Contaminated water treatment system, contaminated water treatment method, predictive model generation method, and trained predictive model

Abstract

At least one of the filter pressure loss or filter flow rate in the carbonate precipitation process is estimated, and the operating conditions are adjusted based on the estimation results. [Solution] The contaminated water treatment system 1 includes one or more computers 2 that acquire a first variable indicating the amount of sodium carbonate and caustic soda added, a second variable indicating the amount of carboxylic acid and nitric acid added, a third variable indicating the properties of the contaminated water, and input the first, second, and third variables as input data 24 to a predictive model 20 that has been trained using training data 26, in which the first, second, and third variables are used as explanatory variables and the fourth variable indicating at least one of the pressure loss of the filter or the flow rate of the filter is used as the objective variable, and output the predicted value of the fourth variable as output data 25 from the predictive model 20, and perform a process to determine whether the predicted value is within a predetermined specification range.

Description

Technical Field

[0004] , , , , , , ,

[0003] , , , , ,

[0001] Embodiments of the present invention relate to contaminated water treatment technology.

Background Art

[0002] Contaminated water generated due to a severe accident in a nuclear facility is subjected to a treatment for removing radioactive substances and heavy metals using membrane treatment, iron coprecipitation treatment, carbonate precipitation treatment, and adsorption treatment. Among these, the carbonate precipitation treatment aims to remove divalent metals of the same group that become inhibitory substances, such as alkaline earth metals, for example, calcium and magnesium, in order to treat radioactive strontium in the subsequent adsorption treatment. This carbonate precipitation treatment mainly employs a method of adding sodium carbonate and caustic soda to the contaminated water. The substances precipitated by the treatment are separated as a slurry-like mixture of mud and removed from the contaminated water. This treatment is generally carried out as a pretreatment for the contaminated water sent to the adsorption treatment. However, depending on the treatment conditions, the precipitated substances may not precipitate well, which may cause an increase in the load on the solid-liquid separation operation arranged in the front stage of the adsorption treatment and a risk of poor discharge in the activated carbon tower used for the adsorption treatment.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Patent Document 3

Patent Document 4

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the solid-liquid separation operation preceding the adsorption treatment, filter filtration is generally performed to reduce the impact of suspended components on the adsorption process, such as clogging of adsorption sites. However, it is known that excessive generation of suspended components places a load on the filter. For example, this can lead to increased filter clogging, decreased treatment efficiency due to an increase in the number of filter cleanings, an increase in acidic waste liquid used for filter cleaning, and an increased rate of filter degradation. Clogging can be resolved by dissolving the deposits by adding carboxylic acid and nitric acid. However, since the amount of suspended components flowing into the filter changes depending on the amount of sodium carbonate and caustic soda added in the carbonate precipitation treatment, it is difficult to accurately estimate the amount of carboxylic acid and nitric acid to add. Considering future contaminated water treatment, reducing the load on the filter and optimizing the amount of chemicals used are required, and these are issues that need to be resolved.

[0005] In the field of radioactive waste disposal, known technologies include techniques for determining and optimizing disposal methods based on radioactivity information of radioactive waste. Furthermore, technologies are known for measuring radionuclide concentrations at the inlet and outlet of a treatment device, controlling the water flow rate to the device, and the timing of adding adsorbents to remove contaminated water, thereby improving removal efficiency and reducing radioactive waste. Additionally, systems have been proposed that measure the iron concentration in the liquid and modify the subsequent water flow process, as well as contaminated water treatment methods that analyze both radioactivity and non-radioactive nuclides in the water and construct the optimal treatment process from a pre-prepared set of treatment methods. These known technologies all involve the treatment and removal of radionuclides and the pre-installation of multiple dedicated facilities; they do not involve predicting filter pressure loss from the operating state of the equipment and adjusting operating conditions. Adjusting operating conditions refers to adjusting the amounts of sodium oxide and caustic soda added, and the amounts of carboxylic acid and nitric acid added.

