Processing device and processing method

Abstract

A processing device 1 comprises: an extraction unit 21 that extracts variables from differential equations of state variables constituting a marine ecosystem model; and a determination unit 22 that determines each term included in the differential equations as a measurement item that may depend on a newly considered variable.

Description

Processing Apparatus, and Processing Method

[0001] The present disclosure relates to a processing apparatus and a processing method.

[0002] Patent Document 1 generates a marine ecosystem model corresponding to the environment from state variables input by a user. The marine ecosystem model infers changes in state variables such as substance amounts or organism amounts over time. In Patent Document 1, a user can modify the substance flow of the marine ecosystem model as necessary.

[0003] For example, when applying an existing marine ecosystem model to an environment different from the existing marine ecosystem model, the substance flow of the marine ecosystem model may be examined with reference to user knowledge or prior papers. Since the environment is different from that of the existing marine ecosystem model, it is necessary to add new elements.

[0004] International Publication No. 2024 / 228229

[0005] However, in order to add new elements, it may be necessary to conduct a huge number of experiments and acquire data in consideration of the combined effects of environmental factors or interactions between organisms.

[0006] The present disclosure has been made in view of the above circumstances, and an object of the present disclosure is to provide a technology capable of reducing the burden of experiments for adding new elements to an existing marine ecosystem model.

[0007] A processing apparatus according to one aspect of the present disclosure includes an extraction unit that extracts variables from differential equations of state variables constituting a marine ecosystem model, and a determination unit that determines each term included in the differential equations as a measurement item that may depend on a newly considered variable.

[0008] In a processing method according to one aspect of the present disclosure, a computer extracts variables from differential equations of state variables constituting a marine ecosystem model, and determines each term included in the differential equations as a measurement item that may depend on a newly considered variable.

[0009] According to the present disclosure, it is possible to provide a technology capable of reducing the burden of experiments for adding new elements to an existing marine ecosystem model.

[0010] Figure 1 is a diagram illustrating the functional blocks of the processing apparatus of this disclosure. Figure 2 is a diagram illustrating the system configuration of the estimation system and the functional blocks of the estimation device that provides differential equation data. Figure 3 is a diagram illustrating an example of a marine ecosystem model. Figure 4 is a diagram illustrating an example of a differential equation for a marine ecosystem model. Figure 5 is a diagram illustrating an example of optimal conditions. Figure 6 is a diagram illustrating an example of the conditions of a tank used in an experiment. Figure 7 is a flowchart illustrating the design process. Figure 8 is a flowchart illustrating the update process. Figure 9 is a diagram illustrating the hardware configuration of the computer used in the processing apparatus.

[0011] Embodiments of this disclosure will be described below with reference to the drawings. In the drawings, the same parts are denoted by the same reference numerals and their descriptions are omitted.

[0012] General marine ecosystem models may differ depending on the area or phenomenon being studied. For example, the parameters used in a marine ecosystem model may vary depending on the area or phenomenon, due to factors such as chemical or physical influences that need to be considered in a particular ocean area. Furthermore, even if a marine ecosystem model aims to reproduce a specific phenomenon, the phenomenon itself may be partially unknown, or the material flows may be unclear.

[0013] When generating a marine ecosystem model for a particular sea area, new elements may be added to an existing model. For example, if the target area differs from the existing model, or if the phenomena to be reproduced differ, elements representing these differences are added to the existing model.

[0014] For example, consider a scenario where we need to construct a new ecosystem model that includes macroalgae and takes into account the effects of the hydrogen ion concentration (pH: potential of hydrogen). When adding a macroalgae model equation to an existing model, it can be unclear which part of the macroalgae model equation to add what function to reproduce the effects of the hydrogen ion concentration.

[0015] The differential equation describing the time evolution of the amount of substance in macroalgae is generally expressed as follows: photosynthesis term - respiration term - extracellular secretion term - death term - grazing term. However, it is unclear which of these terms in the differential equation is affected by the hydrogen ion concentration.

[0016] Therefore, adding new elements requires extensive experimentation and data acquisition.

[0017] The processing device 1 shown in Figure 1 reduces the experimental burden of adding new elements to a new marine ecosystem model. For example, the processing device 1 reduces the experimental burden of identifying terms affected by newly considered variables, such as the hydrogen ion concentration. Furthermore, the processing device 1 updates the differential equations of the marine ecosystem model based on the results obtained from the experiments. The hydrogen ion concentration is a physical quantity that represents the degree of acidity and baseness, and is sometimes expressed as pH.

