Parameter generation device, method, and program

Abstract

In the present invention, an input means 81 receives the input of a condition to be satisfied by a parameter. A model generation means 82 converts the input condition into a model expressed as a Hamiltonian. An annealing processing means 83 generates an Ising model from the model obtained by the conversion, inputs the generated Ising model to an annealing machine, and performs annealing. An output means 84 converts an annealing result to a parameter and outputs the same.

Description

[Technical Field]

[0001] The present invention relates to a parameter generation device, a parameter generation method, and a parameter generation program that generate parameters using an annealing machine, as well as a simulation system and a simulation method that perform trials based on the generated parameters. [Background technology]

[0002] In research into new materials, in order to discover a manufacturing method for a new material that exhibits the desired performance, a huge number of combinations of parameters (parameter sets) are attempted, which indicate the type and amount of material, processing temperature, pressure, time, etc. However, the number of combinations is astronomical, making it impossible to try every possible pattern.

[0003] In recent years, annealing machines, which rapidly try out such combinations, have come into use. Annealing machines are specialized devices for finding the ground state of the Hamiltonian of the Ising model.

[0004] For example, Patent Document 1 describes an integer programming device that uses an annealing machine to find a solution to a constrained integer programming problem. The device described in Patent Document 1 sets a Hamiltonian in the annealing machine, the Hamiltonian including a first term representing the objective function of the constrained integer programming problem, a second term representing a constraint condition and expressed as a linear equation related to multiple spins, and a third term represented as a quadratic equation, and obtains the found values ​​of the multiple spins. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent Publication No. 2021-89596 Summary of the Invention [Problem to be solved by the invention]

[0006] Meanwhile, knowledge has been accumulated from past experience and simulation results, such as which combinations of conditions are expected to produce good results and which combinations of conditions are inappropriate. In research, new parameter sets are determined from within the range that satisfies the conditions of this accumulated knowledge, and the process of obtaining results is repeated by making prototypes and conducting simulations.

[0007] However, when the combination of conditions becomes complex, such as when conditions are interrelated, it can be difficult to find a parameter set that satisfies all of the conditions. Generally, parameter sets that satisfy the conditions are created based on the experience and intuition of experts, but this method is highly dependent on the individual, and there is a problem in that parameters cannot be searched for efficiently without a specific expert.

[0008] Another method is to randomly set parameters and extract a parameter set that satisfies the conditions. However, this method also has the problem of being inefficient because the probability of generating a parameter set that satisfies the conditions becomes low when the conditions become complex.

[0009] Furthermore, the device described in Patent Document 1 is intended to solve a combinatorial optimization problem such as the knapsack problem as an integer programming problem, and is not intended to derive a parameter set that satisfies each condition.

[0010] Therefore, an object of the present invention is to provide a parameter generation device, a parameter generation method, and a parameter generation program that can efficiently generate a parameter set that satisfies desired conditions, as well as a simulation system and a simulation method that perform trials based on the generated parameters. [Means for solving the problem]

[0011] The parameter generation device according to the present invention includes an input means for receiving input of conditions that the parameters must satisfy, a model generation means for converting the input conditions into a model expressed by a Hamiltonian, an annealing processing means for generating an Ising model from the converted model and inputting the generated Ising model into an annealing machine to perform annealing, and an output means for converting the results of the annealing into parameters and outputting them. The input means receives an input of a reference parameter set, and the model generation means generates a Hamiltonian modeled so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes. It is characterized by:

[0012] A simulation system according to the present invention includes a simulator that performs trials based on input parameters, a parameter generation device that generates parameters to be input to the simulator, and an annealing machine that performs annealing based on an Ising model, wherein the parameter generation device includes input means that accepts input of conditions that the parameters should satisfy, model generation means that converts the input conditions into a model expressed by a Hamiltonian, annealing processing means that generates an Ising model from the converted model and outputs the generated Ising model to the annealing machine, and output means that converts the results of annealing performed by the annealing machine into parameters and outputs the parameters to the simulator; The input means receives an input of a reference parameter set, and the model generation means generates a Hamiltonian modeled so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes; The annealing machine performs annealing based on the Ising model input from the parameter generation device, and outputs the results of the annealing to the parameter generation device, and the simulator performs trials on the input parameter set and outputs the results of the trials.