[0006] Embodiments of the present invention have been made in consideration of these circumstances, and aim to estimate at least one of the pressure loss of the filter or the flow rate of the filter in the carbonate precipitation treatment, and to adjust the operating conditions based on the estimation results. [Means for solving the problem]

[0007] The wastewater treatment system according to an embodiment of the present invention includes one or more computers that perform a process, which acquires a first variable indicating the amount of sodium carbonate and caustic soda added in the carbonate precipitation treatment, acquires a second variable indicating the amount of carboxylic acid and nitric acid added in the carbonate precipitation treatment, acquires a third variable indicating the properties of the wastewater in the carbonate precipitation treatment, uses the first, second, and third variables as explanatory variables, and a fourth variable indicating at least one of the pressure loss of the filter or the flow rate of the filter in the carbonate precipitation treatment as an objective variable, inputs the first, second, and third variables as input data to a predictive model trained with learning data, outputs the predicted value of the fourth variable as output data from the predictive model, determines whether the predicted value is within a preset specification range, and continues the carbonate precipitation treatment under predetermined operating conditions if it is determined that the predicted value is within the specification range. [Effects of the Invention]

[0008] According to embodiments of the present invention, at least one of the filter pressure loss or filter flow rate in a carbonate precipitation treatment can be estimated, and the operating conditions can be adjusted based on the estimation results. [Brief explanation of the drawing]

[0009] [Figure 1] A block diagram showing the contaminated water treatment system. [Figure 2] A diagram showing the configuration of the predictive model during training. [Figure 3] A diagram showing the configuration of the predictive model used during operation. [Figure 4] A flowchart illustrating the treatment of contaminated water. [Figure 5] A sequence diagram showing the treatment of contaminated water. [Modes for carrying out the invention]

[0010] The following describes in detail the contaminated water treatment system, the contaminated water treatment method, the predictive model generation method, and embodiments of the trained predictive model, with reference to the drawings.

[0011] Reference numeral 1 in Figure 1 denotes the contaminated water treatment system of this embodiment. The contaminated water treatment method is carried out using this contaminated water treatment system 1.

[0012] The treatment of contaminated water containing radioactive materials includes membrane treatment, iron coprecipitation, carbonate precipitation, and adsorption treatment. Contaminated water treatment system 1 is primarily a system for estimating at least one of the filter pressure loss or filter flow rate in the carbonate precipitation treatment and adjusting the operating conditions based on the estimation results.

[0013] The contaminated water treatment system 1 estimates, for example, the pressure loss of the filter or the flow rate of the filter based on the amounts of sodium carbonate and caustic soda added in the carbonate precipitation treatment, the amounts of carboxylic acid and nitric acid added, and the properties of the contaminated water (e.g., liquid properties), using machine learning. The contaminated water treatment system 1 adjusts and determines the operating conditions based on these estimation results. By continuously monitoring the contaminated water treatment system over time, the contaminated water treatment system 1 can predict and prevent the occurrence of abnormalities in the filter. The machine learning in this embodiment includes supervised learning using training data.

[0014] The configuration of the contaminated water treatment system 1 will be explained with reference to the block diagram shown in Figure 1. Note that the contaminated water treatment system 1 may include components other than those shown in Figure 1, and some of the components shown in Figure 1 may be omitted.

[0015] The contaminated water treatment system 1 includes a control computer 2. The control computer 2 includes an input unit 3, an output unit 4, a communication unit 5, a processing circuit 6, and a storage unit 7.

[0016] Furthermore, the contaminated water treatment system 1 includes a carbonate precipitation treatment device 10, a filter cleaning device 11, a first addition device 12, a second addition device 13, a first acquisition device 14, a second acquisition device 15, a third acquisition device 16, and a fourth acquisition device 17. These devices are connected to a control computer 2 and are controlled. Also, these devices are configured to transmit and receive predetermined information to and from the control computer 2.

[0017] The contaminated water treatment system 1 has hardware resources such as a CPU (Central Processing Unit), GPU (Graphics Processing Unit), ROM (Read Only Memory), RAM (Random Access Memory), HDD (Hard Disk Drive), and SSD (Solid State Drive), and includes a computer in which information processing by software is realized using the hardware resources by the CPU executing various programs.

[0018] Each component of the contaminated water treatment system 1 does not necessarily have to be provided in one computer. For example, one contaminated water treatment system 1 may be realized by a plurality of computers connected to each other via a network. For example, the computer used during machine learning and the computer used during actual operation may each be an individual computer.

[0019] Note that the configuration of the contaminated water treatment system 1 may also be realized as a cloud service. That is, the control computer 2 constituting the contaminated water treatment system 1 may be a server on the cloud. For example, not only the configuration for performing storage processing but also all of the configurations for performing main processing may exist on the cloud, and the user may only perform the setting of the contaminated water treatment system 1 and the confirmation of input and output via an API (Application Programming Interface) or a web browser.

[0020] Input unit 3 receives predetermined information in response to user operations on the control computer 2. This input unit 3 includes input devices such as a mouse, keyboard, and touch panel. In other words, predetermined information is input to the control computer 2 in response to operations on these input devices.

