 may include a thermometer, and the facility-related data may include the air temperature in the area of ​​the water treatment facility 200. The sensor 30 may include an actinometer, and the facility-related data may include the amount of solar radiation in the area of ​​the water treatment facility 200. The sensor 30 may include a hygrometer, and the facility-related data may include the humidity in the area of ​​the water treatment facility 200.

[0069] The data processing device 100, for example, takes facility-related data up to a reference time as input, and generates a learning model that outputs the amount of coagulant to be added to raw water in the mixing basin 222, which should be set any time after the reference time. The data processing device 100, for example, takes facility-related data up to a reference time as input, and generates a learning model that outputs the amount of excess sludge to be disposed of in the sludge treatment facility 229, which should be set any time after the reference time. The data processing device 100, for example, takes facility-related data up to a reference time as input, and generates a learning model that outputs the amount of raw water inflow into the landing tank 221, which should be set any time after the reference time. The data processing device 100, for example, takes facility-related data up to a reference time as input, and generates a learning model that outputs the amount of chlorine to be injected into the chlorine mixing basin 226, which should be set any time after the reference time.

[0070] It should be noted that the configuration of water treatment facility 200 shown in FIG. 3 is just one example, and water treatment facility 200 may include facilities other than those shown in FIG. 3, or may not include some of the facilities shown in FIG.

[0071] 4 shows an example of a schematic functional configuration of the data processing device 100. The data processing device 100 includes a storage unit 110, a facility-related data acquisition unit 120, an environment identification unit 122, a learning model generation unit 124, a model selection unit 126, a control-related data output unit 128, a learning model update unit 130, a control parameter acquisition unit 132, a change parameter generation unit 134, and a parameter selection unit 140. It is not essential that the data processing device 100 includes all of these units. Here, a case where the data processing device 100 processes facility-related data related to a water treatment facility 200 will be mainly taken as an example and described.

[0072] The storage unit 110 stores various data, and includes a facility-related data storage unit 112 that stores facility-related data, and a learning model storage unit 114 that stores a learning model.

[0073] The facility-related data acquiring unit 120 acquires facility-related data. The facility-related data acquiring unit 120 may acquire the facility-related data continuously. The facility-related data acquiring unit 120 stores the acquired facility-related data in the facility-related data storage unit 112.

[0074] The facility-related data acquiring unit 120 may acquire facility-related data of the water treatment facility 200. The facility-related data acquiring unit 120 receives the facility-related data of the water treatment facility 200, for example, from the water treatment facility 200. The facility-related data acquiring unit 120 receives the facility-related data of the water treatment facility 200, for example, from a sensor 30 installed in the water treatment facility 200.

[0075] The facility-related data acquiring unit 120 receives facility-related data of the water treatment facility 200 from, for example, the information providing server 40. The facility-related data acquiring unit 120 receives data indicating the climate of the area of ​​the water treatment facility 200 from, for example, the information providing server 40.

[0076] The environment identifying unit 122 identifies the environment of the water treatment facility 200. The environment identifying unit 122 may identify the environment of the water treatment facility 200 based on the facility-related data acquired by the facility-related data acquiring unit 120.

[0077] The environment identification unit 122 identifies the environment of the water treatment facility 200 from, for example, a plurality of environment candidates for each external factor that affects the water treatment capacity of the water treatment facility 200. The environment identification unit 122 identifies the environment of the water treatment facility 200 from, for example, the environment candidates including two or more of a normal state, a highly turbid raw water state, a rainy state, a cold state, and a high temperature state.

[0078] The candidate environments may include a climate in an area of ​​the water treatment facility 200. The candidate environments may include sunny weather. The candidate environments may include cloudy weather. The candidate environments may include rainy weather. The candidate environments may include sunny weather after rain. The candidate environments may include cloudy weather after rain. The candidate environments may include a midsummer day. The candidate environments may include a midsummer day after rain.

[0079] The environment candidate may include an event occurring in the water treatment facility 200. The environment candidate may include construction work. When construction work is being carried out in the water treatment facility 200, the environment identifying unit 122 identifies the environment of the water treatment facility 200 as construction work.

[0080] The environment candidate may include special. The environment identification unit 122 identifies the environment of the water treatment facility 200 as special, for example, when it is a public holiday. The environment identification unit 122 identifies the environment of the water treatment facility 200 as special, for example, when a disaster such as an earthquake has occurred in the water treatment facility 200. The environment identification unit 122 identifies the environment of the water treatment facility 200 as special, for example, when the area of ​​the water treatment facility 200 is experiencing abnormal weather such as a severe cold wave.

[0081] For example, when the water treatment facility 200 is a sewage treatment plant that is affected when a large amount of wastewater (for example, washing of a large pot in a food factory) flows out from a nearby factory or the like, the environment identification unit 122 identifies the environment of the water treatment facility 200 as special. In this case, the environment identification unit 122 may identify the environment of the water treatment facility 200 as XX factory washing or the like.

[0082] For example, when the water treatment facility 200 is a sewage treatment plant that is affected when a large amount of wastewater (for example, from pool cleaning, a prefectural festival, etc.) flows out in the vicinity, the environment identification unit 122 identifies the environment of the water treatment facility 200 as special. In this case, the environment identification unit 122 may identify the environment of the water treatment facility 200 as "XX event being held, etc."

[0083] For example, the environment identification unit 122 identifies the environment of the water treatment facility 200 as special when the water treatment facility 200 is a water purification plant, the amount of produced water is insufficient, and water is received from a prefectural water supply, etc. In this case, the environment identification unit 122 may identify the environment of the water treatment facility 200 as receiving prefectural water, etc.

[0084] For example, in the case where the water treatment facility 200 is a water purification plant that is affected by the operation of a nearby factory (e.g., a paper mill) that consumes a large amount of water, the environment determination unit 122 determines the environment of the water treatment facility 200 as special if the factory is not in operation. Note that in this case, the environment determination unit 122 may determine the environment of the water treatment facility 200 as "XX factory not in operation" or the like.

[0085] The environment candidate may include an abnormality. For example, when an abnormality or failure occurs in the equipment of the water treatment facility 200, the environment identifying unit 122 identifies the environment of the water treatment facility 200 as abnormal.

[0086] The environment identification unit 122 may identify the environment of the water treatment facility 200 based on manual input. For example, when the facility-related data includes data indicating the environment of the water treatment facility 200, the environment identification unit 122 identifies the environment of the water treatment facility 200 based on the data.

[0087] The environment identification unit 122 may automatically identify the environment of the water treatment facility 200. The environment identification unit 122 identifies the environment of the water treatment facility 200, for example, based on facility-related data acquired from a sensor 30 installed in the water treatment facility 200. As a specific example, the environment identification unit 122 identifies the climate of the area of ​​the water treatment facility 200 based on the facility-related data acquired from the sensor 30.

[0088] The environment determination unit 122 may determine the climate of the area of ​​the water treatment facility 200 according to the installation status of the sensor 30. For example, in the case where only a rain gauge is installed in the water treatment facility 200, the environment determination unit 122 determines that a day with a total rainfall of 0 mm per day is a sunny day, and a day with a total rainfall of more than 0 mm per day is a rainy day. In addition, for example, in the case where a rain gauge and a thermometer are installed in the water treatment facility 200, the environment determination unit 122 determines that a day with a total rainfall of 0 mm per day and a maximum temperature of 20 degrees or more and less than 30 degrees is a sunny day, a day with a total rainfall of 0 mm per day and a maximum temperature of 30 degrees or more is a midsummer day, and a day with a total rainfall of 0 mm per day and a maximum temperature of 20 degrees or less is a cloudy day. In addition, for example, in the case where a rain gauge and an actinometer are installed in the water treatment facility 200, the environment determination unit 122 determines that a day with a total rainfall of 0 mm per day and a solar radiation of 1.5 kW / m 2 If the above conditions exist, it is a midsummer day, with a total rainfall of 0 mm and solar radiation of 0.5 W / m 2 If the total rainfall does not exceed 0.5 mm per day, it is considered cloudy, if the total rainfall is 0.5 mm or more per day, it is considered rainy, and otherwise it is considered sunny. In this manner, the environment determination unit 122 may determine the environment of the water treatment facility 200 by comparing the data acquired from the sensor 30 with the reference value. The reference value may be arbitrarily set and may be changeable.