[0013] The parameter generation method according to the present invention comprises: The computer Accepts input of the conditions that the parameters must satisfy and the reference parameter set, The computer The input conditions are converted into a model expressed in Hamiltonian so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes. The computer Generate an Ising model from the converted model, The computer The generated Ising model is input into an annealing machine to perform annealing. The computerIt is characterized by converting the results of annealing into parameters and outputting them.

[0014] In the simulation method according to the present invention, a parameter generating device generates parameters to be input to a simulator that performs trials based on input parameters, and the parameter generating device inputs conditions that the parameters should satisfy. , and input of the reference parameter set and the parameter generating device receives the The closer the generated parameters are to the reference parameter set, the smaller the energy will be. The input conditions are converted into a model expressed in Hamiltonian, the parameter generation device generates an Ising model from the converted model, the parameter generation device outputs the generated Ising model to an annealing machine, the annealing machine performs annealing based on the Ising model input from the parameter generation device, the annealing machine outputs the annealing execution results to the parameter generation device, the parameter generation device converts the annealing execution results by the annealing machine into parameters and outputs them to a simulator, and the simulator performs a trial for the input parameter set and outputs the trial results.

[0015] The parameter generation program according to the present invention causes a computer to perform an input process of receiving input of conditions that the parameters must satisfy, a model generation process of converting the input conditions into a model expressed by a Hamiltonian, an annealing process of generating an Ising model from the converted model and inputting the generated Ising model into an annealing machine to perform annealing, and an output process of converting the results of the annealing into parameters and outputting them. In the input process, a reference parameter set is input and in the model generation process, a Hamiltonian is generated that is modeled so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes. It is characterized by: [Effects of the Invention]

[0016] According to the present invention, a parameter set that satisfies desired conditions can be efficiently generated. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a block diagram showing a configuration example of a first embodiment of a simulation system of the present invention. [Figure 2] FIG. 10 is an explanatory diagram showing an example of a condition. [Figure 3] FIG. 10 is an explanatory diagram showing an example of a parameter set. [Figure 4] FIG. 10 is an explanatory diagram showing an example of a trial result. [Figure 5] 10 is a flowchart illustrating an example of the operation of the parameter generating device. [Figure 6] 10 is a flowchart illustrating an example of the operation of the simulation system. [Figure 7] FIG. 10 is a block diagram showing a configuration example of a second embodiment of a simulation system according to the present invention. [Figure 8] 1 is a block diagram showing an overview of a parameter generating device according to the present invention; [Figure 9] 1 is a block diagram showing an overview of a simulation system according to the present invention; DETAILED DESCRIPTION OF THE INVENTION

[0018] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0019] Embodiment 1. 1 is a block diagram showing an example of the configuration of a first embodiment of a simulation system of the present invention. The simulation system 100 of this embodiment includes a parameter generating device 10, an annealing machine 20, and a simulator 30.

[0020] The parameter generation device 10 of this embodiment is a device that generates parameters to be input to the simulator 30, and the simulator 30 performs trials based on the generated parameters. The annealing machine 20 is a dedicated device that finds the ground state of the Hamiltonian of the Ising model, and is a device that performs annealing based on the Ising model generated by the parameter generation device 10.

[0021] More specifically, an annealing machine is a device that probabilistically determines the value of a binary variable that minimizes or maximizes the objective function (i.e., Hamiltonian) of an Ising model that takes binary variables as arguments. The binary variables may be realized by classical bits or quantum bits. The annealing machine 20 of this embodiment may take any form. The annealing machine 20 may be configured with any hardware that can probabilistically determine the value of a binary variable that minimizes or maximizes the objective function that takes binary variables as arguments. The annealing machine 20 may be, for example, a non-von Neumann computer in which the objective function is implemented by hardware in the form of an Ising model. The annealing machine 20 may be a quantum annealing machine or a general annealing machine.

[0022] Furthermore, in this embodiment, the term "parameters" refers to the values ​​of each attribute to be generated, and more specifically, refers to the values ​​of attributes to be input into the simulator 30 for trial purposes. For example, when a trial is conducted on a manufacturing method for a new material, the parameters are values ​​indicating, for example, the amount of each material, the processing temperature and pressure to be set in the manufacturing process, the processing time, etc. In addition, when estimating the GDP of each country when the global GDP (Gross Domestic Product) increases, the parameters are values ​​indicating, for example, the economic index of each country, the correlation index of each country's GDP, the amount of imports and exports, demographics, etc.