[0021] Output unit 4 outputs predetermined information. The control computer 2 includes a device for displaying images, such as a display that outputs the analysis results. In other words, output unit 4 controls the images displayed on the display. The display may be separate from the computer body or integrated with it. Additionally or alternatively, the control computer 2 may control images displayed on displays of other computers connected via the network.

[0022] The communication unit 5 communicates with other computers via a communication line such as the Internet. While the control computer 2 and the other computers are connected via the Internet, other configurations are also possible. For example, the control computer 2 and the other computers may be connected via a LAN (Local Area Network), WAN (Wide Area Network), or mobile communication network. Furthermore, each device may be connected to the others via a bus.

[0023] The processing circuit 6 is, for example, a circuit equipped with a CPU, GPU, or a dedicated or general-purpose processor. This processor realizes various functions by executing various programs stored in the memory unit 7. The processing circuit 6 may also be composed of hardware such as an FPGA (Field Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit). Various functions can also be realized by this hardware. Furthermore, the processing circuit 6 can realize various functions by combining software processing by the processor and programs with hardware processing.

[0024] The memory unit 7 stores a predetermined program to be executed by the processing circuit 6. The memory unit 7 also stores various information necessary for contaminated water treatment based on information stored in a predetermined database. Furthermore, the memory unit 7 stores various information necessary for generating the trained prediction model 20.

[0025] The carbonate precipitation treatment apparatus 10 is a device that performs carbonate precipitation treatment based on operating conditions determined by the control computer 2. A filter (not shown) is provided in this carbonate precipitation treatment apparatus 10.

[0026] The filter cleaning device 11 is a device that cleans the filter of the carbonate precipitation treatment device 10 based on the operating conditions determined by the control computer 2. The carbonate precipitation treatment device 10 and the filter cleaning device 11 may be an integrated device.

[0027] The first additive device 12 adjusts the amount of sodium carbonate and caustic soda added in the carbonate precipitation treatment based on the operating conditions determined by the control computer 2. The first additive device 12 is a device that adds sodium carbonate and caustic soda to the carbonate precipitation treatment device 10.

[0028] The second additive device 13 adjusts the amount of carboxylic acid and nitric acid added in the carbonate precipitation treatment based on the operating conditions determined by the control computer 2. The second additive device 13 is a device that adds carboxylic acid and nitric acid to the filter cleaning device 11.

[0029] The first acquisition device 14 acquires the amounts of sodium carbonate and caustic soda added in the carbonate precipitation treatment. The first acquisition device 14 is equipped with a measuring instrument (not shown) for measuring the amounts of sodium carbonate and caustic soda added to the carbonate precipitation treatment apparatus 10. The first acquisition device 14 transmits the measured amounts to the control computer 2. This transmitted information is a first variable indicating the amounts of sodium carbonate and caustic soda added in the carbonate precipitation treatment.

[0030] The second data acquisition device 15 acquires the amounts of carboxylic acid and nitric acid added during the carbonate precipitation treatment. The second data acquisition device 15 is equipped with measuring instruments (not shown) for measuring the amounts of carboxylic acid and nitric acid added to the carbonate precipitation treatment device 10. The second data acquisition device 15 transmits the measured amounts to the control computer 2. This transmitted information is a second variable indicating the amounts of carboxylic acid and nitric acid added during the carbonate precipitation treatment.

[0031] The third acquisition device 16 acquires the properties of the contaminated water during the carbonate precipitation treatment. The third acquisition device 16 is equipped with a measuring instrument (not shown) for measuring the properties of the contaminated water introduced into the carbonate precipitation treatment apparatus 10. The third acquisition device 16 transmits information indicating the properties of the measured contaminated water to the control computer 2. This transmitted information is a third variable indicating the properties of the contaminated water during the carbonate precipitation treatment.

[0032] The fourth acquisition device 17 acquires at least one of the filter pressure loss or filter flow rate during the carbonate precipitation treatment. The fourth acquisition device 17 includes a measuring instrument (not shown) for measuring the filter pressure loss and flow rate provided in the carbonate precipitation treatment apparatus 10. The fourth acquisition device 17 transmits the measured filter pressure loss and flow rate to the control computer 2. This transmitted information is a fourth variable indicating at least one of the filter pressure loss or filter flow rate during the carbonate precipitation treatment.

[0033] As shown in Figure 5, the processing circuit 6 comprises a prediction unit 30, a determination unit 31, and an adjustment unit 32. The prediction unit 30 includes a trained prediction model 20 (Figure 3). The prediction unit 30, determination unit 31, and adjustment unit 32 are realized by the processing circuit 6 executing a program stored in the memory unit 7. Note that the prediction unit 30, determination unit 31, and adjustment unit 32 may each be provided in separate computers.