[0089] The facility-related data storage unit 112 may store the facility-related data of the water treatment facility 200 in the environment identified by the environment identifying unit 122 in association with the environment.

[0090] For example, the facility-related data storage unit 112 stores the facility-related data acquired by the facility-related data acquisition unit 120 in association with a label representing the environment identified by the environment identification unit 122. As a specific example, the facility-related data storage unit 112 stores the facility-related data during a rainy period in association with a rainy weather label. The facility-related data storage unit 112 may store the same facility-related data in association with a plurality of labels. As a specific example, the facility-related data storage unit 112 stores the facility-related data during a rainy period on a public holiday in association with rainy weather labels and rainy weather labels. The facility-related data storage unit 112 may store the facility-related data in association with a label for any period of time. For example, the facility-related data storage unit 112 stores the facility-related data in association with a label for each day. In this case, for example, when a cloudy period and a rainy period exist in one day, the facility-related data may be stored in association with a label corresponding to a longer period of time.

[0091] For example, the facility-related data storage unit 112 may classify and store the facility-related data acquired by the facility-related data acquisition unit 120 according to the environment identified by the environment identification unit 122. As a specific example, the facility-related data storage unit 112 stores facility-related data during a rainy period in a storage area corresponding to rainy weather, and stores facility-related data during a cloudy period in a storage area corresponding to cloudy weather. The facility-related data storage unit 112 may store the same facility-related data in multiple storage areas. As a specific example, the facility-related data storage unit 112 stores facility-related data during a rainy period on a public holiday in a storage area corresponding to the public holiday and a storage area corresponding to rainy weather. Note that the facility-related data storage unit 112 may store facility-related data during a rainy period on a public holiday in a storage area corresponding to the public holiday and a storage area corresponding to rainy weather.

[0092] The learning model generation unit 124 generates a learning model. The learning model generation unit 124 may store the generated learning model in the learning model storage unit 114. The learning model generation unit 124 may generate a learning model using facility-related data stored in the facility-related data storage unit 112. For example, the learning model generation unit 124 uses control-related data set for the water treatment facility 200 by an administrator or the like of the water treatment facility 200 as teacher data, and generates a learning model in which the facility-related data is used as an input and the control-related data is used as an output.

[0093] The learning model generation unit 124 generates a learning model using, for example, a random forest. The learning model generation unit 124 may generate a learning model using a random forest for some or all of the multiple types of facility-related data stored in the facility-related data storage unit 112. The learning method used by the learning model generation unit 124 is not limited to the random forest, and other known learning methods may be used.

[0094] The learning model generation unit 124 may generate a learning model using facility-related data corresponding to the environment for each environment of the water treatment facility 200. The learning model generation unit 124 may store the learning model generated using the facility-related data corresponding to the environment in the learning model generation unit 124 in association with the environment.

[0095] The model selection unit 126 selects a learning model to be used from the multiple types of learning models stored in the learning model storage unit 114. For example, when executing an output process of control-related data based on facility-related data, the model selection unit 126 selects one learning model corresponding to the environment of the water treatment facility 200 identified by the environment identification unit 122 from the multiple types of learning models stored in the learning model storage unit 114.

[0096] The control-related data output unit 128 executes output processing of control-related data based on the facility-related data, using the learning model stored in the learning model storage unit 114. For example, the control-related data output unit 128 displays and outputs the control-related data on a display provided in the data processing device 100. In addition, the control-related data output unit 128 transmits and outputs the control-related data to a communication terminal owned by, for example, the manager of the water treatment facility 200.

[0097] One of the purposes of the data processing device 100 of this embodiment is to brush up the learning model before officially introducing the learning model into the target water treatment facility 200, and after the brush-up is completed, the control-related data output unit 128 may reflect the control of the water treatment facility 200 by transmitting the control-related data to the water treatment facility 200.

[0098] The control-related data output unit 128, for example, inputs facility-related data from before a reference time into one learning model selected by the model selection unit 126, and outputs control-related data from after the reference time. Since water treatment by the water treatment facility 200 is affected by the environment of the water treatment facility 200, it is possible to output control-related data that allows water treatment to be performed more appropriately, compared to a case where a learning model prepared regardless of the environment of the water treatment facility 200 is used.

[0099] As a specific example, the control-related data output unit 128 inputs the facility-related data up to the present time into one learning model and outputs the control-related data one hour later. This makes it possible to notify the control-related data that is desirable to set based on the current environment, situation, etc. of the water treatment facility 200.

[0100] The learning model update unit 130 updates the learning model stored in the learning model storage unit 114. For example, after the model selection unit 126 selects one learning model corresponding to the environment of the water treatment facility 200, the learning model update unit 130 generates a new learning model using facility-related data corresponding to the environment in response to determining that it is time to execute an update process to update the learning model corresponding to the environment. The learning model update unit 130 may generate the learning model using the same learning method as the learning model generation unit 124. Then, the learning model update unit 130 selects the new learning model or the one learning model based on a comparison result between the new learning model and the one learning model. When the learning model update unit 130 selects the new learning model, it changes it to the one learning model, and stores the new learning model in the learning model storage unit 114 in association with the environment. When the learning model update unit 130 selects the one learning model, it may discard the new learning model.

[0101] Since a new learning model can be generated using newer facility-related data than a previous learning model, it is possible to update the learning model to one that is more likely to improve prediction accuracy. Also, for example, if the latest facility-related data contains inappropriate data, which reduces prediction accuracy, it is possible to avoid updating.

[0102] The learning model update unit 130 may compare the new learning model with the one learning model using control-related data manually set by the manager of the water treatment facility 200 in the past. For example, the learning model update unit 130 inputs facility-related data from a time before the manager of the water treatment facility 200 set the control-related data into both the new learning model and the one learning model, and outputs the control-related data. Then, the learning model update unit 130 identifies a difference between the output control-related data and the control-related data set by the manager of the water treatment facility 200, and selects the new learning model or the one learning model with the smaller identified difference. This makes it possible to select a learning model that can output control-related data that is closer to the control-related data based on the experience of the manager of the water treatment facility 200.

[0103] The learning model update unit 130 may compare the new learning model with the first learning model using control-related data that has been manually set most recently by the manager of the water treatment facility 200 or the like. For example, the learning model update unit 130 generates a new learning model using facility-related data other than a predetermined period in the past based on the update timing of the learning model, among the facility-related data corresponding to the environment identified by the environment identification unit 122, and compares the new learning model with the first learning model using facility-related data in a predetermined period in the past based on the update timing. When there is a major change in the water treatment facility 200, an accurate evaluation may not be possible even if an evaluation is performed using facility-related data before the change, but the accuracy of the evaluation can be improved by using the most recent facility-related data.

[0104] The learning model update unit 130 may generate a plurality of different new learning models using facility-related data corresponding to the environment. When the learning model update unit 130 selects one of the plurality of different new learning models based on a comparison result between the generated plurality of different new learning models and one learning model, the learning model update unit 130 may store the selected new learning model in the learning model storage unit 114 in association with the environment, instead of the one learning model.

[0105] The learning model update unit 130 generates a plurality of new learning models with different parameters of the learning method, for example. For example, when a random forest is used as the learning method, the learning model update unit 130 generates a plurality of new learning models while changing the parameters of the random forest. Examples of the parameters of the random forest include the number of decision trees, the number of branches of the decision tree, the number of layers of the decision tree, and branching conditions.

[0106] The learning model update unit 130 generates a plurality of new learning models, for example, by using different types of facility-related data among the plurality of types of facility-related data stored in the facility-related data storage unit 112. The learning model update unit 130 may generate a plurality of new learning models by using different numbers of facility-related data among the plurality of types of facility-related data, or by using different types even if the same number of data is used.