[0023] The parameter generation device 10 of this embodiment generates a parameter set that satisfies desired conditions. Note that, depending on the setting of the conditions, it may be impossible to generate a parameter set that satisfies all constraints. Therefore, in this embodiment, the "parameter set that satisfies the desired conditions" includes not only a parameter set that satisfies all conditions, but also a parameter set that satisfies the set conditions as much as possible (i.e., a parameter set in which constraint violations are within a predetermined allowable range). Note that the allowable degree or range may be predetermined by a user or the like.

[0024] In other words, in this embodiment, the "parameter set that satisfies the desired conditions" includes not only a parameter set that satisfies all conditions, i.e., a perfect score of 100, but also a parameter set that allows some constraint violations, i.e., a score of around 95. In this embodiment, an annealing machine is used to derive the parameters, so the values ​​obtained as a result of deriving the ground state of the Hamiltonian can be said to be a parameter set that satisfies the desired conditions.

[0025] The parameter generating device 10 includes an input unit 11, a model generating unit 12, an annealing processing unit 13, and an output unit 14.

[0026] The input unit 11 receives input of conditions that a parameter should satisfy (hereinafter, may be simply referred to as parameter conditions). Specifically, the input unit 11 receives input of conditions that a parameter itself should satisfy and conditions that a parameter should satisfy in relation to other parameters.

[0027] The content of the conditions and the manner in which they are expressed are arbitrary. For example, when trying out a manufacturing method for a new material, parameter conditions include conditions for processing the material itself (e.g., upper limit temperature, pressure, processing time, etc.), conditions for exclusive materials, and conditions regarding the distribution of material quantities (e.g., the sum of the material quantities, or the specified quantity of each individual material). Other parameter conditions include conditions regarding the selection of material types (e.g., specifying one material from each material group, etc.).

[0028] FIG. 2 is an explanatory diagram showing examples of input conditions. In the example shown in FIG. 2, materials a through z exist, and conditions include the amount of each material, as well as the temperature and pressure for each material. Specifically, the example shown in FIG. 2 shows conditions that the total amount of materials a, c, q, and z is 0.8 or less, and that the total amount of materials c, h, and i is approximately 2.0. In addition, the example shown in FIG. 2 shows conditions that materials a and b, i and j, and p and q are not used simultaneously (are exclusive), that the upper limit temperature for material a is 60, the upper limit temperature for material p is 150, and the upper limit pressure for material c is 1.2.

[0029] In the example shown in FIG. 2, the parameter conditions are given in mathematical expressions, but the parameter conditions may be given in natural language expressions.

[0030] The input unit 11 may also accept an input of the degree of importance to be attached to each condition (hereinafter referred to as the condition weight). The condition weight may be an absolute value or a relative value. Furthermore, the input unit 11 may also accept an input specifying the number of times that the annealing processing unit 13, which will be described later, generates parameters (hereinafter referred to as the number of times of parameter generation).

[0031] The model generation unit 12 converts the input conditions into a model expressed by a Hamiltonian. Specifically, the model generation unit 12 converts each input condition into a model expressed by a Hamiltonian, and generates a final model from the linear sum of each converted model.

[0032] The method of converting the conditions into a model expressed by a Hamiltonian is arbitrary. For example, a conversion template corresponding to the content of the condition may be determined in advance for each condition, and the model generation unit 12 may determine the content of the input condition, select a conversion template, and then apply each element included in the condition to the conversion template to generate a model.

[0033] The method of converting the conditions into a model is not limited to the method using a conversion template. For example, the model generation unit 12 may convert the conditions into a model using a dedicated library that receives input conditions and generates a model represented by a Hamiltonian.

[0034] A specific example in which the model generation unit 12 converts conditions into a model will be described below. In this example, the parameter conditions for a trial manufacturing method for a new material will be used as an example. Here, the parameters assumed are the type (amount) of material and the type of environment (temperature and pressure).

[0035] First, the decision variable (quantum bit) xq representing the type of material is defined as follows: xq[type,quantity] The type indicates the type of material, and the quantity indicates the amount in, for example, 0.1 x 10 bits.

[0036] Similarly, the decision variable (qubit) xe representing the type of environment is defined as follows: xe[type,volume] The type indicates the type of environment, and the volume indicates a value in, for example, 10x10 bits. In addition, auxiliary variables s1[] and s2[] are used as appropriate.