[0018] (Estimation System) The processing device 1 of this disclosure shown in Figure 1 is used in the estimation system 2 shown in Figure 2. The estimation system 2 comprises the processing device 1, the estimation device 100, and the ecosystem database 104.

[0019] The estimation device 100 generates a marine ecosystem model and uses the generated model to predict the quantity of state variables for each time period.

[0020] The estimation device 100 includes differential equation data F, model template data 111, condition data 112, and result data 113, as well as the functions of data acquisition unit 121, model generation unit 122, condition acquisition unit 124, prediction unit 125, and display unit 126. Each data is stored in a storage device such as memory 902 or storage 903. Each function is implemented in the CPU 901.

[0021] The differential equation data F identifies the marine ecosystem model used by the estimation device 100.

[0022] A marine ecosystem model, as shown in Figure 3, for example, represents the balance of substances between nutrients (N), phytoplankton (P), zooplankton (Z), and detritus (D). Nutrients, phytoplankton, zooplankton, and detritus are state variables whose amounts change depending on the state of the ecosystem. A marine ecosystem model is identified by differential equations, as shown in Figure 4, for example. Each differential equation shows the time-dependent increase or decrease of each state variable per unit time.

[0023] The model template data 111 is data that stores the template data for each model generated by the model generation unit 122. The model template data 111, for example, associates the number of classifications for each state variable in the model drawn by each template with the flow between the state variables and the differential equation that identifies the balance of the flow between the state variables.

[0024] Condition data 112 stores the conditions referenced to estimate the amount of substance and biomass from the model. Condition data 112 includes data such as constants, initial values, and prediction time conditions used in the model.

[0025] The result data 113 consists of data on the amount of substance and biomass of each state variable predicted by the prediction unit 125.

[0026] The acquisition unit 121 acquires state variables that constitute the ecosystem model from a user terminal (not shown) or the like. The acquisition unit 121 inputs the acquired state variables to the model generation unit 122.

[0027] The model generation unit 122 generates a model from the state variables acquired by the acquisition unit 121 and modifies the model according to instructions from a user terminal or the like. First, the model generation unit 122 identifies the taxonomic position of the state variables acquired by the acquisition unit 121, for example by referring to the ecosystem database 104. Next, the model generation unit 122 acquires the model template to be generated from the model template data 111 based on the number of taxonomic positions for each classification that were identified. The model generation unit 122 generates a model by incorporating the state variables into the acquired model template. The model generation unit 122 generates differential equation data F corresponding to the generated model.

[0028] The condition acquisition unit 124 acquires data on conditions used in model prediction, such as the initial values ​​of variables used in the flow between state variables.

[0029] The prediction unit 125 uses the generated model and conditions such as initial values ​​acquired by the condition acquisition unit 124 to predict changes in the ecosystem model. The prediction unit 125 predicts the amount of state variables for each time period and records it in the result data 113.

[0030] The display unit 126 displays the result data 113 on a user terminal or the like.

[0031] (Processing device) The processing device 1 shown in Figure 1 acquires differential equation data F that identifies the differential equations in the marine ecosystem model generated by the estimation device 100 shown in Figure 2, thereby reducing the experimental burden of identifying terms affected by newly considered variables.

[0032] As shown in Figure 1, the processing unit 1 includes differential equation data F, variable data 11, measurement item data 12, optimal condition data 13, measurement data 16, and additional function data 17, as well as the functions of an extraction unit 21, a determination unit 22, and an update unit 30. Each data is stored in a storage device such as a memory 902 or storage 903. Each function is implemented in the CPU 901.

[0033] The variable data 11, measurement item data 12, optimal condition data 13, extraction unit 21, and determination unit 22 function as the design unit 20. The design unit 20 reduces the experimental burden by identifying terms in the differential equations showing the time evolution of the amount of macroalgae that are affected by newly considered variables such as the hydrogen ion concentration. The update unit 30 updates the differential equations of the marine ecosystem model based on the results obtained from the experiment.

[0034] The differential equation data F identifies the differential equations of the state variables that constitute the marine ecosystem model. The differential equation of this disclosure is, for example, shown in equation (1). Equation (1) is the differential equation for the biomass of macroalgae, where the state variable is α. In equation (1), α and β are coefficients. Water temperature T and light intensity I are variables. Although not shown in equation (1), a function of nutrients follows the right-hand side of the photosynthesis term in equation (1). The function of nutrients has inorganic nitrogen concentration and inorganic phosphorus concentration as variables.

