JP3245640B2 - Reactor core analysis method and medium recording the reactor core analysis program - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a reactor core analysis method for performing a reactor core analysis by a computer and a medium recording a reactor core analysis program, and more particularly to a reactor core utilizing object-oriented technology. The present invention relates to a medium recording an analysis method and a reactor core analysis program.

[0002]

2. Description of the Related Art Reactor core analysis methods for analyzing effective neutron multiplication factor, power distribution, etc. in a reactor core include a diffusion calculation method based on a neutron diffusion theory, a transport calculation method based on a neutron transport theory, and the like. There is. Both of these calculation methods are analytical calculation methods in nuclear reactor physics, and differ mainly in how neutrons are treated as physical quantities. The diffusion calculation uses an approximation method that assumes that all neutrons in the core are represented by a single direction of motion. On the other hand, the transport calculation method takes into account the multidirectionality of the neutron motion, and therefore, a more accurate calculation result can be obtained as compared with the diffusion calculation method.

In the diffusion calculation method, a core analysis is performed by a numerical solution of a neutron diffusion equation using a computer. At this time, an arbitrary discretized mesh point is set for each parameter, and It solves by deriving a difference equation based on points. In the transport calculation method, a core analysis is performed by a numerical solution of the neutron transport equation using a computer. At this time, for each parameter,
It is common to the diffusion calculation method in that a mesh point is set and a difference equation based on the mesh point is derived and solved. However, in the transport calculation method, in order to consider the multidirectionality of the neutron motion, the number of parameters is larger than in the diffusion calculation method, and the mesh points need to be set in detail. For this reason,
In the transport calculation method, the storage capacity and calculation time used by a computer increase.

In a conventional reactor core analysis method, an analysis calculation is performed by a calculation code (program) based on one of a diffusion calculation method and a transport calculation method in a calculation system specified by an input. Those that output a result are generally known. Specifically, the conventional reactor core analysis method includes a calculation part of one of a three-dimensional coarse mesh calculation code based on the diffusion calculation method and a detailed mesh calculation code based on the transport calculation method, and input, output, The present invention uses a procedural program in which calculation control portions such as determination of calculation convergence are united. However, from the viewpoint of reducing the storage capacity used by the computer and shortening the calculation time, it is common to use a reactor core analysis method using a three-dimensional coarse mesh calculation code mainly based on the diffusion calculation method. .

[0005]

In recent years, the core structure of a nuclear reactor has become diversified and complicated. For example, it is well known that a mixed oxide fuel assembly (MOX fuel assembly) of enriched uranium and plutonium dioxide and a uranium fuel assembly are simultaneously loaded into a reactor core. When analyzing a core loaded with such a mixed fuel assembly,
At the boundary between the two, the neutron flux has a large slope,
In the three-dimensional coarse mesh calculation code based on the diffusion calculation method,
There is a problem that the calculation accuracy is reduced.

In order to obtain a highly accurate calculation result, it is conceivable to perform a reactor core analysis using a detailed mesh calculation code based on a transport calculation method. In this case, however, the detailed mesh points and parameters are increased to increase the number of parameters. There is a problem that the storage capacity used by the computer increases and the calculation time also increases. Furthermore, in the conventional reactor core analysis method based on procedural programming, only one calculation method or calculation code can be selected in the entire calculation system, so that the diffusion calculation method can provide sufficient calculation accuracy. On the other hand, unnecessary analysis is performed, and there is a problem that the calculation time and the storage capacity of the computer increase.

In the conventional reactor core analysis method using procedural programming, it is conceivable to use a plurality of calculation codes. In this case, however, the ratio of the mixture of the fuel assemblies loaded in the reactor core to be analyzed is considered. Therefore, the area to be calculated by the diffusion calculation method and the area to be calculated by the transport calculation method must be determined in advance and incorporated into the procedural programming. For this reason, there is a problem that it is difficult to dynamically change the program processing according to the fuel assemblies loaded in the reactor core.

On the other hand, from the viewpoint of reduction in calculation time and storage capacity, distributed processing or parallel processing using a plurality of computers reduces the storage capacity used in each computer and shortens the calculation time. Although it is conceivable, analysis analysis by a procedural program
Since only one calculation method can be used in the entire calculation system, the calculation time cannot be effectively reduced after all.

In addition, when a calculation code based on another calculation method is added in the future, the entire analysis calculation program must be corrected, and as a result, the test work and debug work of the program accompanying the correction work also increase. And
There is a problem that it is not efficient from the viewpoint of program maintainability.

[0010] The present invention has been made in view of the above-mentioned problems, and the analysis calculation program is not based on the procedural programming technique but is based on the object-oriented programming technique, so that a high-precision calculation result can be obtained. It is a main object of the present invention to provide a reactor core analysis method and a medium recording a reactor core analysis program capable of obtaining a nuclear reactor. Another object of the present invention is to provide a nuclear reactor core analysis method and a medium recording a nuclear reactor core analysis program capable of shortening the calculation time and effectively reducing the storage capacity of a computer. It is. Another object of the present invention is to provide a reactor core analysis method and a medium in which a reactor core analysis program is recorded in which maintenance such as a program change operation is easy and maintenance is excellent in the future. Another object of the present invention is to achieve high-precision calculation results by distributed processing or parallel processing using a plurality of computers, shorten calculation time, and reduce storage capacity. An object of the present invention is to provide a medium recording a reactor core analysis method and a reactor core analysis program capable of facilitating maintenance.