[0037] The model generating unit 12 models each condition as a Hamiltonian. The following description will be given for the constraints shown in FIG. (1) Regarding a+c+q+z≦0.8

number

[0038] (2) Regarding c+h+i≒2.0

number

[0039] (3) Regarding a and b being exclusive

number

[0040] (4) If a is included, the upper limit temperature is 60.

number

[0041] For example, when these Hamiltonians H1 to H4 are generated, the model generation unit 12 generates a linear sum of each model as shown in the following equation 5 to obtain a final model. Note that w1 to w4 are weighting constants indicating the weights of the conditions. The weighting constants shown in equation 5 are arbitrary and may be set randomly, or may be the weights of the conditions input by the input unit 11.

[0042] H=w1*H1+w2*H2+w3*H3+w4*H4 (Formula 5)

[0043] The annealing processing unit 13 generates an Ising model from the converted model. Specifically, the annealing processing unit 13 generates an Ising model from a linear sum of the converted models (for example, the Hamiltonian shown in Equation 5). Here, both the converted model and the linear sum of the converted models are Hamiltonians. Note that, since the method of generating an Ising model from a Hamiltonian is widely known, a detailed description thereof will be omitted here.

[0044] The annealing processing unit 13 then inputs the generated Ising model into the annealing machine 20, causes the annealing to be performed, and obtains an execution result. Note that the annealing processing unit 13 may cause the annealing machine 20 to perform annealing only once, or may cause the annealing to be performed a number of times specified as the number of parameter generation times. In other words, the annealing processing unit 13 may cause the annealing machine 20 to perform annealing on the same Ising model multiple times.

[0045] In the case of general mathematical optimization, an optimal solution is obtained almost uniquely, whereas the results of execution by an annealing machine usually produce so many variations that it is difficult to find the same solution. This is suitable for situations where you need to obtain various variations while satisfying certain conditions, such as when trying out manufacturing methods for new materials.

[0046] The output unit 14 converts the annealing results into a parameter set and outputs it. This is because the annealing results and the parameters are expressed differently. The output unit 14 may convert the annealing results into the original parameter set (in other words, values ​​that can be executed by the simulator 30) based on, for example, defined decision variables (qubits).

[0047] The output unit 14 may output the converted parameter set directly to the simulator 30, or may output it in a file format (for example, CSV (Comma Separated Value) format). Fig. 3 is an explanatory diagram showing an example of a parameter set output by the output unit 14. The example shown in Fig. 3 shows the result of the output unit 14 outputting a parameter set in CSV format when the parameter set includes materials a to z, temperature, and pressure.

[0048] The simulator 30 executes a trial for the input parameter set and outputs the trial result (simulation result). For example, if there are multiple trial results (i.e., if multiple parameter sets are input), the simulator 30 may display the multiple output trial results in a list in association with the parameter sets.

[0049] 4 is an explanatory diagram showing an example of a trial result. The trial result shown in FIG. 4 is a table in which each parameter set is associated with a circle in the column corresponding to the condition that the parameter set satisfies. The simulator 30 may display the parameter sets in descending order of the number of conditions that the parameter sets satisfy, or may display the parameter sets in ascending order of the energy indicated by the Hamiltonian.

[0050] The input unit 11, model generation unit 12, annealing processing unit 13, and output unit 14 are realized by a processor (for example, a CPU (Central Processing Unit)) of a computer that operates according to a program (parameter generation program).

[0051] For example, the program may be stored in a storage unit (not shown) of the parameter generating device 10, and the processor may read the program and operate as the input unit 11, the model generating unit 12, the annealing processing unit 13, and the output unit 14 in accordance with the program. Also, the functions of the parameter generating device 10 may be provided in the form of SaaS (Software as a Service).

[0052] Furthermore, the input unit 11, model generation unit 12, annealing processing unit 13, and output unit 14 may each be realized by dedicated hardware. Furthermore, some or all of the components of each device may be realized by general-purpose or dedicated circuits, processors, etc., or a combination thereof. These may be configured by a single chip, or by multiple chips connected via a bus. Some or all of the components of each device may be realized by a combination of the above-mentioned circuits, etc., and a program.

[0053] Furthermore, when some or all of the components of the parameter generation device 10 are realized by a plurality of information processing devices, circuits, etc., the plurality of information processing devices, circuits, etc. may be centrally or decentralized. For example, the information processing devices, circuits, etc. may be realized as a client-server system, a cloud computing system, or the like, in a form in which each is connected via a communication network.