[0035]

[0036] The variable data 11 identifies the variables in the differential equation. For equation (1), the variable data 11 includes, for example, water temperature T and light intensity I.

[0037] Measurement item data 12 identifies the measurement items in the experiment. The measurement items are each term included in the differential equation. Each term included in the differential equation may depend on newly considered variables, and therefore becomes a measurement item. For equation (1), measurement item data 12 includes the photosynthesis term, the extracellular secretion term, the death term, and the grazing term.

[0038] The optimal condition data 13 associates the optimal conditions with each variable. The optimal condition data 13 may associate the optimal conditions with each individual variable within each variable. The optimal condition data 13 may associate the optimal conditions with some of the variables within each variable.

[0039] The measurement data 16 is data that correlates the measured values ​​of each measurement item presented by the design department 20.

[0040] The additional function data 17 identifies the functions to be added to the differential equations for the terms of the measurement items that depend on the newly considered variables. The functions to be added depend on the newly considered variables.

[0041] The extraction unit 21 extracts variables from the differential equations of state variables constituting the marine ecosystem model. The extraction unit 21 uses, for example, AI (Artificial Intelligence) to extract variables included in the differential equations specified by the differential equation data F. Specifically, the extraction unit 21 excludes coefficients from the parameters included in the differential equations to extract the variables.

[0042] For Equation (1), the extraction unit 21 extracts variables such as the biomass ALG of macroalgae, water temperature T, inorganic nitrogen concentration N, and inorganic phosphorus concentration P as variables. The extraction unit 21 outputs variable data 11 including each of the extracted variables.

[0043] The determination unit 22 determines and outputs the conditions for the experiment. The determination unit 22 outputs measurement item data 12 that specifies measurement items, and optimal condition data 13 that specifies the optimal conditions for the variables.

[0044] The determination unit 22 determines each term included in the differential equations as a measurement item that may depend on a newly considered variable. The determination unit 22 uses, for example, AI to extract the terms included in the differential equations of the state variables. Here, a term is a part separated by + or - in the differential equation.

[0045] The determination unit 21 extracts, as terms for Equation (1), a photosynthesis term, a respiration term, an extracellular secretion term, a mortality term, and a predation term. The determination unit 22 outputs measurement item data 12 including each of the extracted terms.

[0046] The determination unit 22 further determines optimal conditions. The determination unit 22 associates the determined optimal conditions with the names of the variables to which the optimal conditions are applied, and outputs the result to the optimal condition data 13.

[0047] The determination unit 22 takes a term including a variable in the differential equation as a function of that variable, and determines the value of the variable at which the value of the function is maximum as the optimal condition for the variable. If the value of the function does not converge, the determination unit 22 does not determine the optimal condition for the variable of that function.

[0048] The determination unit 22 identifies the portion of the differential equation containing that variable as a single function for each variable. For equation (1), the determination unit 22 outputs the set of functions shown in equation (2) for the biomass ALG of macroalgae, water temperature T, inorganic nitrogen concentration N, and inorganic phosphorus concentration P.

[0049]

[0050] Each set of functions in equation (2) comprises five sets of functions f1-f5. Function f1 takes the biomass of macroalgae ALG as the variable. Function f2 takes the water temperature T as the variable. Function f3 takes the light intensity I as the variable. Function f4 takes the inorganic nitrogen concentration N as the variable. Function f5 takes the inorganic phosphorus concentration P as the variable. Functions f4 and f5 are functions of nutrients.

[0051] For example, the set of functions for water temperature T is the part that contains only one variable, water temperature T. The function of water temperature T is the function f2 in equation (2). In equation (1), the function following the right-hand side of function f2 is the function f3 of light intensity I.

[0052] Next, the determination unit 22 calculates the values ​​of each variable when the value of each function is maximized. Here, the range of the variables is the range that can occur in the sea area where the ecosystem model is constructed. The determination unit 22 outputs the function and optimal conditions for water temperature T and light intensity I as variables, for example, as shown in Figure 5. Note that among the set of functions shown in equation (2), the function for the biomass ALG of macroalgae is a function of the left side of the differential equation, and is therefore excluded from the items to be controlled as conditions.