[0011]

According to one aspect of the present invention, there is provided a reactor core for analyzing a reactor core by a computer based on a program based on a predetermined calculation code. In the analysis method,
Depending on the input procedure for inputting analysis information and the core structure area
A plurality of diffusion calculation object having a calculation method for performing core analysis calculated based on the diffusion calculation method Te, Ryo core structure
A plurality of transport calculation objects having a calculation method for performing a core analysis calculation based on a transport calculation method according to a region and a physical quantity object having physical quantity data as object data are generated in a storage area expressing a single calculation system. An object generation step, a calculation system construction step of generating a holding object having a calculation activation method for activating the calculation method for the calculation object having the calculation object and the physical quantity object as object data on a storage area,
The holding object selectively selects the calculation start method
By launching on the basis of the selected calculation method
Each calculation object selectively activates the calculation method, inputs and exchanges the physical quantity objects with each other to perform an analysis calculation, and outputs a calculation result by the analysis calculation procedure. And an output procedure for performing the output.

According to the present invention, a calculation code of a different type or a calculation code of a different mesh setting is obtained by converting a calculation code for performing a core analysis and physical quantity data necessary for the core analysis into objects and performing a process based on object-oriented programming. This is a reactor core analysis method that can combine the processes by

Here, an object-oriented programming technique will be briefly described. An object is a basic module that constitutes a program designed and created in object-oriented programming.
It is defined as a data structure that also has a method as a description of a procedure for executing the data. That is, the object is an integrated data and processing procedure. An object that defines and defines the data structure and processing procedure of the object in advance is called a class, and an object that is materialized based on the class is called an instance.

The properties of an object have the following characteristics. An object has an internal state and is managed by the object itself. Also, the object itself interprets the message and executes the corresponding method. There is no other way to access an object except to send a message to the object, and the object has a data hiding property that direct access to the internal state of the object is not allowed except for the methods registered in the object. are doing.

In object-oriented programming, programming is performed using the flow of such objects and messages. Specifically, an object is generated from the class, and a message for activating the method is passed to the object as shown in FIG. The object receiving the message interprets the content of the message by itself, performs processing, and reflects the processing result. Then, a processing result is processed and used as an argument of a message to another object, and a series of flows is programmed.

Therefore, in object-oriented programming, unlike procedural programming in which data and procedures are treated as separate entities, if a procedure corresponding to data is described in advance as a method and defined as a class, data processing can be performed. You don't need to write the procedure in the program, just what you want to do as a result.

Therefore, in the object-oriented programming, since the programming is performed for each object, parts are advanced, and a method is called by a message, so that the independence of each object is increased.
Maintainability is improved.

As the analysis information to be input in the input procedure of the present invention, information such as initial conditions necessary for analysis calculation, calculation conditions such as information relating to a material data map or mesh, boundary conditions, and convergence determination conditions can be used. . still,
The input of the analysis information can be read from an input device such as a keyboard device. Alternatively, the analysis information may be stored in advance as a file in a storage device such as a floppy disk device or a hard disk device, and the file may be input by reading a file in which the analysis information is stored from the storage device.

In the present invention, a calculation object based on a calculation code is generated for each calculation code by the object generation procedure. Here, "to generate a calculation object in a storage area" means that a calculation object having the same data structure as a class in which a calculation code as a predetermined analysis program is defined as a calculation method is materialized in the storage area. Generate as an instance.

The calculation codes include calculation programs based on the same or different calculation methods with different mesh settings, in addition to calculation programs based on different calculation methods. The configuration of the calculation code is not particularly limited as long as it is an analysis calculation program for performing a core analysis of a nuclear reactor. For example, a diffusion calculation program based on an existing three-dimensional mesh calculation, a transport calculation program based on a detailed mesh calculation, and the like. , A diffusion or transport node calculation program, etc., and each is generated as a separate calculation object. Further, the calculation method of the calculation object can be constituted by only a portion corresponding to a subroutine for solving an actual problem in the analysis calculation program. For example, a subroutine for calculating the neutron flux density can be used as the calculation method.

In the present invention, a physical quantity object is generated from physical quantity data by an object generating procedure. Here, “generate a physical quantity object in the storage area” means that the physical quantity object is registered in the storage area by registering the physical quantity data as object data while maintaining the same data structure as the class defined in advance. To generate as an instance. Note that "object data" refers to data such as numerical values and character strings registered as data in an object, and the term "object data" is used below to distinguish it from data in ordinary procedural programming.

As the physical quantity data, for example, neutron density, directionality, neutron energy distribution, cross-sectional area, etc. can be used, and analysis information input in the input procedure, for example, initial conditions, boundary conditions, etc., is treated as physical quantity data. Can also. Further, a physical quantity object can be generated for each physical quantity data.

The holding object generated by the calculation system construction procedure includes a plurality of calculation objects and physical quantity objects generated on the storage area used in the analysis calculation procedure, and is used as a target of the core analysis of the reactor. It corresponds to a single calculation system.

Further, the analysis calculation by the calculation object is controlled by the calculation activation method of the holding object. The configuration of the holding object is not particularly limited as long as it has a calculation object and a physical quantity object as object data and has a calculation activation method. For example, other data necessary for the analysis calculation can be registered as the object data of the held object. Further, as a method to be registered in the holding object, another method such as confirmation of a calculation result for controlling the analysis calculation can be defined.

In the analysis calculation procedure, the calculation is started by passing a message for starting a calculation start method to the held object. For example, a message can be passed to the holding object by a control program that controls the whole or a part of the procedure from the input procedure to the output procedure, but is not limited to this. The part that performs the calculation control according to the analysis calculation procedure can be converted into an object as a calculation control object. In this case, the maintenance of the calculation control itself is improved.

Next, the holding object that has received the message interprets the contents of the message by the holding object itself and activates the calculation activation method. A message for starting the calculation method is passed to the calculation object in the holding object by the calculation start method. Each computation object that has received the message interprets the content of the message by itself and activates a computation method. When the calculation object starts the calculation method, calculation based on the calculation method or calculation mesh described in the calculation code of each calculation object is performed.

In the analysis calculation procedure, each calculation object that has started the calculation method is input while exchanging the physical quantity object in the holding object. Here, “input while exchanging physical quantity objects” means that physical quantity objects including data necessary for analysis calculation are alternately input to a plurality of calculation objects, and a message of an input request is received from the calculation objects. This can be achieved by the physical quantity object interpreting the message by the physical quantity object itself and passing object data to the calculation object.