[0107] The learning model update unit 130 generates a plurality of different new learning models, for example, by using facility-related data corresponding to different time periods among the facility-related data stored in the facility-related data storage unit 112. For example, the learning model update unit 130 uses different facility-related data on a yearly, monthly, or daily basis. In addition, the learning model update unit 130 uses different facility-related data on a weekday / weekend basis, a day of the week, or a season basis.

[0108] The learning model update unit 130 generates a plurality of different new learning models with different learning orders, for example. The learning model update unit 130 generates a learning model with each of the plurality of orders, for example. The number of variations of the order may be any number.

[0109] As described above, by configuring to select a learning model from one learning model and various types of new learning models, it is possible to generate a learning model with higher prediction accuracy.

[0110] The control parameter acquisition unit 132 acquires control parameters to be set in the water treatment facility 200. The control parameter acquisition unit 132 acquires, for example, control parameters that have been decided to be set for the water treatment facility 200. The control parameter acquisition unit 132 acquires, for example, control parameters that have already been set for the water treatment facility 200. The control parameter acquisition unit 132 may acquire control parameters included in the control-related data output by the control-related data output unit 128.

[0111] The change parameter generating unit 134 generates a plurality of change parameters by changing the control parameters acquired by the control parameter acquiring unit 132. For example, the control parameters are the amount of air to be fed to the activated sludge in the reaction tank 213, and the amount of air to be fed is 20 m 3 / min, the change parameter generating unit 134 changes the 3 / min etc., 10, 30, 40m 3 A plurality of change parameters are generated in units of 100 / minute, etc. A method for generating the plurality of ON change parameters from the control parameters may be any method. Furthermore, the change parameter generating unit 134 may generate the plurality of change parameters according to an instruction from an administrator of the data processing device 100, etc.

[0112] The parameter selection unit 140 uses one learning model selected by the model selection unit 126 to select one of a plurality of candidate parameters including the control parameter acquired by the control parameter acquisition unit 132 and a plurality of change parameters generated by the change parameter generation unit 134. The parameter selection unit 140 may include a candidate parameter identification unit 142 and a cost estimator 144.

[0113] The candidate parameter identifying unit 142 identifies a candidate parameter from among a plurality of candidate parameters. For example, first, the candidate parameter identifying unit 142 sets the objective variable of one learning model selected by the model selecting unit 126 as a first parameter.

[0114] The first parameter may be the quality of the water discharged from the water treatment facility 200. If the water treatment facility 200 is a wastewater treatment plant, the first parameter may be the quality of the water discharged from the chlorine contactor 216 to a river, ocean, etc. If the water treatment facility 200 is a water purification plant, the first parameter may be the quality of the water discharged from the pumping station 228.

[0115] The first parameter may be the moisture content of the sludge in the water treatment facility 200. If the water treatment facility 200 is a sewage treatment plant, the first parameter may be the moisture content of the sludge in the reaction tank 213. If the water treatment facility 200 is a sewage treatment plant, the first parameter may be the moisture content of the sludge in the settling tank 224.

[0116] The candidate parameter identification unit 142 identifies a candidate parameter that satisfies a predetermined condition when the first parameter is input to a learning model in which the objective variable is the first parameter and output from the learning model. When the first parameter is the quality of water discharged from the water treatment facility 200, the first parameter satisfies the predetermined condition when the water quality is better than a predetermined threshold. For example, when the first parameter indicates a higher value as the water quality is better, the first parameter satisfies the predetermined condition when the first parameter is higher than the predetermined threshold. When the first parameter is the moisture content of sludge in the water treatment facility 200, the first parameter satisfies the predetermined condition when the moisture content is lower than a predetermined threshold. When one candidate parameter is identified by the candidate parameter identification unit 142, the parameter selection unit 140 may cause the control-related data output unit 128 to output the identified candidate parameter.

[0117] When a plurality of candidate parameters are identified by the candidate parameter identification unit 142, the cost estimation unit 144 estimates the cost of the consumption object that will be consumed when each of the plurality of candidate parameters is set in the water treatment facility 200. For example, the cost estimation unit 144 estimates the cost of electricity that will be consumed when the candidate parameters are set in the water treatment facility 200. For example, the cost estimation unit 144 estimates the amount of electricity consumption. For example, the cost estimation unit 144 estimates the electricity fee. The cost estimation unit 144 may estimate the cost of electricity using a pre-stored evaluation function. The evaluation function can be generated using, for example, the relationship between the control parameters and the amount of electricity usage when the control parameters are set for each piece of equipment in the water treatment facility 200. The cost estimation unit 144 may store an evaluation function acquired from outside, or may generate the evaluation function based on the facility-related data stored in the facility-related data storage unit 112.

[0118] The cost estimation unit 144 estimates, for example, the cost of chemicals consumed when the candidate parameters are set in the water treatment facility 200. Examples of chemicals include chlorine used in the chlorine contact tank 216, a coagulant used in the mixing tank 222, and chlorine used in the chlorine mixing tank 226. The cost estimation unit 144 may estimate the cost of chemicals using a pre-stored evaluation function. The evaluation function can be generated, for example, using the relationship between the control parameters and the amount of chemicals used when the control parameters are set for each piece of equipment in the water treatment facility 200. The cost estimation unit 144 may store an evaluation function acquired from outside, or may generate the evaluation function based on the facility-related data stored in the facility-related data storage unit 112.

[0119] The parameter selection unit 140 may select one of the multiple candidate parameters based on the cost estimated by the cost estimation unit 144. For example, the parameter selection unit 140 selects the candidate parameter with the lowest estimated cost from among the multiple candidate parameters. This makes it possible to identify a candidate parameter with lower electricity costs or chemical costs while ensuring the quality of the water discharged from the water treatment facility 200, thereby contributing to efficient operation of the water treatment facility 200.

[0120] Alternatively, the cost estimation unit 144 may estimate the cost of the consumables that will be consumed when the multiple candidate parameters are set in the water treatment facility 200, and the candidate parameter identification unit 142 may input the multiple candidate parameters to a learning model in ascending order of cost until a first parameter output by the learning model with the objective variable as the first parameter satisfies a predetermined condition, thereby identifying a candidate parameter that satisfies the condition. This makes it possible to reduce the overall processing load when the processing load of determining the cost of electricity or the cost of chemicals is higher than the processing load of determining whether the first parameter satisfies the condition.

[0121] Fig. 5 shows an example of a processing flow by the data processing device 100. Fig. 5 explains a processing flow in which the data processing device 100 changes the learning model to be used as appropriate and updates the learning model while collecting facility-related data. Note that here, a state in which the environment of the water treatment facility 200 is once identified and a single learning model corresponding to the environment is used to display and output control-related data is explained as a start state.

[0122] In step (sometimes abbreviated to S) 102, the facility-related data acquisition unit 120 collects facility-related data. In S104, the environment identification unit 122 identifies a label that represents the environment of the water treatment facility 200.

[0123] In S106, the control-related data output unit 128 determines that the learning model is to be changed if the environment of the water treatment facility 200 has changed. If it is determined that the learning model is to be changed, the process proceeds to S108, and if the environment of the water treatment facility 200 has not changed, the process proceeds to S110. In S108, the control-related data output unit 128 selects a learning model corresponding to the label identified in S104 from the learning model storage unit 114.

[0124] In S110, the learning model update unit 130 determines whether it is time to update the learning model. For example, when a predetermined period has passed since the currently used learning model was generated, the learning model update unit 130 determines that it is time to update the learning model. If it is determined that it is time to update, the process proceeds to S112, and if it is determined that it is not time to update, the process proceeds to S114. In S112, the learning model update unit 130 executes an update process for the learning model.

[0125] In S114, it is determined whether or not to end the process. For example, when the data processing device 100 has received an end instruction from the administrator of the data processing device 100 or the administrator of the water treatment facility 200, the data processing device 100 determines to end the process. When it is determined not to end the process, the process returns to S102.