[0053] Figure 5 shows that the water temperature T takes values ​​in the range of 15 degrees Celsius to 20 degrees Celsius. It also shows that the function f2 of water temperature in equation (2) takes its maximum value when the water temperature T is 20 degrees Celsius. The light intensity I takes values ​​in the range of 70 to 100 degrees Celsius. It also shows that the function f3 of light intensity I in equation (2) takes its maximum value when the light intensity I is 100.

[0054] Furthermore, if the determination unit 22 determines that a variable is unnecessary as an experimental condition item if the function value diverges without converging and the value of the variable at which it reaches its maximum cannot be identified, it determines that the variable diverges without converging.

[0055] The experimental conditions are output as a result of processing by the determination unit 22.

[0056] The experimental conditions indicate that a control experiment will be conducted using multiple tanks with the optimal conditions identified in the optimal condition data 13. For example, as shown in Figure 6, when one control tank and two treatment tanks are prepared, the water temperature T and light intensity I are set to the same conditions. The hydrogen ion concentration is set to approximately 8.2 in the control tank and to 7.5 and 6.0 in the two treatment tanks.

[0057] The experimental conditions indicate that the following will be measured: (1) time-series data of the state variable and (2) time-series data of each item specified in the measurement item data 12. Furthermore, the measurement targets will also include (3) time-series data of each variable set in the optimal condition data 13. This is to confirm that each tank is controlled under optimal conditions.

[0058] The experiment is conducted according to the experimental conditions determined by the determination unit 22, and measurement data 16 is generated.

[0059] The update unit 30 acquires measurement data 16, which is the result of the experiment. The measurement data 16 includes (1) time series data of the state variable, (2) time series data of each item specified in the measurement item data 12, and (3) time series data of each variable set in the optimal condition data 13. The update unit 30 uses (1) the time series data of the state variable and (2) the time series data of each item specified in the measurement item data 12 to identify additional functions to be added to the differential equation.

[0060] The update unit 30 identifies terms among the measurement items that show a significant difference in response to the change in the newly considered variable. The update unit 30 adds the function in which the identified term depends on the variable to the differential equation. Here, the update unit 30 defines the function in which the identified term changes depending on the newly considered variable as the additional function.

[0061] The update unit 30 identifies that, among the various measurement items, the photosynthesis term depends on the hydrogen ion concentration. The update unit 30 generates an additional function that depends on the hydrogen ion concentration and adds it to the photosynthesis term as shown in equation (3).

[0062]

[0063] The update unit 30 obtains three or more hydrogen ion index and photosynthesis rate measurements from the measurement data 16. The update unit 30 identifies a formulated function with hydrogen ion index as a variable from the relationship between the three or more hydrogen ion index and photosynthesis rate measurements. If each measurement can be expressed by a Gaussian function, the update unit 30 expresses the hydrogen ion index-dependent function using the Gaussian function of equation (4).

[0064]

[0065] The update unit 30 may express the relationship between the hydrogen ion concentration and the amount of photosynthesis using two Gaussian functions. For example, the update unit 30 calculates the hydrogen ion concentration at the branching point using the least squares method with the values ​​of multiple measurement points. The update unit 30 calculates a first Gaussian function that shows a range smaller than the hydrogen ion concentration at the branching point, and a second Gaussian function that shows a range larger than the hydrogen ion concentration at the branching point, as functions that show the relationship between the hydrogen ion concentration and the amount of photosynthesis.

[0066] The update unit 30 may be expressed using a function other than a Gaussian function, based on the relationship between the hydrogen ion index of three or more points and the measured amount of photosynthesis.

[0067] (Processing Method) The design process relating to this disclosure will be explained with reference to Figure 7. The design process is performed by each part of the design unit 20.

[0068] In step S101, the extraction unit 21 extracts variables from the differential equation of the state variable. In step S102, the determination unit 22 determines each term of the differential equation as a measurement item.

[0069] Steps S103 to S106 are executed for each function of the variable extracted in step S101.

[0070] In step S103, the determination unit 22 determines whether the function of the target variable converges or not. If it converges, the process proceeds to step S104; otherwise, the process proceeds to step S105.

[0071] In step S104, the determination unit 22 determines the value of the variable that maximizes the value of the function of the target variable as the optimal condition for the target variable. In step S105, the determination unit 22 does not set an optimal condition for the target variable.

[0072] After executing steps S103 to S106 for the function of the target variable, steps S103 to S106 are executed for the function of the next variable. After executing steps S103 to S106 for the function of each variable, the process is terminated.

[0073] Referring to Figure 8, the update process related to this disclosure will be explained. The update process is performed by the update unit 30.