Therefore, one physical quantity object can be processed between different calculation objects. Then, the calculation result of the analysis calculation of one calculation object is reflected on the physical quantity object, and then input to another calculation object to perform the analysis calculation.

Such an analysis calculation is performed on the calculation object in the holding object, and the calculation result is output by the output procedure. The output of the calculation result in the output procedure can be output to an output device such as a display device or a printer device. Further, the calculation result can be output as a file to a recording device such as a floppy disk device or a hard disk device.

Therefore, a message for activating the calculation activation method of the held object may be described in the control program or the like, and the subsequent analysis calculation is performed between the calculation object and the physical quantity object. It is not necessary to determine the type of the physical quantity object and the type of the physical quantity object, and describe a procedure corresponding to each type in the control program.

Further, since it is possible to perform analysis calculation by combining different calculation methods or different calculation meshes which could not be realized until now, a calculation method according to each region of the core structure can be freely selected. can do.
As a result, even in the case of a nuclear reactor having a complicated core structure, highly accurate core analysis can be performed, and the calculation time can be reduced and the storage capacity of the computer can be reduced.

Further, even when a calculation code based on another calculation method or a calculation code based on a different mesh setting is newly added, it is only necessary to define a class corresponding to the new calculation code. Need not be corrected, and the maintainability can be improved. In particular, by inheriting the class corresponding to the existing calculation code according to the present invention, all or a part of the class is taken over as it is, and only the part to be changed or added is described as a difference, and the calculation corresponding to the new calculation code is performed. Since a class may be defined, extensibility and maintainability can be improved.

The nuclear reactor core analysis method of the present invention can be applied to, for example, all nuclear reactor analyzes using calculation codes, in addition to nuclear characteristic analysis, nuclear thermal hydraulic characteristic analysis, dynamic characteristic analysis, and the like.

According to a second aspect of the present invention, in the reactor core analysis method according to the first aspect, the physical quantity object further includes a conversion method for converting the physical quantity data into an input format of the calculation object. The analysis calculation procedure, when the calculation object makes an input request to the physical quantity object, the physical quantity object activates the conversion method and converts the physical quantity data into the input format of the input calculation request, The method further comprises a conversion procedure of inputting the converted physical quantity data to the calculation object.

According to the present invention, by generating a physical quantity object having a conversion method, even when different calculation objects have mutually different input formats, analysis calculation can be performed by combining different calculation codes. It is.

In the analysis calculation procedure, when the calculation object inputs a physical quantity object, the calculation object makes an input request to the physical quantity object. At this time, the physical quantity object that has received the input request message from the calculation object interprets the message by itself and starts the conversion method, thereby converting the object data into the input format of the input request calculation object and converting the object data into the calculation object. Pass to. Therefore, even when the input parameters of a plurality of calculation objects have different formats, one physical quantity data can be handled by a plurality of different calculation codes, and a highly accurate calculation result can be obtained.

The configuration of the conversion method of the physical quantity object is not particularly limited as long as it converts the object data corresponding to the input format of a plurality of calculation objects. For example, if the calculation object is a calculation code based on the diffusion calculation method, the physical quantity data is 1
The conversion method can be configured to convert to three-dimensional data when the calculation object is converted to three-dimensional data and the calculation object is a calculation code based on the transportation calculation method.

The medium on which the program according to the reactor core analysis method of the present invention is recorded is as described in claim 3, and the medium in which the computer analyzes the reactor core analysis program for performing a reactor core analysis is provided. ,
A plurality of diffusion calculation classes having a calculation method for performing a core analysis calculation based on a diffusion calculation method according to a region of the core structure, and a plurality of calculation classes having a calculation method for performing a core analysis calculation based on a transport calculation method according to a region of the core structure A transport calculation class, a physical quantity class having a conversion method for converting object data into an input format of the calculation code of the calculation class, and a holding class having a calculation start method for starting the calculation method are defined. An object for realizing a calculation object based on the calculation class, and a physical quantity object having physical quantity data as object data based on the physical quantity class on a storage area expressing a single calculation system Based on the materialization procedure and the holding class, the calculation object And calculus build procedure materializing in a storage area holding an object with bets and the physical quantity object as the object data, wherein the holding object, by invoking the calculation start method selectively, to the calculation method selected Based on the calculation
The object selectively invokes the calculation method,
A calculation procedure of performing an analysis calculation by inputting the physical quantity objects while exchanging the physical quantity objects with each other; and, when the calculation object requests input to the physical quantity object, the physical quantity object activates the conversion method and requests the calculation. An analysis calculation procedure having a conversion procedure of converting the physical quantity data into an input format of an object and inputting the converted physical quantity data to the calculation object, and an output procedure of outputting a calculation result by the analysis calculation procedure. It is characterized by having.

The medium having the reactor core analysis program of the present invention recorded thereon and having the same configuration as that of the above-described invention reads out the reactor core analysis program recorded on the recording medium of the present invention onto a storage area of a computer. When they are executed, the same operation and effect as those of the above-described invention can be obtained.

In the present invention, a calculation class having a calculation method and a physical quantity class having a conversion method are defined for each calculation code. Based on the data structure of each class, a calculation object is converted into an instance by an object materialization procedure. The generated physical quantity data and read in the input procedure are registered as object data, and the physical quantity object is generated as an instance.

Here, the calculation code includes a calculation program based on the same or different calculation method with different mesh settings, in addition to a calculation program based on a different calculation method. The configuration of the calculation code is not particularly limited as long as it is an analysis calculation program for performing a core analysis of a nuclear reactor. For example, a diffusion calculation program based on an existing three-dimensional mesh calculation, a transport calculation program based on a detailed mesh calculation, and the like. , A diffusion or transport node calculation program, etc., and each is generated as a separate calculation object.