[0034] As shown in Figures 2 and 3, the prediction model 20 is a so-called neural network.

[0035] In this embodiment, the contaminated water treatment system 1 includes artificial intelligence (AI) that performs machine learning. The contaminated water treatment system 1 also includes a deep learning unit that extracts a specific pattern from multiple patterns based on deep learning.

[0036] Analysis using control computer 2 can utilize analytical techniques based on artificial intelligence learning. For example, pre-trained models generated by machine learning using neural networks, pre-trained models generated by other machine learning methods, deep learning algorithms, and mathematical algorithms such as regression analysis can be used. Furthermore, forms of machine learning include clustering and deep learning.

[0037] The contaminated water treatment system 1 includes a control computer 2 equipped with artificial intelligence that performs machine learning. For example, this contaminated water treatment system 1 may consist of one computer equipped with a neural network, or it may consist of multiple computers equipped with neural networks.

[0038] Here, a neural network is a mathematical model that represents the characteristics of brain function through computer simulation. For example, it shows a model in which nodes (artificial neurons) that form a network through synaptic connections change the strength of their synaptic connections through learning, thereby acquiring problem-solving abilities. Furthermore, neural networks acquire problem-solving abilities through deep learning.

[0039] As shown in Figure 2, the pre-trained prediction model 20 of this embodiment includes an input layer 21, a hidden layer 22, and an output layer 23. Input data 24 is input to the input layer 21. The parameters of the hidden layer 22 have been pre-trained using training data 26. The output layer 23 outputs output data 25 that shows the results processed by the hidden layer 22 in response to the input data 24 input to the input layer 21.

[0040] The predictive model 20 is trained using machine learning with training data 26, which takes input data 24 or data mimicking it as input and output data 25 or data mimicking it as output.

[0041] For example, a neural network is provided with an intermediate layer 22 having multiple layers. Each layer of this intermediate layer 22 is composed of multiple units. Furthermore, by pre-training the multi-layer neural network with training data 26, i.e., supervising data, it is possible to automatically extract features from patterns of changes in the state of a circuit or system. In addition, the number of intermediate layers, the number of units, the learning rate, the number of training iterations, and the activation function of the multi-layer neural network can be set arbitrarily on the user interface.

[0042] Furthermore, a reward function may be set for each information item to be learned, and deep reinforcement learning, in which the information item with the highest value is extracted based on the reward function, may be used in the neural network.

[0043] Furthermore, there are various machine learning techniques, such as autoencoders, LSTM (Long Short-Term Memory), SDF (Signed Distance Function), GAN (Generative Adversarial Network), and RNN (Recurrent Neural Network). These techniques may be applied to the machine learning in this embodiment.

[0044] As shown in Figure 3, in the prediction model 20 used in actual operation, the input layer 21 receives the first, second, and third variables as input data 24. The output layer 23 outputs the predicted value of the fourth variable as output data 25. As shown in Figure 2, the parameters of the hidden layer 22 are machine-trained using training data 26, with the first, second, and third variables as explanatory variables and the fourth variable as the target variable.

[0045] Here, the first variable represents the amount of sodium carbonate and caustic soda added in the carbonate precipitation treatment. The second variable represents the amount of carboxylic acid and nitric acid added in the carbonate precipitation treatment. The third variable represents the properties of the contaminated water in the carbonate precipitation treatment. The fourth variable represents at least one of the filter pressure loss or filter flow rate in the carbonate precipitation treatment. Note that the fourth variable included in the output data 25 output from output layer 23 is a predicted value of at least one of the filter pressure loss or filter flow rate in the future.

[0046] As shown in Figure 2, during the learning phase, i.e., when generating the prediction model 20, the control computer 2 first acquires the first, second, third, and fourth variables. Next, the control computer 2 inputs the learning data 26, in which the first, second, and third variables are explanatory variables and the fourth variable is the target variable, into the prediction model 20 to perform machine learning. This machine learning uses data affecting the pressure loss and / or flow rate of the filter as input data 24, and the affected pressure loss and / or flow rate of the filter as output data 25.

[0047] The control computer 2 inputs the first, second, and third variables as input data 24 to the prediction model 20. Here, each node is weighted. Furthermore, if there is an error between the output data 25 output from the prediction model 20 and the fourth variable of the training data 26, the control computer 2 corrects the weight of each node and repeats this process to perform machine learning on the prediction model 20.