[0074] In step S201, the update unit 30 identifies the items among the measurement items of the measurement data 16 that show a significant difference in relation to the change in the variable.

[0075] In step S202, the update unit 30 identifies the function that the term identified in step S201 depends on a variable as an additional function. The update unit 30 adds the additional function to the differential equation and then terminates the process.

[0076] The processing unit 1 of this disclosure outputs the minimum necessary data for model generation and its acquisition method (experimental design) from the differential equations of state variables, even when there are points that are unclear from the information of existing models when generating an ecosystem model for a certain sea area. The processing unit 1 enables the unanswered questions to be expressed as functions from experimental measurement data based on the output data. As a result, the processing unit 1 can easily generate an ecosystem model for a specific sea area or phenomenon from an existing ecosystem model for a target area or phenomenon.

[0077] Thus, the processing apparatus 1 of this disclosure can reduce the burden of experiments required to add new elements to existing marine ecosystem models.

[0078] The processing unit 1 described above in this disclosure is, for example, a general-purpose computer system comprising a CPU (Central Processing Unit, processor) 901, memory 902, storage 903 (HDD: Hard Disk Drive, SSD: Solid State Drive), communication device 904, input device 905, and output device 906. In this computer system, each function of the processing unit 1 is realized by the CPU 901 executing a program loaded onto the memory 902.

[0079] The processing unit 1 may be implemented on a single computer, or it may be implemented on multiple computers. Furthermore, the processing unit 1 may be a virtual machine implemented on a computer.

[0080] The program of the processing unit 1 can be stored on a computer-readable recording medium such as an HDD, SSD, USB (Universal Serial Bus) memory, CD (Compact Disc), or DVD (Digital Versatile Disc), or it can be distributed over a network. A computer-readable recording medium is, for example, a non-transitory recording medium.

[0081] The functionality of each element of this disclosure may be implemented using circuits or processing circuits that include general-purpose processors, special-purpose processors, integrated circuits, ASICs (Application Specific Integrated Circuits), FPGAs (Field Programmable Gate Arrays), general circuits, and / or combinations thereof that are programmed using one or more programs stored in one or more memories, or otherwise configured to perform the disclosed functionality. A processor is considered a circuit or processing circuit because it includes transistors and other circuits. A processor may be a programmed processor that executes programs stored in memory. In this disclosure, a circuit, unit, or means is hardware that performs the enumerated functionality, or hardware programmed to perform the enumerated functionality. Hardware may be any hardware of this disclosure that is programmed or configured to perform the enumerated functionality.

[0082] Memory stores computer programs, including computer instructions. These computer instructions provide logic and routines that enable hardware (e.g., circuits or processing circuits) to perform the methods of this disclosure. This computer program can be implemented in common forms, such as computer-readable recording media, computer program products, memory devices, recording media such as CD-ROMs and DVDs, and / or memory of FPGAs or ASICs.

[0083] This disclosure is not limited to the embodiments described above, and numerous modifications are possible within the scope of its essence.

[0084] 1 Processing unit 2 Estimation system 11 Variable data 12 Measurement item data 13 Optimal condition data 16 Measurement data 17 Additional function data 20 Design unit 21 Extraction unit 22 Decision unit 30 Update unit 100 Estimation device 104 Ecosystem database 111 Model template data 112 Condition data 113 Result data 121 Acquisition unit 122 Model generation unit 124 Condition acquisition unit 125 Prediction unit 126 Display unit 901 CPU 902 Memory 903 Storage 904 Communication device 905 Input device 906 Output device F Differential equation data

Claims

1. A processing device comprising: an extraction unit for extracting variables from differential equations of state variables constituting a marine ecosystem model; and a determination unit for determining each term included in the differential equations as measurement items that may depend on newly considered variables.

2. The processing apparatus according to claim 1, wherein the determination unit determines the value of the variable that maximizes the value of the function, treating the term in the differential equation that includes the variable as a function of the variable, and if the value of the function does not converge, it does not determine the optimal condition for the variable of the function.

3. The processing apparatus according to claim 1, further comprising an update unit that identifies a term among the measurement items that shows a significant difference in response to the change in the newly considered variable, and adds a function to the differential equation in which the identified term depends on the variable.

4. A method for a computer to extract variables from differential equations of state variables that constitute a marine ecosystem model, and to determine each term included in the differential equations as a measurement item that may depend on newly considered variables.

Images