[0126] 6 shows an example of the flow of the environment determination process of the water treatment facility 200 by the environment determination unit 122. Here, a specific example is given of the environment determination unit 122 determining the climate of the day at the end of the day based on the amount of solar radiation and the amount of rainfall obtained from an actinometer and a rain gauge.

[0127] In S202, the environment determination unit 122 determines whether the amount of rainfall is 0.5 mm or more per day. If it is 0.5 mm or more per day, it is determined to be rainy, and if it is not 0.5 mm or more per day, the process proceeds to S204. In S204, the environment determination unit 122 determines whether it was rainy the previous day. If it was rainy the previous day, the process proceeds to S206, and if it was not rainy, the process proceeds to S210.

[0128] In S206, the environment determination unit 122 determines whether the maximum solar radiation in a day is 0.5 kW / m 2 The environment determination unit 122 determines whether the 2 If it is judged to be less than 0.5kW / m, it is determined to be cloudy after rain. 2 If it is determined that the number is not less than the predetermined number, the process proceeds to S208.

[0129] In S208, the environment determination unit 122 determines whether the amount of solar radiation in a day is 1.5 kW / m 2 The environment determination unit 122 determines whether or not the above-mentioned times exist. If it determines that the above-mentioned times exist, it determines that the weather is a midsummer day after rain, and if it determines that the above-mentioned times do not exist, it determines that the weather is a sunny day after rain.

[0130] In S210, the environment determination unit 122 determines whether the maximum solar radiation in a day is 0.5 kW / m 2 The environment determination unit 122 determines whether the 2 If it is judged to be below 0.5kW / m, it is determined to be cloudy. 2 If it is determined that the difference is not less than the predetermined value, the process proceeds to S212.

[0131] In S212, the environment determination unit 122 determines whether the amount of solar radiation in a day is 1.5 kW / m 2 The environment determination unit 122 determines whether or not the above-mentioned times exist. If it is determined that the above-mentioned times exist, it determines that the day is a midsummer day, and if it is determined that the above-mentioned times do not exist, it determines that the day is a fine day.

[0132] Fig. 7 illustrates an example of the flow of a learning model update process by the learning model update unit 130. In Fig. 7, a process in which the learning model update unit 130 generates multiple types of new and different learning models and determines whether to update the learning model will be described.

[0133] In S302, the learning model update unit 130 generates a plurality of learning models using different types of facility-related data from among a plurality of types of facility-related data. In S304, the learning model update unit 130 generates a plurality of learning models using facility-related data corresponding to different periods. In S306, the learning model update unit 130 generates a plurality of learning models using different parameters of the learning method.

[0134] In S308, the learning model update unit 130 compares the multiple learning models generated in S302, S304, and S306 with the previous model to determine whether the evaluation result has improved. If it is determined that the evaluation result has improved, the process proceeds to S310. In S310, the learning model update unit 130 replaces the learning model from the previous model with the learning model with the highest evaluation among the multiple newly generated learning models.

[0135] 8 shows an example of a processing flow by the data processing device 100. Here, a processing flow for identifying control parameters capable of reducing the costs of consumption items such as electricity and chemicals while ensuring the quality of water discharged from the water treatment facility 200 will be described.

[0136] In S402, the control-related data output unit 128 selects a learning model corresponding to the environment of the water treatment facility 200 identified by the environment identification unit 122 from the learning model storage unit 114. In S404, the control parameter acquisition unit 132 acquires control parameters to be set in the water treatment facility 200.

[0137] In S406, the change parameter generation unit 134 generates a plurality of change parameters by changing the control parameters acquired by the control parameter acquisition unit 132 in S404. In S408, the candidate parameter identification unit 142 inputs each of a plurality of candidate parameters including the control parameters acquired in S404 and the plurality of change parameters generated in S406 into a learning model in which the objective variable is the water quality of the water discharged from the water treatment facility 200. In S410, the candidate parameter identification unit 142 identifies candidate parameters for which the water quality output from the learning model satisfies the condition. If a plurality of candidate parameters are identified (YES in S412), the process proceeds to S414, and if not a plurality of candidate parameters (NO in S412), the process proceeds to S418.

[0138] In S414, the cost estimation unit 144 estimates the cost of the consumption object that will be consumed when each of the multiple candidate parameters identified in S410 is set in the water treatment facility 200. In S416, the parameter selection unit 140 selects one of the multiple candidate parameters based on the cost estimated in S414. The parameter selection unit 140 selects, for example, the candidate parameter with the lowest cost estimated in S414.

[0139] In S418, the control-related data output unit 128 outputs the candidate parameter. If only one candidate parameter is identified in S412, the control-related data output unit 128 outputs the candidate parameter, and if multiple candidate parameters are identified, the control-related data output unit 128 outputs the candidate parameter selected in S416.

[0140] 9 shows an example of a processing flow by the data processing device 100. Here, a processing flow will be described for identifying control parameters capable of reducing the costs of consumption items such as electricity and chemicals while ensuring the quality of water discharged from the water treatment facility 200.

[0141] In S502, the control-related data output unit 128 selects, from the learning model storage unit 114, a learning model corresponding to the environment of the water treatment facility 200 identified by the environment identification unit 122. In S504, the control parameter acquisition unit 132 acquires control parameters to be set in the water treatment facility 200.

[0142] In S506, the change parameter generation unit 134 generates a plurality of change parameters by changing the control parameters acquired in S504 by the control parameter acquisition unit 132. In S508, the cost estimation unit 144 estimates the cost of a consumption object that will be consumed when set in the water treatment facility 200 for each of a plurality of candidate parameters including the control parameters acquired in S504 and the plurality of change parameters generated in S506.

[0143] In S510, the candidate parameter identification unit 142 inputs the candidate parameter with the lowest cost estimated in S508 out of the multiple candidate parameters into a learning model in which the objective variable is the water quality of the water discharged from the water treatment facility 200. In S512, the candidate parameter identification unit 142 determines whether the water quality output from the learning model satisfies the condition. If it is determined that the condition is not satisfied, the process proceeds to S514, and if it is determined that the condition is satisfied, the process proceeds to S516.

[0144] If the determination has been completed for all of the multiple candidate parameters (YES in S514), the process ends, and if the determination has not been completed (NO in S514), the process returns to S510. In S510, the next candidate parameter is input to the learning model in ascending order of cost estimated in S508.

[0145] In S516, the control-related data output unit 128 outputs the candidate parameters that are determined to satisfy the water quality requirements in S512.

[0146] 10 is a schematic diagram showing an example of a hardware configuration of a computer 1200 functioning as the data processing device 100. A program installed on the computer 1200 can cause the computer 1200 to function as one or more "parts" of the device according to the present embodiment, or cause the computer 1200 to execute operations or one or more "parts" associated with the device according to the present embodiment, and / or cause the computer 1200 to execute a process or steps of the process according to the present embodiment. Such a program can be executed by the CPU 1212 to cause the computer 1200 to execute specific operations associated with some or all of the blocks of the flowcharts and block diagrams described herein.

[0147] The computer 1200 according to this embodiment includes a CPU 1212, a RAM 1214, and a graphics controller 1216, which are connected to each other by a host controller 1210. The computer 1200 also includes input / output units such as a communication interface 1222, a storage device 1224, a DVD drive, and an IC card drive, which are connected to the host controller 1210 via an input / output controller 1220. The DVD drive may be a DVD-ROM drive, a DVD-RAM drive, etc. The storage device 1224 may be a hard disk drive, a solid state drive, etc. The computer 1200 also includes a legacy input / output unit such as a ROM 1230 and a keyboard, which are connected to the input / output controller 1220 via an input / output chip 1240.

[0148] The CPU 1212 operates according to a program stored in the ROM 1230 and the RAM 1214, thereby controlling each unit. The graphic controller 1216 acquires image data generated by the CPU 1212 into a frame buffer or the like provided in the RAM 1214 or into itself, and causes the image data to be displayed on the display device 1218.