[0054] Next, the operation of this embodiment will be described. First, the operation of the parameter generating device 10 of this embodiment will be described. FIG. 5 is a flowchart showing an example of the operation of the parameter generating device 10.

[0055] The input unit 11 receives input of parameter conditions (step S11). The model generation unit 12 converts the input conditions into a model expressed by a Hamiltonian (step S12). The annealing processing unit 13 generates an Ising model from the converted model (step S13), and inputs the generated Ising model into an annealing machine to execute annealing (step S14). Then, the output unit 14 converts the annealing execution result into a parameter set and outputs it (step S15).

[0056] Next, a description will be given of the operation of the simulation system 100 of this embodiment. Fig. 6 is a flowchart showing an example of the operation of the simulation system 100. In this example of operation, it is assumed that the output result from the parameter generation device 10 is input to the simulator 30.

[0057] The process from converting the input parameter conditions into a model to generating an Ising model is the same as the process from step S11 to step S13 illustrated in Fig. 5. The annealing processing unit 13 outputs the generated Ising model to the annealing machine (step S21).

[0058] The annealing machine 20 performs annealing based on the input Ising model (step S22). The annealing machine 20, for example, finds multiple spin values ​​through annealing. Then, the annealing machine 20 outputs the annealing results to the parameter generation device 10 (step S23). The output unit 14 converts the annealing results into a parameter set and outputs it to the simulator 30 (step S24). The simulator 30 performs a trial for the input parameter set and outputs the trial results (step S25).

[0059] As described above, in this embodiment, the input unit 11 accepts input of parameter conditions, and the model generation unit 12 converts the input conditions into a model expressed by a Hamiltonian. The annealing processing unit 13 generates an Ising model from the converted model and inputs the generated Ising model into an annealing machine to perform annealing. The output unit 14 then converts the results of the annealing into parameters and outputs them. This makes it possible to efficiently generate a parameter set that satisfies desired conditions.

[0060] Next, a modified example of the simulation system of this embodiment will be described. In the above embodiment, the case where the input unit 11 accepts input of so-called constraint conditions as parameter conditions has been described. On the other hand, in this modified example, the case where the input unit 11 accepts input of conditions as parameter conditions that, when satisfied, result in better results, i.e., objective function-like conditions, will be described.

[0061] Such objective function conditions can be obtained, for example, by using a large amount of experimental data and having AI (Artificial Intelligence) perform machine learning. An example of the objective function conditions is a linear regression model that predicts a performance index. In this modification, the simulation system 100 introduces this linear regression model into the Hamiltonian and uses it as a condition for achieving desired performance.

[0062] The linear regression model is expressed as a linear regression equation with the performance index (hereinafter sometimes simply referred to as index) as the objective variable and the parameters described above as explanatory variables. For example, if the predicted value of the i-th index is y i Then, y i is expressed as a linear sum of each parameter. Performance indicators include hardness, flexibility, and heat resistance.

[0063] The input unit 11 receives the above-mentioned objective function condition as an input of the condition that the parameters should satisfy, that is, the linear regression model y i Accepts input.

[0064] The model generation unit 12 converts the input objective function-like conditions into a model expressed in Hamiltonian. For example, when the input unit 11 receives an input of a linear regression model showing N performance indexes, the model generation unit 12 converts the received linear regression model into a Hamiltonian H as shown in the following formula 6: O It may be modeled as

[0065]

number

[0066] In Equation 6, Lmed i is the target median of the i-th index. For example, if the hardness index is considered desirable in the range of 20 to 25, the target median is 23. Also, W iare weights that are set in advance according to the importance of each index. For example, if heat resistance is more important than hardness, the weight of heat resistance is set greater than the weight of hardness.

[0067] Then, the model generation unit 12 calculates the above H O The linear sum of each model, including the addition of the , is generated to obtain the final model. The subsequent processing is the same as in the above embodiment.

[0068] As described above, in this modification, the input unit 11 receives an input of a linear regression model expressed by a linear regression equation in which the performance index is the objective variable and the parameters are the explanatory variables, and the model generation unit 12 converts the input linear regression model into a model expressed in Hamiltonian. Therefore, in addition to the effects of the above embodiment, it is possible to efficiently generate a parameter set that satisfies more desirable conditions.