The calculation method of the calculation object can be constituted by only a portion corresponding to a subroutine for solving an actual problem in the analysis calculation program. For example, a subroutine for calculating the neutron flux density can be used as the calculation method.

Therefore, when a new calculation code is used, it is sufficient to define a new calculation class using the new calculation code as a calculation method, and it is not necessary to modify the entire reactor core analysis program. In particular, by inheriting the class corresponding to the existing calculation code, all or part of the class is inherited as it is, and only the part to be changed or added is described as a difference to define the calculation class corresponding to the new calculation code. Therefore, expandability and maintainability can be improved.

The configuration of the "medium" is not particularly limited as long as the program can record a reactor core analysis program as binary data. For example, a floppy disk, a CD-ROM, a hard disk of a computer, or the like can be used as a medium.

The medium on which the nuclear reactor core analysis program of the present invention is recorded includes, for example, nuclear characteristic analysis, nuclear thermal hydraulic characteristic analysis,
In addition to dynamic characteristics analysis, it can be applied to all reactor core analyzes using calculation codes.

According to a fourth aspect of the present invention, there is provided a reactor core analysis method in which a plurality of processors perform a core analysis of a reactor by program processing based on a predetermined calculation code. Input procedure and the core structure in all processors.
A plurality of diffusion calculation objects having a calculation method for performing a core analysis calculation based on a diffusion calculation method according to a structure region, and a plurality of calculation methods having a calculation method for performing a core analysis calculation based on a transport calculation method according to a region of a core structure . A transport calculation object, a physical quantity object having physical quantity data as object data, a calculation control object having a construction method for constructing a calculation system and a calculation execution method for controlling analysis calculation, and converting any object into one-dimensional data Storing a transmission method for transmitting to another processor and a transmission / reception object having a reception method for generating the object from the one-dimensional data received from the other processor in a storage area of its own processor expressing a single computation system Object generation procedure to be generated on the area, In any one of the processors, when the calculation control object activates the construction method, the computation control object has the computation object and the physical quantity object as object data based on the analysis information, and activates the computation method of the computation object. A calculation system construction procedure for generating a holding object having an activation method on a storage area, and in all processors, the calculation control object activates the calculation execution method, so that in the holding object in each processor, Each calculation object
A calculation procedure for selectively activating the calculation method and inputting and exchanging the physical quantity object in the holding object to perform an analysis calculation, and a calculation procedure for a predetermined number of physical quantity objects, each time the calculation procedure is completed. Invoking the transmission method of the transmission / reception object, transmitting the physical quantity object updated with the calculation result by the calculation object to another processor as one-dimensional data, and in the other processor, By activating the receiving method, an analysis calculation procedure having an object update procedure that generates a physical quantity object from the received one-dimensional data and reflects a calculation result in another processor, and in a processor other than the first processor,
The transmission / reception object activates the transmission method, transmits a physical quantity object in which a calculation result in each processor is stored to the first processor as one-dimensional data, and in the first processor, the transmission / reception object By activating the reception method, a calculation result aggregation procedure of generating a physical quantity object in which a calculation result is reflected from the received one-dimensional data on a storage area, and, in the first processor, calculation results in all processors. And an output procedure for outputting.

The present invention performs a reactor core analysis based on an object-oriented programming method by a process using a plurality of processors, for example, a parallel process or a distributed process.

In the present invention, a processor that inputs analysis information, generates an area object, and outputs a calculation result among a plurality of processors is defined as a first processor. Here, the plurality of processors includes each computer when a plurality of computers are connected via a network, and a computer having a plurality of processors.

The analysis information input in the input procedure in the first processor includes initial conditions necessary for analysis calculation, calculation conditions such as information on material data maps or meshes, information such as boundary conditions and convergence judgment conditions, and information of each processor. It is possible to use the arrangement information of the computation objects distributed to each processor based on the processing capacity. still,
The input of the analysis information can be read from an input device such as a keyboard device. Alternatively, the analysis information may be stored in advance as a file in a storage device such as a floppy disk device or a hard disk device, and the file may be input by reading a file in which the analysis information is stored from the storage device.

In the present invention, a calculation object based on a calculation code is generated for each calculation code by the object generation procedure. Here, "to generate a calculation object in a storage area" means that a calculation object having the same data structure as a class in which a calculation code as a predetermined analysis program is defined as a calculation method is materialized in the storage area. Generate as an instance.

The calculation code includes a calculation program based on the same or different calculation method with different mesh settings, in addition to a calculation program based on a different calculation method. The configuration of the calculation code is not particularly limited as long as it is an analysis calculation program for performing a core analysis of a nuclear reactor. For example, a diffusion calculation program based on an existing three-dimensional mesh calculation, a transport calculation program based on a detailed mesh calculation, and the like. , A diffusion or transport node calculation program, etc., and each is generated as a separate calculation object.

The calculation method of the calculation object can be constituted by only a portion corresponding to a subroutine for solving an actual problem in the analysis calculation program. For example, a subroutine for calculating the neutron flux density can be used as the calculation method.

In the present invention, a physical quantity object is generated from physical quantity data by an object generating procedure. Here, “generate a physical quantity object in the storage area” means that the physical quantity object is registered in the storage area by registering the physical quantity data as object data while maintaining the same data structure as the previously defined class. Generate as an instance.

As the physical quantity data, for example, neutron density, directionality, neutron energy distribution, cross-sectional area, etc. can be used. Analysis information input in the input procedure, such as initial conditions, boundary conditions, etc., is treated as physical quantity data. Can also. Further, a physical quantity object can be generated for each physical quantity data.

The transmission / reception object generated by the object generation means of the present invention is used for transmission / reception of an object between a plurality of processors. For example, when parallel processing or distributed processing is performed by a plurality of processors, it is necessary to exchange objects such as physical quantity objects with each other. However, the data structure of the object itself is 2
When the values are converted and transmitted between the processors, the internal information can be referred to from other objects, so that the concealment of the data of the objects to be transmitted and received cannot be maintained.