[0048] Next, the contaminated water treatment performed by the contaminated water treatment system 1 will be explained using the flowchart in Figure 4 and the sequence diagram in Figure 5. Note that the aforementioned diagrams may be referenced.

[0049] This process is repeated at regular intervals. The repeated process executes the contaminated water treatment method in the contaminated water treatment system 1. This process may also be interrupted and executed while the contaminated water treatment system 1 is performing other main processes. Furthermore, the contaminated water treatment is continuously executed and repeated while the carbonate precipitation treatment device 10 (Figure 5) and the filter cleaning device 11 (Figure 5) are operating.

[0050] As shown in Figure 4, first, in step S1, the prediction unit 30 (Figure 5) of the control computer 2 acquires a first variable, which is data transmitted from the first acquisition device 14 (Figure 5), indicating the amount of sodium carbonate and caustic soda added in the carbonate precipitation treatment. Sodium carbonate and caustic soda are chemicals added from the first addition device 12 (Figure 5) to the carbonate precipitation treatment device 10 (Figure 5). The first acquisition device 14 measures the amount added by the first addition device 12 and transmits this to the prediction unit 30.

[0051] In the next step S2, the prediction unit 30 of the control computer 2 acquires a second variable, which is data transmitted from the second acquisition device 15 (Figure 5), indicating the amount of carboxylic acid and nitric acid added in the carbonate precipitation treatment. Note that carboxylic acid and nitric acid are chemicals for filter cleaning that are added from the second addition device 13 (Figure 5) to the filter cleaning device 11. The second acquisition device 15 measures the amount added by the second addition device 13 and transmits this to the prediction unit 30.

[0052] In the next step S3, the prediction unit 30 of the control computer 2 acquires a third variable, which is data transmitted from the third acquisition device 16 (Figure 5), indicating the properties of the contaminated water in the carbonate precipitation treatment. The third acquisition device 16 acquires measured values ​​indicating the properties of the contaminated water from the carbonate precipitation treatment device 10 and the filter cleaning device 11, and transmits this to the prediction unit 30.

[0053] The properties of contaminated water include at least one piece of information: temperature, pH, oxidation-reduction potential, carbon dioxide concentration, and turbidity. This allows for a proper understanding of the contaminated water's properties and enables adjustment of the additive dosage.

[0054] In the next step S4, the prediction unit 30 of the control computer 2 estimates a predicted value for a fourth variable that indicates at least one of the future pressure loss or flow rate of the filter. Here, the control computer 2 inputs the first, second, and third variables as input data 24 to the prediction model 20 (Figure 3), and causes the prediction model 20 to output the predicted value of the fourth variable as output data 25.

[0055] The predicted values ​​represent the fourth variable in the future, which is expected to change when sodium carbonate and caustic soda are added to the current carbonate precipitation treatment apparatus 10 (Figure 5), and / or when carboxylic acid and nitric acid are added.

[0056] In the next step S5, the determination unit 31 (Figure 5) in the control computer 2 determines whether the predicted value is within a preset specification range. The specification range is set by the user as thresholds indicating its upper and lower limits. If the predicted value is within the specification range (YES in step S5), the process proceeds to step S10. On the other hand, if the predicted value is not within the specification range (NO in step S5), the process proceeds to step S6.

[0057] If the initial predicted value is determined to be within the specified range, the control computer 2 continues the carbonate precipitation process under the default operating conditions. On the other hand, if the predicted value is determined to be outside the specified range, the control computer 2 adjusts the amount of chemicals added. Here, the control computer 2 adjusts the amount of sodium carbonate and caustic soda added in the carbonate precipitation process, and / or the amount of carboxylic acid and nitric acid added, so that the pressure loss of the filter decreases or the flow rate of the filter increases.

[0058] In step S6, which proceeds if the answer in step S5 is NO, it is determined whether the amount of additive cannot be adjusted. For example, this is determined by checking whether the number of times the prediction model 20 has output a predicted value is n. This n number is set in advance by the user. For example, n can be 2 or n can be 100. If the predicted value is not within the specified range even after step S4 has been repeated a predetermined number of times, the amount of additive can be considered unadjustable. If the amount of additive cannot be adjusted (if the answer in step S6 is YES), the process proceeds to step S12. On the other hand, if the amount of additive is not unadjustable (if the answer in step S6 is NO), the process proceeds to step S7.

[0059] If the result in step S6 is NO, in step S7, the determination unit 31 notifies the adjustment unit 32 (Figure 5) of an error. At this point, the control computer 2 notifies the user of a warning. For example, the control computer 2 displays a predetermined image indicating the warning on the display, which serves as the output unit 4 (Figure 1).