[0149] The communication interface 1222 communicates with other electronic devices via a network. The storage device 1224 stores programs and data used by the CPU 1212 in the computer 1200. The DVD drive reads programs or data from a DVD-ROM or the like and provides them to the storage device 1224. The IC card drive reads programs and data from an IC card and / or writes programs and data to an IC card.

[0150] The ROM 1230 stores therein a boot program or the like that is executed by the computer 1200 upon activation, and / or a program that depends on the hardware of the computer 1200. The input / output chip 1240 may also connect various input / output units to the input / output controller 1220 via a USB port, a parallel port, a serial port, a keyboard port, a mouse port, and the like.

[0151] The programs are provided by a computer-readable storage medium such as a DVD-ROM or an IC card. The programs are read from the computer-readable storage medium, installed in the storage device 1224, the RAM 1214, or the ROM 1230, which are also examples of computer-readable storage media, and executed by the CPU 1212. Information processing described in these programs is read by the computer 1200, and brings about cooperation between the programs and the various types of hardware resources described above. An apparatus or method may be constructed by implementing operations or processing of information according to the use of the computer 1200.

[0152] For example, when communication is performed between the computer 1200 and an external device, the CPU 1212 may execute a communication program loaded in the RAM 1214 and instruct the communication interface 1222 to perform communication processing based on the processing described in the communication program. Under the control of the CPU 1212, the communication interface 1222 reads transmission data stored in a transmission buffer area provided in the RAM 1214, the storage device 1224, a DVD-ROM, or a recording medium such as an IC card, and transmits the read transmission data to the network, or writes reception data received from the network to a reception buffer area or the like provided on the recording medium.

[0153] Furthermore, the CPU 1212 may cause all or a necessary portion of a file or database stored in an external recording medium such as the storage device 1224, a DVD drive (DVD-ROM), an IC card, etc. to be read into the RAM 1214, and perform various types of processing on the data on the RAM 1214. The CPU 1212 may then write back the processed data to the external recording medium.

[0154] Various types of information, such as various types of programs, data, tables, and databases, may be stored in the recording medium and undergo information processing. The CPU 1212 may perform various types of processing on the data read from the RAM 1214, including various types of operations, information processing, conditional judgment, conditional branching, unconditional branching, information search / replacement, etc., as described throughout this disclosure and specified by the instruction sequence of the program, and write back the results to the RAM 1214. The CPU 1212 may also search for information in a file, database, etc. in the recording medium. For example, when a plurality of entries each having an attribute value of a first attribute associated with an attribute value of a second attribute are stored in the recording medium, the CPU 1212 may search for an entry whose attribute value of the first attribute matches a specified condition from among the plurality of entries, read the attribute value of the second attribute stored in the entry, and thereby obtain the attribute value of the second attribute associated with the first attribute that satisfies a predetermined condition.

[0155] The above-described programs or software modules may be stored in a computer-readable storage medium on the computer 1200 or in the vicinity of the computer 1200. In addition, a recording medium such as a hard disk or a RAM provided in a server system connected to a dedicated communication network or the Internet can be used as a computer-readable storage medium, thereby providing the programs to the computer 1200 via the network.

[0156] The blocks in the flowcharts and block diagrams in the present embodiment may represent stages of a process in which an operation is performed or "parts" of an apparatus responsible for performing the operation. Particular stages and "parts" may be implemented by dedicated circuitry, programmable circuitry provided with computer-readable instructions stored on a computer-readable storage medium, and / or a processor provided with computer-readable instructions stored on a computer-readable storage medium. The dedicated circuitry may include digital and / or analog hardware circuits, and may include integrated circuits (ICs) and / or discrete circuits. The programmable circuitry may include reconfigurable hardware circuits, such as, for example, field programmable gate arrays (FPGAs), programmable logic arrays (PLAs), and the like, including AND, OR, XOR, NAND, NOR, and other logical operations, flip-flops, registers, and memory elements.

[0157] A computer-readable storage medium may include any tangible device capable of storing instructions that are executed by a suitable device, such that a computer-readable storage medium having instructions stored thereon comprises an article of manufacture that includes instructions that can be executed to create means for performing the operations specified in the flowcharts or block diagrams. Examples of computer-readable storage media may include electronic storage media, magnetic storage media, optical storage media, electromagnetic storage media, semiconductor storage media, and the like. More specific examples of computer-readable storage media may include floppy disks, diskettes, hard disks, random access memories (RAMs), read-only memories (ROMs), erasable programmable read-only memories (EPROMs or flash memories), electrically erasable programmable read-only memories (EEPROMs), static random access memories (SRAMs), compact disk read-only memories (CD-ROMs), digital versatile disks (DVDs), Blu-ray disks, memory sticks, integrated circuit cards, and the like.

[0158] The computer readable instructions may include either assembler instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk®, JAVA®, C++, etc., and conventional procedural programming languages ​​such as the “C” programming language or similar programming languages.

[0159] Computer readable instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, or to a programmable circuit, either locally or over a local area network (LAN), a wide area network (WAN), such as the Internet, etc., to cause the processor of the general purpose computer, special purpose computer, or other programmable data processing apparatus, or to a programmable circuit, to execute the computer readable instructions to generate means for performing the operations specified in the flowcharts or block diagrams. Examples of processors include computer processors, processing units, microprocessors, digital signal processors, controllers, microcontrollers, etc.

[0160] Although the present invention has been described above using the embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments. It is clear to those skilled in the art that various modifications and improvements can be made to the above embodiments. It is clear from the description of the claims that such modifications and improvements can also be included in the technical scope of the present invention.

[0161] It should be noted that the order of execution of each process, such as operations, procedures, steps, and stages, in the devices, systems, programs, and methods shown in the claims, specifications, and drawings is not specifically stated as "before" or "prior to," and that the process may be performed in any order unless the output of a previous process is used in a later process. Even if the operational flow in the claims, specifications, and drawings is described using "first," "next," etc. for convenience, it does not mean that the process must be performed in this order. [Explanation of symbols]

[0162] 10 system, 20 treatment facility, 30 sensor, 40 information server, 100 data processing device, 110 memory unit, 112 facility-related data memory unit, 114 learning model memory unit, 120 facility-related data acquisition unit, 122 environment identification unit, 124 learning model generation unit, 126 model selection unit, 128 control-related data output unit, 130 learning model update unit, 132 control parameter acquisition unit, 134 change parameter generation unit, 140 parameter selection unit, 142 candidate parameter identification unit, 144 cost estimation unit, 200 water treatment facility, 210 management device, 211 gritification basin, 212 primary sedimentation basin, 213 reaction tank, 214 final sedimentation basin, 215 sludge treatment facility, 216 chlorine contact tank, 220 management device, 221 receiving tank, 222 mixing basin, 223 Flocculation tank, 224, settling tank, 225, filtration tank, 226, chlorination tank, 227, distribution tank, 228, pumping station, 1200, computer, 1210, host controller, 1212, CPU, 1214, RAM, 1216, graphic controller, 1218, display device, 1220, input / output controller, 1222, communication interface, 1224, storage device, 1230, ROM, 1240, input / output chip

Claims

1. A facility-related data acquisition unit that acquires facility-related data related to the water treatment facility; A learning model storage unit that stores a plurality of types of learning models that receive the facility-related data as input and output control-related data related to control of the water treatment facility; an environment identification unit that identifies an environment of the water treatment facility based on the facility-related data; a model selection unit that selects one learning model corresponding to the environment of the water treatment facility identified by the environment identification unit from the multiple types of learning models stored in the learning model storage unit; a control-related data output unit that inputs the facility-related data from a reference time to the one learning model selected by the model selection unit, and outputs the control-related data from the reference time onward; Equipped with The learning model storage unit stores the learning model generated using facility-related data corresponding to a climate for each climate in an area of ​​the water treatment facility, or the learning model generated using facility-related data corresponding to an event for each event occurring in the water treatment facility, The environment identification unit identifies the climate of the area of ​​the water treatment facility from climate candidates including a plurality of sunny days, cloudy days, rainy days, sunny days after rain, cloudy days after rain, midsummer days, and midsummer days after rain, as the environment of the water treatment facility, or identifies an event occurring at the water treatment facility from event candidates including a plurality of construction works, disasters, abnormal weather, and abnormalities or failures of equipment at the water treatment facility. Data processing device.