[0069] Embodiment 2. Next, a second embodiment of the simulation system of the present invention will be described. Fig. 7 is a block diagram showing an example of the configuration of the second embodiment of the simulation system of the present invention. A simulation system 200 of this embodiment includes a parameter generation device 110, an annealing machine 20, and a simulator 30. The aspects of the annealing machine 20 and the simulator 30 are the same as those of the first embodiment.

[0070] The parameter generation device 110 includes an input unit 111, a model generation unit 112, an annealing processing unit 13, and an output unit 14. That is, the parameter generation device 110 of this embodiment differs from the parameter generation device 10 of the first embodiment in that it includes the input unit 111 and the model generation unit 112 instead of the input unit 11 and the model generation unit 12. Other configurations are the same as those of the first embodiment.

[0071] The input unit 111 accepts input of parameter conditions, similar to the input unit 11 of the first embodiment. Furthermore, the input unit 111 of this embodiment accepts input of a reference parameter set in addition to the parameter conditions. An example of a reference parameter set is a parameter set that has produced good results in the past. By accepting input of a parameter set that has produced good results in the past, it becomes possible to generate a parameter set that is close to this parameter set.

[0072] The reference parameter set is not limited to a parameter set that has produced good results in the past. The reference parameter set may be, for example, a parameter set currently in operation. By accepting input of a parameter set currently in operation, it is possible to generate a parameter set that does not significantly change the current operation.

[0073] The model generation unit 112 converts the input conditions into a model expressed by a Hamiltonian, similar to the model generation unit 12 of the first embodiment. Furthermore, the model generation unit 112 generates a Hamiltonian that is modeled so that the closer the generated parameters are to a reference parameter set, the smaller the energy becomes. Then, the model generation unit 112 generates a final model from the linear sum of each model.

[0074] For example, if the Hamiltonian generated in this way is H5 and the weight of H5 is w5, the linear sum of each model is generated as shown in Equation 7 below.

[0075] H=w1*H1+w2*H2+w3*H3+w4*H4+w5*H5 (Formula 7)

[0076] The input unit 111, the model generation unit 112, the annealing processing unit 13, and the output unit 14 are realized by a processor of a computer that operates according to a program (parameter generation program).

[0077] As described above, in this embodiment, the input unit 111 receives input of a reference parameter set, and the model generation unit 112 generates a Hamiltonian modeled so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes. Thus, in addition to the effects of the first embodiment, a parameter set can be generated that is closer to the reference parameter set.

[0078] Next, an overview of the present invention will be described. Fig. 8 is a block diagram showing an overview of a parameter generation device according to the present invention. A parameter generation device 80 according to the present invention includes input means 81 (e.g., input unit 11) that accepts input of conditions to be satisfied by parameters, model generation means 82 (e.g., model generation unit 12) that converts the input conditions into a model expressed by a Hamiltonian, annealing processing means 83 (e.g., annealing processing unit 13) that generates an Ising model from the converted model and inputs the generated Ising model into an annealing machine to perform annealing, and output means 84 (e.g., output unit 14) that converts the results of the annealing into parameters and outputs them.

[0079] Such a configuration allows efficient generation of parameter sets that satisfy desired conditions.

[0080] Specifically, the model generation means 82 may convert each input condition into a model expressed by a Hamiltonian, and the annealing processing means 83 may generate an Ising model from the linear sum of each converted model.

[0081] Alternatively, the input means 81 may receive input of a reference parameter set, and the model generation means 82 may generate a Hamiltonian modeled so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes. Such a configuration makes it possible to generate a parameter set close to the reference parameter set.

[0082] Furthermore, the annealing processing means 83 may cause the annealing machine to perform annealing on the same Ising model multiple times. Depending on the characteristics of the annealing machine, it becomes possible to obtain a wide variety of parameter sets.

[0083] The input means 81 may, for example, accept input of at least one of the following conditions to be satisfied by the parameters: a condition for processing the material itself, a condition for exclusive material, a condition regarding the distribution of the amount of material, and a condition regarding the selection of the type of material.

[0084] 9 is a block diagram showing an overview of a parameter generation device according to the present invention. The simulation system 1 according to the present invention includes a simulator 70 (e.g., simulator 30) that performs trials based on input parameters, a parameter generation device 180 (e.g., parameter generation device 10) that generates parameters to be input to the simulator 30, and an annealing machine 90 (annealing machine 20) that performs annealing based on an Ising model.