[0060] In the next step S8, the adjustment unit 32 of the control computer 2 adjusts the amount of sodium carbonate and caustic soda to be added. Here, the adjustment unit 32 acquires at least one piece of information, such as the weight, mass, and concentration of the added chemicals, and reflects the acquired information in the adjustment of the amount to be added. In other words, the adjustment unit 32 measures the amount of sodium carbonate and caustic soda to be added and reflects this in the input amount of the first adding device 12. In this way, the amount of sodium carbonate and caustic soda to be added can be optimized.

[0061] In the next step S9, the adjustment unit 32 of the control computer 2 adjusts the amounts of carboxylic acid and nitric acid added. Here, the adjustment unit 32 acquires information on at least one of the following as additives: carboxylic acid, dicarboxylic acid, tricarboxylic acid, oxycarboxylic acid, aminocarboxylic acid, nitric acid, formic acid, and amide sulfuric acid, and reflects the acquired information in adjusting the amount added. In other words, the adjustment unit 32 measures the amount of carboxylic acid and nitric acid added and reflects this in the input amount of the second additive device 13. In this way, the amounts of carboxylic acid and nitric acid added can be optimized. After executing step S9, the control computer 2 returns to step S4.

[0062] In other words, the control computer 2 issues a warning if it determines that the predicted value is outside the specified range. This allows the user to be aware of the warning. Furthermore, the control computer 2 adjusts at least one of the amounts of sodium carbonate and caustic soda added, or the amounts of carboxylic acid and nitric acid added. The control computer 2 inputs the adjusted first, second, and third variables into the prediction model 20 and outputs the predicted value from the prediction model 20. The control computer 2 then repeats the adjustment of at least one of the amounts of sodium carbonate and caustic soda added, or the amounts of carboxylic acid and nitric acid added, until the predicted value is within the specified range. In this way, the adjustment can be made to achieve the optimal amount of additive. In other words, the operating conditions can be optimized. The adjustment unit 32 manipulates the explanatory variables in a simulated manner to determine the conditions under which the filter's pressure loss and / or flow rate reach their optimal levels.

[0063] If the prediction unit 30 estimates a new predicted value, it transmits this new predicted value to the determination unit 31. The determination unit 31 performs a new determination, and if the predicted value is within the specified range, it transmits a notification to the adjustment unit 32 indicating that it is normal. The adjustment unit 32 transmits information indicating the optimal amount to the first adding device 12 and the second adding device 13, respectively. Based on this transmitted information, the first adding device 12 adds the chemical to the carbonate precipitation treatment device 10, and the second adding device 13 adds the chemical to the filter cleaning device 11.

[0064] In step S12, which proceeds if the answer in step S6 is YES, the control computer 2 notifies the user that adjustment is not possible. In other words, the control computer 2 notifies the user that adjustment is not possible if it is not possible to adjust at least one of the amounts of sodium carbonate and caustic soda added, or the amounts of carboxylic acid and nitric acid added. For example, the control computer 2 displays a predetermined image indicating that adjustment is not possible on the display, which serves as the output unit 4 (Figure 1). In this way, the user can understand that adjustment is not possible. The control computer 2 then terminates the contaminated water treatment.

[0065] In step S10, which proceeds if the answer to step S5 is YES, the control computer 2 acquires a fourth variable, which is data transmitted from the fourth acquisition device 17 (Figure 1) and indicates at least one of the current pressure loss of the filter or the flow rate of the filter.

[0066] In the next step S11, the control computer 2 performs machine learning on the predictive model 20 (Figure 2) again using training data 26 (Figure 2) which includes at least one of the first, second, third, and fourth variables acquired during the carbonate precipitation treatment. In this way, each time the carbonate precipitation treatment is performed, additional machine learning can be performed in accordance with the actual operating conditions, thereby improving the accuracy of the predictive model 20. Then, the control computer 2 terminates the contaminated water treatment.

[0067] While the flowchart above illustrates a configuration where each step is executed sequentially, the order of each step is not necessarily fixed, and the order of some steps may be reversed. Furthermore, some steps may be executed in parallel with others. Also, the steps included in the flowchart above represent at least a subset of the steps, and other steps may be included in the flowchart.

[0068] The arrows in the sequence diagram above are just one example of how the process flow can be represented, and there may be other processing flows besides those indicated by the arrows. Furthermore, the order of operations is not necessarily fixed, and the order of operations may be reversed. Also, some operations may be executed in parallel with others.