In the present invention, since the transmission / reception object in which the part for transmitting / receiving the object is an object is used, the transmission / reception of the object can be performed only by activating the transmission method and the reception method for the transmission / reception object. Therefore, concealment of the object to be transmitted / received can be ensured.

When transmitting an object,
The transmission method converts object information into one-dimensional data. Here, the “one-dimensional data” is obtained by binarizing all information about the object such as the type information and the data value of the object to be transmitted. Further, the one-dimensional data may be configured to include the identification numbers of the destination and source processors, for example, the addresses of the processors on the network. Therefore, the processor that has received the one-dimensional data can activate the reception method of the transmission / reception object and generate an object identical to the object on the transmission side based on the information.

Therefore, the transmission / reception object makes it possible to transmit / receive the object to / from another processor while ensuring the concealment of data. Even when a plurality of processors such as parallel processing or distributed processing are used, the object-oriented programming can be performed. Reactor core analysis based on the method becomes possible.

The configuration of the transmission method of the transmission / reception object is not particularly limited as long as the information of the object is converted into one-dimensional data and transmitted to another processor. For example, after converting the information of the object into one-dimensional data, by passing the one-dimensional data to a communication control program that is operating in advance on the storage area, a communication board or a communication terminal, a communication cable, and the like can be transmitted from the communication control program. Via the other processor.

The configuration of the receiving method of the transmission / reception object is not particularly limited as long as it receives one-dimensional data from another processor and generates an object. For example, a communication control program operating a reception method in advance on a storage area receives one-dimensional data via a communication cable, a communication board or a communication terminal, and executes the reception method so that the one-dimensional data is passed to the transmission / reception object. Can be configured.

The holding objects generated by the calculation system construction procedure include a plurality of calculation objects and physical quantity objects used in the analysis calculation procedure performed by each processor. The holding object is generated based on the analysis information by activating the construction method of the calculation control object for each processor. That is, when the analysis calculation is performed on all the physical quantity objects and the calculation objects by one processor, the calculation time may be long or the calculation may not be performed due to the limitation of the storage capacity and the processing capacity. For this reason, calculation objects are distributed to the processors based on the analysis information.

When the arrangement information of the calculation objects distributed to each processor based on the processing capability of the processor is used as the analysis information, the processing capability of the plurality of processors used for the reactor core analysis may be different. Even so, the optimal number of calculation objects is arranged for each processor by the calculation system construction procedure, so that the calculation time can be reduced.

The analysis calculation by the calculation object is controlled by the calculation start method of the holding object. The configuration of the holding object is not particularly limited as long as it has a calculation object and a physical quantity object as object data and has a calculation activation method. For example, other data necessary for the analysis calculation can be registered as the object data of the held object. Further, as a method of the holding object, another method such as confirmation of a calculation result for controlling the analysis calculation can be defined.

In the analysis calculation procedure, the calculation is started by the calculation control object first passing a message for starting the calculation start method to the holding object. According to the present invention, since the calculation control object is generated and the calculation is controlled as described above, the maintainability of the calculation control itself is also improved.

Next, the holding object that has received the message interprets the contents of the message by the holding object itself and activates the calculation activation method. A message for starting the calculation method is passed to the calculation object in the holding object by the calculation start method. Each computation object that has received the message interprets the content of the message by itself and activates a computation method. When the computer object starts the calculation method, calculation based on the calculation method of each calculation object is performed.

In the analysis calculation procedure, each calculation object that has started the calculation method inputs data while exchanging physical quantity data in the held object. Here, “input while exchanging physical quantity objects” means that physical quantity objects including data necessary for analysis calculation are alternately input to a plurality of calculation objects, and a message of an input request is received from the calculation objects. This can be achieved by the physical quantity object interpreting the message by the physical quantity object itself and passing object data to the calculation object.

Therefore, one physical quantity object can be processed by different calculation objects. And
The calculation result of the analysis calculation of one calculation object is reflected on the physical quantity object. Such a calculation procedure is performed in each processor. For this reason, the physical quantity object in which the calculation result is reflected by the predetermined number of calculation procedures in each processor is reflected in the physical quantity objects in all other processors via the transmission / reception object by the object update procedure. For this reason, there is no inconsistency in the content of the physical quantity object for each processor, and a high-precision calculation result can be obtained by parallel processing using a plurality of processors or distributed processing as in the case of performing analysis calculation with a single processor. Obtainable. Further, since the physical quantity object is transmitted / received by the transmission / reception object, concealment of data of the physical quantity object is also maintained.

The configuration of the object update procedure is not particularly limited as long as the transmission method of the transmission / reception object is activated each time the predetermined number of calculation procedures are completed. For example, in the object update procedure, the transmission method of the transmission / reception method can be activated each time one calculation procedure ends. In this case, in order to reflect the calculation result to another processor for each calculation procedure, for example, even when performing the analysis calculation until the calculation result converges to a certain value, the same convergence as the analysis calculation in the single processor is performed. Nature can be secured.

Further, the object update procedure can be configured so that the transmission method of the transmission / reception object is activated when a plurality of calculation procedures are completed. In this case, even when the ratio of the time required for communication between a plurality of processors to the processing time of the entire analysis calculation is large, communication is not performed during the processing of a certain number of calculation procedures. Can be improved.

After that, for each processor, this physical quantity object is input to another calculation object and an analysis calculation is performed. Such an analysis calculation is performed for all calculation objects existing in the holding object.

All the results obtained by the calculation by the analysis calculation procedure for each processor are transmitted to the first processor and aggregated by the calculation result aggregation procedure. Then, according to the output procedure of the first processor, the calculation result in the entire calculation system is output.

For example, a neutron flux or the like is calculated for each processor, an effective neutron multiplication factor or a fission rate is calculated from the calculation result, and the neutrons are aggregated in the first processor.
Can be output.