2. The data processing device according to claim 1 , wherein the environment specifying unit specifies the environment of the water treatment facility from a plurality of environment candidates for each external factor that affects a treatment capacity of the water treatment.

3. The data processing device according to claim 2 , wherein the environment specifying unit specifies the environment of the water treatment facility from candidate environments including two or more of a normal state, a highly turbid raw water state, a rainy state, a cold state, and a high temperature state.

4. A facility-related data acquisition unit that acquires facility-related data related to a water treatment facility; A learning model storage unit that stores a plurality of types of learning models that receive the facility-related data as input and output control-related data related to control of the water treatment facility; an environment identification unit that identifies an environment of the water treatment facility based on the facility-related data; a model selection unit that selects one learning model corresponding to the environment of the water treatment facility identified by the environment identification unit from the multiple types of learning models stored in the learning model storage unit; a control-related data output unit that inputs the facility-related data from a reference time to the one learning model selected by the model selection unit, and outputs the control-related data from the reference time onward; Equipped with a facility-related data storage unit that stores the facility-related data of the water treatment facility in the environment identified by the environment identification unit in association with the environment; a learning model update unit that generates a new learning model using the facility-related data corresponding to the environment in response to determining that it is time to execute an update process to update the learning model corresponding to the environment after the one learning model corresponding to the environment has been selected by the model selection unit, and when the new learning model is selected based on a comparison result between the new learning model and the one learning model, associates the new learning model with the environment and stores it in the learning model storage unit instead of the one learning model; Further comprising: Data processing device.

5. The data processing device according to claim 4 , further comprising a learning model generation unit that generates the learning model for each environment of the water treatment facility using the facility-related data corresponding to the environment and stores the learning model in the learning model storage unit.

6. The data processing device according to claim 4 or 5, wherein the learning model update unit generates the new learning model using facility-related data corresponding to the environment other than the facility-related data for a predetermined period in the past based on the update timing, and compares the new learning model with the one learning model using the facility-related data for the predetermined period in the past based on the update timing.

7. The data processing device of claim 6, wherein the learning model update unit generates a plurality of different new learning models using the facility-related data corresponding to the environment, and when the learning model update unit selects one of the plurality of different new learning models based on a comparison result between the plurality of different new learning models and the one learning model, the learning model update unit associates the environment with the selected new learning model and stores it in the learning model memory unit instead of the one learning model.

8. The data processing device according to claim 7 , wherein the learning model update unit generates the plurality of different new learning models with different parameters of a learning method.

9. The data processing device according to claim 7 or 8, wherein the learning model update unit generates the plurality of different new learning models by using different types of facility-related data from among the plurality of types of facility-related data stored in the facility-related data storage unit.

10. 10. The data processing device according to claim 7, wherein the learning model update unit generates the plurality of different new learning models by using facility-related data corresponding to different periods among the facility-related data stored in the facility-related data storage unit.

11. The data processing device according to claim 7 , wherein the learning model update unit generates the plurality of different new learning models with different degrees of learning.

12. A facility-related data acquisition unit that acquires facility-related data related to a water treatment facility; A learning model storage unit that stores a plurality of types of learning models that receive the facility-related data as input and output control-related data related to control of the water treatment facility; an environment identification unit that identifies an environment of the water treatment facility based on the facility-related data; a model selection unit that selects one learning model corresponding to the environment of the water treatment facility identified by the environment identification unit from the multiple types of learning models stored in the learning model storage unit; a control-related data output unit that inputs the facility-related data from a reference time to the one learning model selected by the model selection unit, and outputs the control-related data from the reference time onward; Equipped with a control parameter acquisition unit that acquires control parameters to be set in the water treatment facility; a change parameter generating unit for generating a plurality of change parameters by changing the control parameters; a parameter selection unit that selects one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the one learning model selected by the model selection unit; Further comprising: The parameter selection unit a candidate parameter identification unit that identifies, for each of the plurality of candidate parameters, a candidate parameter that satisfies a predetermined condition, the first parameter being input to the one learning model in which a response variable is a first parameter and output; a cost estimating unit that, when a plurality of the candidate parameters are identified by the candidate parameter identifying unit, estimates a cost of a consumption object that will be consumed when each of the plurality of candidate parameters is set in the water treatment facility; having selecting one of the plurality of candidate parameters based on the cost; Data processing device.

13. A facility-related data acquisition unit that acquires facility-related data related to a water treatment facility; A learning model storage unit that stores a plurality of types of learning models that receive the facility-related data as input and output control-related data related to control of the water treatment facility; an environment identification unit that identifies an environment of the water treatment facility based on the facility-related data; a model selection unit that selects one learning model corresponding to the environment of the water treatment facility identified by the environment identification unit from the multiple types of learning models stored in the learning model storage unit; a control-related data output unit that inputs the facility-related data from a reference time to the one learning model selected by the model selection unit, and outputs the control-related data from the reference time onward; Equipped with a control parameter acquisition unit that acquires control parameters to be set in the water treatment facility; a change parameter generating unit for generating a plurality of change parameters by changing the control parameters; a parameter selection unit that selects one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the one learning model selected by the model selection unit; Further comprising: The parameter selection unit a cost estimating unit that estimates the cost of a consumption object that will be consumed when the plurality of candidate parameters are set in the water treatment facility; a candidate parameter identification unit that inputs the plurality of candidate parameters to the one learning model in ascending order of cost until the first parameter to be output by the one learning model having a response variable as a first parameter satisfies a predetermined condition, and identifies a candidate parameter that satisfies the condition; having Data processing device.

14. The data processing device according to claim 12 , wherein the control parameter acquisition section acquires the control parameters included in the control-related data output by the control-related data output section.

15. The data processing device according to claim 12 , wherein the cost estimator estimates a cost of electricity consumed when the candidate parameters are set for the water treatment facility.

16. The data processing device according to claim 12 , wherein the cost estimating unit estimates costs of chemicals consumed when the candidate parameters are set in the water treatment facility.

17. 17. The data processing device according to claim 12, wherein the first parameter is the water quality of water discharged from the water treatment facility, and the predetermined condition is satisfied when the water quality is better than a predetermined threshold value.

18. 17. The data processing device according to claim 12, wherein the first parameter is a moisture content of sludge in the water treatment facility, and the predetermined condition is satisfied when the moisture content is lower than a predetermined threshold value.

19. a facility-related data acquisition unit that acquires facility-related data related to a treatment facility; a learning model storage unit that stores a plurality of types of learning models that receive the facility-related data as an input and output control-related data related to control of the treatment facility; an environment identification unit that identifies an environment of the treatment facility based on the facility-related data; A model selection unit that selects one learning model corresponding to the environment of the treatment facility identified by the environment identification unit from the multiple types of learning models stored in the learning model storage unit; a control-related data output unit that inputs the facility-related data from a reference time to the one learning model selected by the model selection unit, and outputs the control-related data from the reference time onward; Equipped with The learning model storage unit stores the learning model generated using facility-related data corresponding to a climate for each climate in an area of ​​the treatment facility, or the learning model generated using facility-related data corresponding to an event for each event occurring in the treatment facility, The environment identification unit identifies a climate of the area of ​​the treatment facility from candidate climates including a plurality of sunny days, cloudy days, rainy days, sunny days after rain, cloudy days after rain, midsummer days, and midsummer days after rain, as the environment of the treatment facility, or identifies an event occurring in the treatment facility from candidate events including a plurality of construction, disasters, abnormal weather, and abnormalities or failures of equipment of the treatment facility. Data processing device.