[0069] The aforementioned contaminated water treatment system 1 comprises a control device, a memory device, an output device, an input device, and a communication interface. Here, the control device includes a highly integrated processor such as a CPU (Central Processing Unit), GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), or a dedicated chip. The memory device includes ROM (Read Only Memory), RAM (Random Access Memory), HDD (Hard Disk Drive), SSD (Solid State Drive), etc. The output device includes a display panel, a head-mounted display, a projector, a printer, etc. The input device includes a mouse, a keyboard, a touch panel, etc. This contaminated water treatment system 1 can be implemented with a hardware configuration using a standard computer.

[0070] Furthermore, the program or learned model to be executed in the aforementioned contaminated water treatment system 1 is provided pre-loaded into ROM or similar media. Additionally or alternatively, this program or learned model is provided as an installable or executable file stored on a computer-readable, non-temporary storage medium. This storage medium includes CD-ROMs, CD-Rs, memory cards, DVDs, flexible disks (FDs), and the like.

[0071] Furthermore, the program or trained model executed in this contaminated water treatment system 1 may be stored on a computer connected to a network such as the Internet and provided for download via the network. In other words, the program or trained model may be provided from cloud computing resources. Alternatively, a server on the cloud may execute the program or trained model, and only the processing results may be provided via the cloud. In addition, this contaminated water treatment system 1 can also be configured by connecting and combining separate modules, each independently performing the function of its components, via a network or dedicated line.

[0072] According to the embodiments described above, the contaminated water treatment system 1 includes a predictive model 20 that has been trained using training data 26, with at least one of the filter pressure loss or filter flow rate in the carbonate precipitation treatment as the target variable. This allows for the estimation of at least one of the filter pressure loss or filter flow rate in the carbonate precipitation treatment, and the adjustment of operating conditions based on the estimation results.

[0073] While several embodiments of the present invention have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These embodiments can be carried out in a variety of other forms, and various omissions, substitutions, modifications, and combinations are possible without departing from the spirit of the invention. These embodiments or their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. Where there is a singular expression, it does not exclude plurals unless clearly indicated otherwise in the context. Furthermore, conjunctions such as "and," "or," and "and / or" are inclusive unless clearly indicated otherwise in the context. [Explanation of Symbols]

[0074] 1...Contaminated water treatment system, 2...Control computer, 3...Input unit, 4...Output unit, 5...Communication unit, 6...Processing circuit, 7...Storage unit, 10...Carbonate precipitation treatment device, 11...Filter cleaning device, 12...First addition device, 13...Second addition device, 14...First acquisition device, 15...Second acquisition device, 16...Third acquisition device, 17...Fourth acquisition device, 20...Prediction model, 21...Input layer, 22...Intermediate layer, 23...Output layer, 24...Input data, 25...Output data, 26...Training data, 30...Prediction unit, 31...Decision unit, 32...Adjustment unit.

Claims

1. The first variable, which indicates the amount of sodium carbonate and caustic soda added in the carbonate precipitation treatment, is obtained. A second variable is obtained that indicates the amount of carboxylic acid and nitric acid added in the carbonate precipitation treatment. A third variable indicating the properties of the contaminated water in the carbonate precipitation treatment is obtained, A predictive model, trained using training data in which the first, second, and third variables are explanatory variables and a fourth variable representing at least one of the pressure loss of the filter or the flow rate of the filter in the carbonate precipitation process is the target variable, is input with the first, second, and third variables as input data, and the predicted value of the fourth variable is output from the predictive model as output data. Determine whether the predicted value falls within the pre-set specification range. If it is determined that the predicted value is within the specified range, the carbonate precipitation treatment is continued under the default operating conditions. A system comprising one or more computers that perform processing, Contaminated water treatment system.

2. The aforementioned computer, If it is determined that the predicted value is not within the specified range, a warning is issued. Adjust the amount of sodium carbonate and caustic soda added, or at least one of the amount of carboxylic acid and nitric acid added. The adjusted first, second, and third variables are input into the prediction model, and the prediction values ​​are output from the prediction model. The amount of sodium carbonate and caustic soda added, or the amount of carboxylic acid and nitric acid added, is repeatedly adjusted until the predicted value falls within the specified range. It is what executes the process. The contaminated water treatment system according to claim 1.

3. The aforementioned computer, If it is not possible to adjust the amount of sodium carbonate and caustic soda added, or the amount of carboxylic acid and nitric acid added, a notification will be given indicating that adjustment is not possible. It is what executes the process. The contaminated water treatment system according to claim 2.