The output of the calculation result in the output procedure is:
For example, it can be output to an output device such as a display device or a printer device. Further, the calculation result can be output as a file to a recording device such as a floppy disk device or a hard disk device.

As described above, the reactor core analysis can be performed based on the object-oriented programming method by the parallel processing or the distributed processing using a plurality of processors. Alternatively, analysis calculation can be performed by combining different calculation meshes, and a calculation method according to each region of the core structure can be freely selected. As a result, even in the case of a reactor having a complicated core structure, highly accurate core analysis can be performed, and the calculation time can be reduced and the storage capacity of the processor can be reduced.

Further, even when a new calculation code based on another calculation method is added, it is only necessary to define a class corresponding to the new calculation code. Performance can be improved. In particular, by inheriting the class corresponding to the existing calculation code according to the present invention, all or a part of the class is taken over as it is, and only the part to be changed or added is described as a difference, and the calculation corresponding to the new calculation code is performed. Since a class may be defined, extensibility and maintainability can be improved.

The reactor core analysis method of the present invention can be applied to, for example, all nuclear reactor analyzes using calculation codes, in addition to nuclear characteristic analysis, nuclear thermal hydraulic characteristic analysis, dynamic characteristic analysis, and the like.

In the present invention, the configuration and connection form of the computer are not particularly limited as long as the reactor core analysis is performed using a plurality of processors. For example, a plurality of computers can be connected to each other via a network, and a reactor core analysis can be performed by distributed processing. Further, the reactor core analysis can be performed by parallel processing of computers of a plurality of processor systems. Also, the number of processors is not particularly limited.

According to a fifth aspect of the present invention, in the reactor core analysis method according to the fourth aspect, the physical quantity object further includes a conversion method for converting the physical quantity data into an input format of the calculation object. An area object generation procedure for generating, in the first processor, an area object having physical quantity data based on analysis information as object data in a storage area corresponding to each processor; By activating the transmission method, the object transmits each area object to the corresponding processor as one-dimensional data,
In a processor other than the first processor, the transmission / reception object activates the reception method, and an area object transmission / reception procedure for generating the area object from a received one-dimensional data on a storage area, further comprising: A calculation system constructing step for generating the holding object in the storage area based on the area object, and the analysis calculation step includes, when the calculation object makes an input request to the physical quantity object, The object further comprises a conversion procedure of activating the conversion method, converting the physical quantity data into an input format of the calculation object requested to be input, and inputting the converted physical quantity data to the calculation object.

According to the present invention, by generating a physical quantity object having a conversion method, even when different calculation objects have mutually different input formats, analysis calculation can be performed by combining different calculation codes. It is.

In the analysis calculation procedure, when the calculation object inputs a physical quantity object, the calculation object makes an input request to the physical quantity object. At this time, the physical quantity object receiving the input request message from the calculation object by the conversion procedure converts the object data into the input format of the requested calculation object by interpreting the message by itself and invoking the conversion method. And enter it in the calculation object. Therefore, even when the format of input parameters of a plurality of calculation objects is different depending on the conversion method and the conversion procedure, one physical quantity data can be handled by a plurality of different calculation codes, and a highly accurate calculation result can be obtained. it can.

The configuration of the conversion method of the physical quantity object is not particularly limited as long as it converts the object data corresponding to the input format of a plurality of calculation objects. For example, if the calculation object is a calculation code based on the diffusion calculation method, the physical quantity data is 1
The conversion method can be configured to convert to three-dimensional data when the calculation object is converted to three-dimensional data and the calculation object is a calculation code based on the transportation calculation method.

Further, according to the present invention, an area object is generated by using the physical quantity data based on the analysis information as object data by the area object generation procedure, and the information for generating the holding object is converted into an object.

The area object stores analysis information as object data in the storage area of the first processor.
It is generated for each processor. Then, according to the area object transmission / reception procedure, the data is transmitted to the corresponding processor by the transmission / reception object while keeping the data concealment.

In the held object generated by the calculation system construction procedure in each processor, a calculation object and a physical quantity object are registered as object data based on the received information of the area object. Therefore, when the arrangement information of the calculation objects distributed to each processor based on the processing capability of the processor is used as the analysis information, the processing capabilities of the plurality of processors used for the reactor core analysis may be different. However, since the optimum number of calculation objects are arranged for each processor by the calculation system construction procedure, the calculation time can be reduced. Also, since the first processor generates an area object based on the physical information of each processor and transmits it to each processor, it is possible to dynamically input necessary information together with physical quantity data at the time of input. it can.