20. A facility-related data acquisition unit that acquires facility-related data related to a treatment facility; a learning model storage unit that stores a plurality of types of learning models that receive the facility-related data as an input and output control-related data related to control of the treatment facility; an environment identification unit that identifies an environment of the treatment facility based on the facility-related data; A model selection unit that selects one learning model corresponding to the environment of the treatment facility identified by the environment identification unit from the multiple types of learning models stored in the learning model storage unit; a control-related data output unit that inputs the facility-related data from a reference time to the one learning model selected by the model selection unit, and outputs the control-related data from the reference time onward; Equipped with a facility-related data storage unit that stores the facility-related data of the treatment facility in the environment identified by the environment identification unit in association with the environment; a learning model update unit that generates a new learning model using the facility-related data corresponding to the environment in response to determining that it is time to execute an update process to update the learning model corresponding to the environment after the model selection unit has selected the one learning model corresponding to the environment of the treatment facility, and when the new learning model is selected based on a comparison result between the new learning model and the one learning model, associates the environment with the new learning model and stores it in the learning model storage unit instead of the one learning model; Further comprising: Data processing device.

21. A facility-related data acquisition unit that acquires facility-related data related to a treatment facility; a learning model storage unit that stores a plurality of types of learning models that receive the facility-related data as an input and output control-related data related to control of the treatment facility; an environment identification unit that identifies an environment of the treatment facility based on the facility-related data; A model selection unit that selects one learning model corresponding to the environment of the treatment facility identified by the environment identification unit from the multiple types of learning models stored in the learning model storage unit; a control-related data output unit that inputs the facility-related data from a reference time to the one learning model selected by the model selection unit, and outputs the control-related data from the reference time onward; Equipped with a control parameter acquisition unit that acquires control parameters to be set in the treatment facility; a change parameter generating unit for generating a plurality of change parameters by changing the control parameters; a parameter selection unit that selects one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the one learning model selected by the model selection unit; Further comprising: The parameter selection unit a candidate parameter identification unit that identifies, for each of the plurality of candidate parameters, a candidate parameter that satisfies a predetermined condition, the first parameter being input to the one learning model in which a response variable is a first parameter and output; a cost estimating unit that, when a plurality of the candidate parameters are identified by the candidate parameter identifying unit, estimates a cost of a consumption object that will be consumed when each of the plurality of candidate parameters is set in the treatment facility; having selecting one of the plurality of candidate parameters based on the cost; Data processing device.

22. A facility-related data acquisition unit that acquires facility-related data related to a treatment facility; a learning model storage unit that stores a plurality of types of learning models that receive the facility-related data as an input and output control-related data related to control of the treatment facility; an environment identification unit that identifies an environment of the treatment facility based on the facility-related data; A model selection unit that selects one learning model corresponding to the environment of the treatment facility identified by the environment identification unit from the multiple types of learning models stored in the learning model storage unit; a control-related data output unit that inputs the facility-related data from a reference time to the one learning model selected by the model selection unit, and outputs the control-related data from the reference time onward; Equipped with a control parameter acquisition unit that acquires control parameters to be set in the treatment facility; a change parameter generating unit for generating a plurality of change parameters by changing the control parameters; a parameter selection unit that selects one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the one learning model selected by the model selection unit; Further comprising: The parameter selection unit a cost estimating unit that estimates a cost of a consumption object that will be consumed when the plurality of candidate parameters are set in the treatment facility; a candidate parameter identification unit that inputs the plurality of candidate parameters to the one learning model in ascending order of cost until the first parameter to be output by the one learning model having a response variable as a first parameter satisfies a predetermined condition, and identifies a candidate parameter that satisfies the condition; having Data processing device.

23. A facility-related data acquisition unit that acquires facility-related data related to the water treatment facility; A learning model storage unit that stores a learning model in which the facility-related data is input and control-related data related to control of the water treatment facility is output; a control parameter acquisition unit that acquires control parameters to be set in the water treatment facility; a change parameter generating unit for generating a plurality of change parameters by changing the control parameters; a parameter selection unit that selects one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the learning model stored in the learning model storage unit; A data processing device comprising:

24. The parameter selection unit a candidate parameter identification unit that identifies, for each of the plurality of candidate parameters, a candidate parameter that satisfies a predetermined condition, the first parameter that is input to the learning model in which a response variable is a first parameter and output from the learning model; a cost specifying unit that specifies, when a plurality of the candidate parameters are specified by the candidate parameter specifying unit, a cost of a consumption object that will be consumed when each of the plurality of candidate parameters is set in the water treatment facility; having 24. A data processing apparatus as claimed in claim 23, further comprising means for selecting one of said plurality of candidate parameters based on said cost.

25. The parameter selection unit a cost specification unit that specifies a cost of a consumption object that will be consumed when the plurality of candidate parameters are set in the water treatment facility; a candidate parameter identification unit that inputs the plurality of candidate parameters to the learning model in ascending order of cost until the first parameter output by the learning model having a response variable as a first parameter satisfies a predetermined condition, and identifies a candidate parameter that satisfies the condition; 24. A data processing apparatus according to claim 23, comprising:

26. A program for causing a computer to function as the data processing device according to any one of claims 1 to 25.

27. 1. A computer-implemented data processing method, comprising: a facility-related data acquisition stage for acquiring facility-related data related to the water treatment facility; a learning model storage step of storing a plurality of types of learning models in a learning model storage unit, the learning model having the facility-related data as input and control-related data related to the control of the water treatment facility as output; an environment identification step of identifying an environment of the water treatment facility based on the facility-related data; a model selection step of selecting one learning model corresponding to the environment of the water treatment facility identified in the environment identification step from the plurality of types of learning models stored in the learning model storage unit; a control-related data output step of inputting the facility-related data from a reference time into the one learning model selected in the model selection step, and outputting the control-related data from the reference time onward; Equipped with The learning model storage step stores the learning model generated using facility-related data corresponding to a climate for each climate in the area of ​​the water treatment facility, or the learning model generated using facility-related data corresponding to an event for each event occurring in the water treatment facility, in the learning model storage unit; The environmental identification step identifies the climate of the area of ​​the water treatment facility from candidate climates including a plurality of sunny days, cloudy days, rainy days, sunny days after rain, cloudy days after rain, midsummer days, and midsummer days after rain, as the environment of the water treatment facility, or identifies an event occurring in the water treatment facility from candidate events including a plurality of construction works, disasters, abnormal weather, and abnormalities or failures of equipment of the water treatment facility. Data processing methods.

28. A computer-implemented data processing method comprising: a facility-related data acquisition stage for acquiring facility-related data related to the water treatment facility; a learning model storage step of storing a plurality of types of learning models in a learning model storage unit, the learning model having the facility-related data as input and control-related data related to the control of the water treatment facility as output; an environment identification step of identifying an environment of the water treatment facility based on the facility-related data; a model selection step of selecting one learning model corresponding to the environment of the water treatment facility identified in the environment identification step from the plurality of types of learning models stored in the learning model storage unit; a control-related data output step of inputting the facility-related data from a reference time into the one learning model selected in the model selection step, and outputting the control-related data from the reference time onward; Equipped with a facility-related data storage step of storing the facility-related data of the water treatment facility in the environment identified in the environment identification step in a facility-related data storage unit in association with the environment; a learning model update step of generating a new learning model using the facility-related data corresponding to the environment in response to determining that it is time to execute an update process to update the learning model corresponding to the environment after the one learning model corresponding to the environment has been selected in the model selection step, and, when the new learning model is selected based on a comparison result between the new learning model and the one learning model, associating the new learning model with the environment and storing the new learning model in the learning model storage unit instead of the one learning model; Further comprising: Data processing methods.