[0085] The parameter generation device 180 includes input means 181 (e.g., input unit 11) that accepts input of conditions that the parameters must satisfy, model generation means 182 (e.g., model generation unit 12) that converts the input conditions into a model expressed in Hamiltonian, annealing processing means 183 (e.g., annealing processing unit 13) that generates an Ising model from the converted model and outputs the generated Ising model to the annealing machine 90, and output means 184 (e.g., output unit 14) that converts the results of annealing performed by the annealing machine 90 into parameters and outputs them to the simulator 70.

[0086] The annealing machine 90 performs annealing based on the Ising model input from the parameter generation device 180, and outputs the results of the annealing to the parameter generation device 180. The simulator 70 then performs a trial on the input parameter set, and outputs the results of the trial.

[0087] With this configuration, parameter sets that satisfy desired conditions can be generated efficiently, making it possible to carry out trials efficiently.

[0088] A part or all of the above-described embodiments can be described as, but not limited to, the following supplementary notes.

[0089] (Appendix 1) An input means for receiving input of conditions that the parameters should satisfy; a model generation means for converting input conditions into a model expressed by a Hamiltonian; an annealing processing means for generating an Ising model from the converted model and inputting the generated Ising model into an annealing machine to perform annealing; and an output means for converting the results of the annealing into the parameters and outputting them. A parameter generation device characterized by:

[0090] (Appendix 2) The model generation means converts each input condition into a model expressed by a Hamiltonian, The annealing processing means generates an Ising model from the linear sum of each transformed model. 2. The parameter generating device of claim 1.

[0091] (Supplementary Note 3) The input means receives an input of a reference parameter set; The model generating means generates a Hamiltonian modeled so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes. 3. The parameter generating device according to claim 1 or 2.

[0092] (Appendix 4) The annealing processing means causes the annealing machine to perform annealing on the same Ising model multiple times. 4. A parameter generating device according to any one of claims 1 to 3.

[0093] (Appendix 5) The input means receives input of at least one of the conditions to be satisfied by the parameters, including a condition for processing the material itself, a condition for exclusive material, a condition for the distribution of the amount of material, and a condition for the selection of the type of material. 5. A parameter generating device according to any one of claims 1 to 4.

[0094] (Appendix 6) The input means receives input of a linear regression model expressed by a linear regression formula in which the performance index is a response variable and the parameters are explanatory variables; The model generation means converts the input linear regression model into a model expressed in Hamiltonian terms. 6. A parameter generating device according to any one of Supplementary Note 1 to Supplementary Note 5.

[0095] (Appendix 7) A simulator that performs trials based on input parameters; a parameter generation device for generating parameters to be input to the simulator; an annealing machine that performs annealing based on an Ising model, The parameter generating device an input means for receiving an input of a condition that the parameter should satisfy; a model generation means for converting input conditions into a model expressed by a Hamiltonian; an annealing processing means for generating an Ising model from the converted model and outputting the generated Ising model to the annealing machine; an output means for converting a result of annealing performed by the annealing machine into the parameters and outputting the parameters to the simulator; The annealing machine comprises: performing annealing based on the Ising model input from the parameter generation device, and outputting the annealing execution result to the parameter generation device; The simulator executes a trial for the input parameter set, and Output the results A simulation system comprising:

[0096] (Appendix 8) The simulator displays multiple output trial results in a list, corresponding to the parameter sets. 8. The simulation system according to claim 7.

[0097] (Appendix 9) Accepts input of conditions that the parameters must satisfy, The input conditions are converted into a model expressed in Hamiltonian. Generate an Ising model from the converted model, The generated Ising model is input into an annealing machine to perform annealing. The results of the annealing are converted into the parameters and output. A parameter generation method comprising:

[0098] (Appendix 10) A parameter generation device that generates parameters to be input to a simulator that performs trials based on input parameters receives input of conditions that the parameters must satisfy; the parameter generation device converts the input conditions into a model expressed in Hamiltonian; the parameter generation device generates an Ising model from the converted model; the parameter generation device outputs the generated Ising model to an annealing machine; the annealing machine performs annealing based on the Ising model input from the parameter generation device; the annealing machine outputs the annealing execution result to the parameter generation device; the parameter generation device converts the results of annealing performed by the annealing machine into the parameters and outputs them to the simulator; The simulator executes a trial for the input parameter set and outputs the result of the trial. A simulation method comprising:

[0099] (Appendix 11) To the computer, an input process that accepts input of conditions that the parameters must satisfy; A model generation process that converts the input conditions into a model expressed in Hamiltonian. An annealing process in which an Ising model is generated from the converted model, and the generated Ising model is input into an annealing machine to perform annealing; and An output process of converting the results of the annealing into the parameters and outputting them. A parameter generation program for executing the above.