The medium on which the program according to the nuclear reactor core analysis method of the present invention is recorded is as described in claim 6, in which a nuclear reactor core analysis program for performing a nuclear reactor core analysis by a plurality of processors is recorded. In the medium, a plurality of diffusion calculation classes that have a calculation method for performing core analysis calculation based on the diffusion calculation method according to the region of the core structure, and a calculation method that performs core analysis calculation based on the transport calculation method according to the region of the core structure A plurality of transport calculation classes having a physical quantity class having a conversion method for converting the calculation code into an input form, a calculation control class having a construction method for constructing a calculation system and a calculation execution method for controlling an analysis calculation, and A transmission method for converting an object into one-dimensional data and transmitting the one-dimensional data to another processor; A transmission / reception class having a reception method for generating the object from the one-dimensional data received from the server, an area class, and a holding class having a calculation start method for starting a calculation method of the calculation object are defined. In the processor, an input procedure for inputting analysis information, and in all processors, a calculation object based on the calculation class, a physical quantity object having the physical quantity data as object data based on the physical quantity class, and a calculation control class. A single computation control object based on the
An object materialization procedure for materializing on a storage area expressing a calculation system, and, in the first processor, an area object having physical quantity data based on the analysis information as object data based on the area class, to each processor. An area object generation procedure correspondingly generated on the storage area, and in the first processor, the transmission / reception object activates the transmission method, thereby transmitting each area object to the corresponding processor as one-dimensional data. In addition, in a processor other than the first processor, the transmission / reception object activates the reception method to generate the region object from the received one-dimensional data in the storage area. In the above, the calculation control object activates the construction method, so that the computation control object has the computation object and the physical quantity object as object data based on the area object, and activates the calculation method of the computation object. and procedures calculus build to generate a holding objects in a storage area having, in all processors, by the calculation control object activates the calculation execution method, within the holding objects in each processor, each calculated object
Is a calculation procedure for selectively activating the calculation method and performing an analysis calculation by inputting and exchanging the physical quantity objects in the holding object, and when the calculation object requests input to the physical quantity object. A conversion procedure for invoking the conversion method to convert the physical quantity data into the input format of the calculation object requested to be input, and inputting the converted physical quantity data to the calculation object; Each time the calculation procedure for the object is completed, the transmission / reception object activates the transmission method, transmits the physical quantity object updated with the calculation result by the calculation object as one-dimensional data to another processor, and The transmission / reception object An object update procedure for generating the physical quantity object from the received one-dimensional data by the object invoking the reception method, and reflecting the calculation result in another processor; and the first processor In the other processor, the transmitting and receiving object activates the transmission method, transmits the physical quantity object in which the calculation result in each processor is stored as one-dimensional data to the first processor, and the first processor In the above, the transmission / reception object activates the reception method, a calculation result aggregation procedure of generating a physical quantity object in which a calculation result is reflected from the received one-dimensional data of the physical quantity object in a storage area, and the first processor In all processors Characterized by comprising an output procedure for outputting the calculation result that.

The medium having the reactor core analysis program of the present invention recorded thereon and having the same configuration as that of the above-described invention reads out the reactor core analysis program recorded on the recording medium of the present invention into a storage area of a computer. When they are executed, the same operation and effect as those of the above-described invention can be obtained.

In the present invention, a calculation class having a calculation method and a physical quantity class having a conversion method are defined for each calculation code, and a calculation object is generated by an object materialization procedure based on the data structure of each class. The physical quantity data read in the input procedure is registered as object data to generate a physical quantity object.

Here, the calculation codes include calculation programs based on the same or different calculation methods with different mesh settings, in addition to calculation programs based on different calculation methods. The configuration of the calculation code is not particularly limited as long as it is an analysis calculation program for performing a core analysis of a nuclear reactor. For example, a diffusion calculation program based on an existing three-dimensional mesh calculation, a transport calculation program based on a detailed mesh calculation, and the like. , A diffusion or transport node calculation program, etc., each of which is generated as a separate calculation object.

Further, the calculation method of the calculation object can be constituted by only a portion corresponding to a subroutine for solving an actual problem in the analysis calculation program. For example, a subroutine for calculating the neutron flux density can be used as the calculation method.

Therefore, when a new calculation code is used, it is sufficient to define a new calculation class using the new calculation code as a calculation method, and it is not necessary to modify the entire reactor core analysis program. Therefore, according to the present invention,
The maintainability of the program can be improved. In particular,
By inheriting the class corresponding to the existing calculation code according to the present invention, all or a part of the class is inherited as it is, and only the part to be changed or added is described as a difference to calculate the calculation class corresponding to the new calculation code. Since it suffices to define, scalability and maintainability can be improved.

Here, the configuration of the “medium” is not particularly limited as long as it can record a reactor core analysis program as binary data. For example, a floppy disk, a CD-ROM, a hard disk, or the like can be used as a medium.

The medium on which the nuclear reactor core analysis program of the present invention is recorded includes, for example, nuclear characteristic analysis, nuclear thermal hydraulic characteristic analysis,
In addition to dynamic characteristics analysis, it can be applied to all reactor core analyzes using calculation codes.

In the present invention, as long as the reactor core analysis uses a plurality of processors, the configuration and connection form of the processors are not particularly limited. For example, a plurality of computers can be connected to each other via a network to form a cluster, and can be read and executed in a storage area of a computer in which a reactor core analysis is performed by distributed processing. Further, it can also be executed by reading into a storage area of a computer in which a reactor core analysis is performed by parallel processing of a plurality of processor systems.

[0094]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a mixed oxide fuel assembly (hereinafter referred to as a “MOX fuel assembly”) of enriched uranium and plutonium dioxide and a reactor in which a uranium fuel assembly is simultaneously loaded. It is a flowchart figure in case the reactor core analysis method is used.

In the present embodiment, two computers A and B are connected to a network to perform analysis and calculation in parallel. FIG. 2 shows a schematic configuration of a computer used in the reactor core analysis of the present embodiment. The computer A has a hard disk 1 in which an executable file of a reactor core analysis program is stored, a floppy disk device 3 as an input device for inputting arrangement map information of calculation objects as analysis information, and an analysis device. A printer device 5 as an output device for outputting a calculation result is connected.

The computer B has a hard disk 7 in which an executable file of the reactor core analysis program is stored. Computer A and computer B have built-in Ethernet boards 6, 8
And a cable 9. A communication program operates on the storage areas 11 and 13 of the computers A and B, respectively. In addition, the storage areas 11 and 13
Is a part where the core analysis program of the present embodiment is read and various objects are generated during execution of the program. For this reason, computers A and B
Core analysis programs can be executed in parallel,
In addition, data can be mutually transmitted and received.

As shown in FIG. 1, in the reactor core analysis in the present embodiment, the computer A uses the input file reading step (procedure), the object materializing step, the area object generation step, the area object transmission step, the calculation system construction step , Calculation start step, inner iteration step, convergence determination step, update step, outer iteration step, effective multiplication factor convergence determination step, calculation result receiving step, and output step. In the computer B, the object materializing step, the area object receiving step, the calculation system construction step, the calculation starting step, the inner iteration step, the convergence determination step, the update step, the outer iteration step, the effective multiplication factor convergence determination step, and the transmission of the calculation result are performed. It consists of steps.