29. A computer-implemented data processing method comprising: a facility-related data acquisition stage for acquiring facility-related data related to the water treatment facility; a learning model storage step of storing a plurality of types of learning models in a learning model storage unit, the learning model having the facility-related data as input and control-related data related to the control of the water treatment facility as output; an environment identification step of identifying an environment of the water treatment facility based on the facility-related data; a model selection step of selecting one learning model corresponding to the environment of the water treatment facility identified in the environment identification step from the plurality of types of learning models stored in the learning model storage unit; a control-related data output step of inputting the facility-related data from a reference time into the one learning model selected in the model selection step, and outputting the control-related data from the reference time onward; Equipped with A control parameter acquisition step of acquiring control parameters to be set in the water treatment facility; a modified parameter generating step of generating a plurality of modified parameters by modifying the control parameters; a parameter selection step of selecting one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the one learning model selected in the model selection step; Further comprising: The parameter selection step comprises: a candidate parameter identification step of inputting and outputting, for each of the plurality of candidate parameters, a candidate parameter that satisfies a predetermined condition, the first parameter being input to the one learning model having a response variable as a first parameter; a cost estimation step of estimating a cost of a consumption object to be consumed when each of the plurality of candidate parameters is set in the water treatment facility, when a plurality of the candidate parameters are identified in the candidate parameter identification step; having selecting one of the plurality of candidate parameters based on the cost; Data processing methods.

30. A computer-implemented data processing method comprising: a facility-related data acquisition stage for acquiring facility-related data related to the water treatment facility; a learning model storage step of storing a plurality of types of learning models in a learning model storage unit, the learning model having the facility-related data as input and control-related data related to the control of the water treatment facility as output; an environment identification step of identifying an environment of the water treatment facility based on the facility-related data; a model selection step of selecting one learning model corresponding to the environment of the water treatment facility identified in the environment identification step from the plurality of types of learning models stored in the learning model storage unit; a control-related data output step of inputting the facility-related data from a reference time into the one learning model selected in the model selection step, and outputting the control-related data from the reference time onward; Equipped with A control parameter acquisition step of acquiring control parameters to be set in the water treatment facility; a modified parameter generating step of generating a plurality of modified parameters by modifying the control parameters; a parameter selection step of selecting one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the one learning model selected in the model selection step; Further comprising: The parameter selection step comprises: a cost estimation step of estimating a cost of a consumption object that will be consumed when the plurality of candidate parameters are set in the water treatment facility; a candidate parameter identification step of inputting the plurality of candidate parameters to the one learning model in ascending order of cost until the first parameter to be output by the one learning model having a response variable as a first parameter satisfies a predetermined condition, and identifying a candidate parameter that satisfies the condition; having Data processing methods.

31. A computer-implemented data processing method comprising: a facility-related data acquisition stage for acquiring facility-related data related to a treatment facility; a learning model storage step of storing a plurality of types of learning models in a learning model storage unit, the learning model having the facility-related data as input and control-related data related to the control of the treatment facility as output; an environment identification step for identifying an environment of the treatment facility based on the facility-related data; a model selection step of selecting one learning model corresponding to the environment of the treatment facility identified in the environment identification step from the plurality of types of learning models stored in the learning model storage unit; a control-related data output step of inputting the facility-related data from a reference time into the one learning model selected in the model selection step, and outputting the control-related data from the reference time onward; Equipped with The learning model storage step stores in the learning model storage unit a learning model generated using facility-related data corresponding to a climate for each climate in the area of ​​the treatment facility, or a learning model generated using facility-related data corresponding to an event for each event occurring in the treatment facility, The environmental identification step identifies a climate of the area of ​​the treatment facility from candidate climates including a plurality of sunny days, cloudy days, rainy days, sunny days after rain, cloudy days after rain, midsummer days, and midsummer days after rain, as the environment of the treatment facility, or identifies an event occurring in the treatment facility from candidate events including a plurality of construction works, disasters, abnormal weather, and abnormalities or failures of equipment of the treatment facility. Data processing methods.

32. A computer-implemented data processing method comprising: a facility-related data acquisition stage for acquiring facility-related data related to a treatment facility; a learning model storage step of storing a plurality of types of learning models in a learning model storage unit, the learning model having the facility-related data as input and control-related data related to the control of the treatment facility as output; an environment identification step for identifying an environment of the treatment facility based on the facility-related data; a model selection step of selecting one learning model corresponding to the environment of the treatment facility identified in the environment identification step from the plurality of types of learning models stored in the learning model storage unit; a control-related data output step of inputting the facility-related data from a reference time into the one learning model selected in the model selection step, and outputting the control-related data from the reference time onward; Equipped with a facility-related data storage step of storing the facility-related data of the treatment facility in the environment identified in the environment identification step in a facility-related data storage unit in association with the environment; a learning model update step of generating a new learning model using the facility-related data corresponding to the environment in response to determining that it is time to execute an update process to update the learning model corresponding to the environment after the one learning model corresponding to the environment has been selected in the model selection step, and, when the new learning model is selected based on a comparison result between the new learning model and the one learning model, associating the new learning model with the environment and storing the new learning model in the learning model storage unit instead of the one learning model; Further comprising: Data processing methods.

33. A computer-implemented data processing method comprising: a facility-related data acquisition stage for acquiring facility-related data related to a treatment facility; a learning model storage step of storing a plurality of types of learning models in a learning model storage unit, the learning model having the facility-related data as input and control-related data related to the control of the treatment facility as output; an environment identification step for identifying an environment of the treatment facility based on the facility-related data; a model selection step of selecting one learning model corresponding to the environment of the treatment facility identified in the environment identification step from the plurality of types of learning models stored in the learning model storage unit; a control-related data output step of inputting the facility-related data from a reference time into the one learning model selected in the model selection step, and outputting the control-related data from the reference time onward; Equipped with a control-related data output step of inputting the facility-related data from a reference time into the one learning model selected in the model selection step, and outputting the control-related data from the reference time onward; Equipped with A control parameter acquisition step of acquiring control parameters to be set in the treatment facility; a modified parameter generating step of generating a plurality of modified parameters by modifying the control parameters; a parameter selection step of selecting one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the one learning model selected in the model selection step; Further comprising: The parameter selection step comprises: a candidate parameter identification step of inputting and outputting, for each of the plurality of candidate parameters, a candidate parameter that satisfies a predetermined condition, the first parameter being input to the one learning model having a response variable as a first parameter; a cost estimation step of estimating a cost of a consumption object to be consumed when each of the plurality of candidate parameters is set in the treatment facility when the plurality of candidate parameters is identified in the candidate parameter identification step; having selecting one of the plurality of candidate parameters based on the cost; Data processing methods.

34. A computer-implemented data processing method comprising: a facility-related data acquisition stage for acquiring facility-related data related to a treatment facility; a learning model storage step of storing a plurality of types of learning models in a learning model storage unit, the learning model having the facility-related data as input and control-related data related to the control of the treatment facility as output; an environment identification step for identifying an environment of the treatment facility based on the facility-related data; a model selection step of selecting one learning model corresponding to the environment of the treatment facility identified in the environment identification step from the plurality of types of learning models stored in the learning model storage unit; a control-related data output step of inputting the facility-related data from a reference time into the one learning model selected in the model selection step, and outputting the control-related data from the reference time onward; Equipped with A control parameter acquisition step of acquiring control parameters to be set in the treatment facility; a modified parameter generating step of generating a plurality of modified parameters by modifying the control parameters; a parameter selection step of selecting one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using the one learning model selected in the model selection step; Further comprising: The parameter selection step comprises: a cost estimation step of estimating a cost of a consumption object that will be consumed when the plurality of candidate parameters are set in the treatment facility; a candidate parameter identification step of inputting the plurality of candidate parameters to the one learning model in ascending order of cost until the first parameter to be output by the one learning model having a response variable as a first parameter satisfies a predetermined condition, and identifying a candidate parameter that satisfies the condition; having Data processing methods.

35. 1. A computer-implemented data processing method, comprising: a facility-related data acquisition stage for acquiring facility-related data related to the water treatment facility; A control parameter acquisition step of acquiring control parameters to be set in the water treatment facility; a modified parameter generating step of generating a plurality of modified parameters by modifying the control parameters; a parameter selection step of selecting one of a plurality of candidate parameters including the control parameter and the plurality of change parameters by using a learning model in which the facility-related data is input and control-related data related to the control of the water treatment facility is output; A data processing method comprising:

Images