[0100] Although the present invention has been described above with reference to the embodiments and examples, the present invention is not limited to the above-described embodiments and examples. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

[0101] This application claims priority based on Japanese Patent Application No. 2021-190203, filed on November 24, 2021, the entire disclosure of which is incorporated herein by reference. [Explanation of symbols]

[0102] 10,110 Parameter Generator 11,111 Input section 12,112 Model generation section 13 Annealing processing section 14 Output section 20 Annealing Machine 30 Simulator 100,200 Simulation System

Claims

1. an input means for receiving an input of a condition that the parameter should satisfy; a model generation means for converting input conditions into a model expressed in Hamiltonian; an annealing processing means for generating an Ising model from the converted model and inputting the generated Ising model into an annealing machine to perform annealing; an output means for converting the results of the annealing into the parameters and outputting the parameters; the input means accepts input of a reference parameter set; The model generation means generates a Hamiltonian modeled so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes. A parameter generation device characterized by:

2. The model generation means converts each input condition into a model expressed by a Hamiltonian, The annealing processing means generates an Ising model from the linear sum of each transformed model. The parameter generating device according to claim 1 .

3. The annealing processing means causes the annealing machine to perform annealing on the same Ising model multiple times.

3. The parameter generating device according to claim 1.

4. The input means receives input of at least one of a condition for processing the material itself, a condition for exclusive material, a condition regarding the distribution of the amount of material, and a condition regarding the selection of the type of material as a condition to be satisfied by the parameter.

3. The parameter generating device according to claim 1.

5. a simulator that performs trials based on input parameters; a parameter generation device for generating parameters to be input to the simulator; an annealing machine that performs annealing based on an Ising model, The parameter generating device an input means for receiving an input of a condition that the parameter should satisfy; a model generation means for converting input conditions into a model expressed in Hamiltonian; an annealing processing means for generating an Ising model from the converted model and outputting the generated Ising model to the annealing machine; an output means for converting a result of annealing performed by the annealing machine into the parameters and outputting the parameters to the simulator; the input means accepts input of a reference parameter set; the model generation means generates a Hamiltonian modeled so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes; The annealing machine comprises: performing annealing based on the Ising model input from the parameter generation device, and outputting the annealing execution result to the parameter generation device; The simulator executes a trial for the input parameter set, and Output the results A simulation system comprising:

6. The simulator displays multiple output trial results in a list, corresponding to the parameter sets. The simulation system according to claim 5.

7. A computer receives input of conditions that the parameters must satisfy and input of a reference parameter set, The computer converts the input conditions into a model expressed by a Hamiltonian so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes; the computer generates an Ising model from the converted model; the computer inputs the generated Ising model into an annealing machine to perform annealing; The computer converts the results of the annealing into the parameters and outputs them. A parameter generation method comprising:

8. A parameter generation device that generates parameters to be input to a simulator that performs trials based on input parameters receives input of conditions that the parameters must satisfy and input of a reference parameter set; the parameter generation device converts the input conditions into a model expressed by a Hamiltonian so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes; the parameter generation device generates an Ising model from the converted model; the parameter generation device outputs the generated Ising model to an annealing machine; the annealing machine performs annealing based on the Ising model input from the parameter generation device; the annealing machine outputs the annealing execution result to the parameter generation device; the parameter generation device converts the results of annealing performed by the annealing machine into the parameters and outputs them to the simulator; The simulator executes a trial for the input parameter set and outputs the result of the trial. A simulation method comprising:

9. On the computer, an input process that accepts input of conditions that the parameters must satisfy; A model generation process that converts the input conditions into a model expressed in Hamiltonian. An annealing process in which an Ising model is generated from the converted model, and the generated Ising model is input into an annealing machine to perform annealing; and Execute an output process of converting the results of the annealing into the parameters and outputting them; In the input process, an input of a reference parameter set is accepted; In the model generation process, a Hamiltonian is generated that is modeled so that the closer the generated parameters are to the reference parameter set, the smaller the energy becomes. Parameter generation program for

Images