4. The aforementioned computer, When adjusting the amount of sodium carbonate and caustic soda added, obtain at least one piece of information regarding the weight, mass, and concentration of the added reagents of the added agents. The acquired information is used to adjust the amount of additive. It is what executes the process. A contaminated water treatment system according to claim 2 or claim 3.

5. The aforementioned computer, When adjusting the amount of the carboxylic acid and the nitric acid added, information is obtained on at least one of the following as an additive: the carboxylic acid, dicarboxylic acid, tricarboxylic acid, oxycarboxylic acid, aminocarboxylic acid, nitric acid, formic acid, and amide sulfuric acid. The acquired information is used to adjust the amount of additive. It is what executes the process. A contaminated water treatment system according to claim 2 or claim 3.

6. The aforementioned computer, The machine learning of the prediction model is performed again using the training data, which includes at least one of the first, second, third, and fourth variables obtained during the carbonate precipitation treatment. It is what executes the process. A contaminated water treatment system according to claim 1 or claim 2.

7. The properties of the contaminated water include at least one piece of information: temperature, pH, oxidation-reduction potential, carbon dioxide concentration, and turbidity of the contaminated water. A contaminated water treatment system according to claim 1 or claim 2.

8. A first acquisition device for acquiring the amounts of sodium carbonate and caustic soda added, A second acquisition apparatus for acquiring the amounts of the carboxylic acid and the nitric acid added, A third acquisition device for acquiring the properties of the contaminated water, A fourth acquisition device for acquiring at least one of the pressure loss of the filter or the flow rate of the filter, A first additive device that adjusts the amount of sodium carbonate and caustic soda added based on the operating conditions determined by the computer, A second additive device that adjusts the amount of carboxylic acid and nitric acid added based on the operating conditions determined by the computer, Equipped with, A contaminated water treatment system according to claim 1 or claim 2.

9. The first variable, which indicates the amount of sodium carbonate and caustic soda added in the carbonate precipitation treatment, is obtained. A second variable is obtained that indicates the amount of carboxylic acid and nitric acid added in the carbonate precipitation treatment. A third variable indicating the properties of the contaminated water in the carbonate precipitation treatment is obtained, A predictive model, trained using training data in which the first, second, and third variables are explanatory variables and a fourth variable representing at least one of the pressure loss of the filter or the flow rate of the filter in the carbonate precipitation process is the target variable, is input with the first, second, and third variables as input data, and the predicted value of the fourth variable is output from the predictive model as output data. Determine whether the predicted value falls within the pre-set specification range. If it is determined that the predicted value is within the specified range, the carbonate precipitation treatment is continued under the default operating conditions. The process is performed by one or more computers. Methods for treating contaminated water.

10. A first variable indicating the amount of sodium carbonate and caustic soda added in the carbonate precipitation treatment, a second variable indicating the amount of carboxylic acid and nitric acid added in the carbonate precipitation treatment, and a third variable indicating the properties of the contaminated water in the carbonate precipitation treatment are input to the prediction model as input data, and the prediction value of a fourth variable indicating at least one of the pressure loss of the filter or the flow rate of the filter in the carbonate precipitation treatment is output from the prediction model as output data. Determine whether the predicted value falls within the pre-set specification range. If it is determined that the predicted value is within the specified range, the carbonate precipitation treatment is continued under the default operating conditions. A method for generating the prediction model used in one or more computers that perform processing, Obtain the earlier 1st variable, earlier 2nd variable, earlier 3rd variable, and earlier 4th variable, Machine learning is performed by inputting training data, in which the first, second, and third variables are used as explanatory variables and the fourth variable is used as the target variable, into the prediction model. The computer performs the processing. Methods for generating predictive models.

11. A first variable indicating the amount of sodium carbonate and caustic soda added in the carbonate precipitation treatment, a second variable indicating the amount of carboxylic acid and nitric acid added in the carbonate precipitation treatment, and a third variable indicating the properties of the contaminated water in the carbonate precipitation treatment are input to the prediction model as input data, and the prediction value of a fourth variable indicating at least one of the pressure loss of the filter or the flow rate of the filter in the carbonate precipitation treatment is output from the prediction model as output data. Determine whether the predicted value falls within the pre-set specification range. If it is determined that the predicted value is within the specified range, the carbonate precipitation treatment is continued under the default operating conditions. The prediction model is used in one or more computers that perform the processing, An input layer into which the first variable, the second variable, and the third variable are input as input data, An output layer that outputs the predicted value of the fourth variable as output data, An intermediate layer whose parameters have been machine-trained using training data in which the first, second, and third variables are explanatory variables and the fourth variable is the target variable, including, A pre-trained predictive model.

Images