In the computer A and the computer B, the core analysis program stored in each of the hard disks 1 and 7 is activated, so that
The core analysis program is read out and executed in 1 and 13. Thus, the core analysis is started. Here, the core analysis program may be stored only in the hard disk 1 of the computer A and read into the storage area of the computer B via a network.

Next, in the input file reading step, from the floppy disk device 3 of the computer A,
An input file storing initial conditions, boundary conditions, and arrangement information of calculation objects to be distributed to the computers A and B is input and read into a storage area. This arrangement information also includes information for performing a transport calculation object for neutron flow data relating to the MOX fuel assembly and performing a diffusion calculation object for neutron flow data relating to the uranium fuel assembly. Here, as the neutron flow data, neutron density, directionality, energy distribution, cross-sectional area, and the like are used, and these are stored in the hard disks 1 and 7 in advance. The initial conditions and boundary conditions of the analysis information input from the floppy disk device 3 are also used as neutron flow data.

In the core analysis program, a plurality of diffusion calculation classes having different mesh intervals, a plurality of transport calculation classes having different mesh intervals, a neutron flow class, a holding class, a calculation control class, and a transmission / reception class are defined.

FIG. 3 is a diagram showing a relationship between objects generated on the storage areas of the computer A and the computer B in the present embodiment.

When the core analysis program is executed, an object is first materialized from each class in an object materializing step. That is, diffusion calculation objects 15A and 15B and transportation calculation objects 16A and 16B are generated in the storage areas of the computers A and B for each mesh. The neutron flow objects 17A and 17B, the calculation control object 21A,
21B and transmission / reception objects 23A and 23B are generated.

Here, in each of the diffusion calculation classes, a calculation method for performing an analysis calculation using the diffusion calculation code is registered, and in each of the transportation calculation classes, a calculation method for performing the analysis calculation using the transportation calculation code is registered. I have.

Further, in the neutron flow class, a conversion method for converting the object data into the input format of the diffusion calculation code and the transport calculation code is registered.

In the transmission / reception class, information of a neutron flow object or area object is converted into one-dimensional data including binary data corresponding to data type information, object data and binary data of a method, and converted into another computer. A transmission method to be transmitted to the server and a reception method to generate an object having the same structure as the transmitted object by analyzing the type information, object data, and method of the corresponding data from the one-dimensional data received from another computer are registered. .

The holding class includes a calculation start method for passing a message for starting a calculation method to the diffusion calculation object or the transport calculation object, a determination method for determining convergence of the calculation result, and an object data of the neutron flow object. And an inspection method for checking the convergence of the effective multiplication factor by a calculation process based on each calculation result of the neutron flow object are registered.

In the calculation control class, a construction method for constructing a calculation system by materializing a held object and a calculation execution method for controlling a calculation by activating an analysis calculation and transmitting a message to the held object are registered. I have.

Next, in the area object generation step,
In the computer A, an area object A and a computer B corresponding to the computer A are obtained from the area class based on the arrangement information of the calculation objects in the computer.
Is generated in the storage area of the computer A. Each region object is associated with an arrangement information map of calculation objects, that is, each computer and a calculation object used by the computer.

In the area object transmitting step, the area object B is transmitted to the computer B. In other words, the calculation control object 21A activates the transmission method with respect to the transmission / reception object 23A using the area object B and the address of the transmission destination computer B on the network as arguments. Then, the transmission / reception object 23A converts the area object B into one-dimensional data including binary data corresponding to the data type information and binary data of the object data, and transmits the converted data to the computer B by a communication program via a network. Sent. The structure of the one-dimensional data is not limited to this, and may be a data structure including information such as a network address of a transmission source computer and a network address of a transmission destination computer.

On the other hand, in the area object receiving step of the computer B, the transmitting / receiving object 23B in the computer B receives the one-dimensional data transmitted from the computer A by activating the receiving method,
The one-dimensional data is analyzed, and an area object B is generated on the storage area from the data type information and the binarized data of the object data. Therefore, the area object is converted into one-dimensional data and transmitted. However, since transmission and reception can be performed only by the transmission method and the reception method of the transmission / reception object, the concealment of the data of the area object is maintained. Transmission and reception are possible.

Thereafter, the computer A generates the holding object 19A in the storage area by activating the construction method of the calculation control object 21A, and refers to the information of the area object A to calculate the calculation object used for the analysis calculation. 15A, 16A and the neutron flow object 17A are registered as object data. Similarly, in the computer B, the holding object 19B is generated in the storage area by the calculation control object 21B, and the calculation objects 15B and 16B and the neutron flow object used for the analysis calculation are referred to by referring to the information of the area object B. 17B is registered as object data. In this way, since the calculation objects 15 and 16 used in the computers A and B are distributed and the calculation system is constructed, the calculation can be designed according to the processing capacity and the storage capacity of the computer, and the calculation time can be reduced. It is possible to reduce the number of storage units used.

Next, the computer A and the computer B
In each calculation start step, a message for starting the calculation execution method is passed to the calculation control object 21. When the calculation control object 21 starts the calculation execution method, the calculation control object 21 passes a message for starting the calculation start method to the holding object 19. Holding object 1 that received this
9 passes a message for activating a calculation method to arbitrary calculation objects 15 and 16 existing in the holding object. Thereby, the calculation object 15,
In step 16, the calculation method is activated to start the analysis calculation.

In the inner iteration step, the calculation objects 15 and 16 that have started the calculation method input and exchange the necessary neutron flow objects 17 with each other to perform the analysis calculation. That is, the neutron flow object 17 including the data necessary for the analysis calculation is
5 and 16 are input alternately. Calculation object 15,
The neutron flow object 17, which has received the input request message from the computer 16, interprets the message by itself and activates the conversion method, thereby converting the object data into the input format of the input requested calculation object and passing it to the calculation object. Specifically, when the calculation object is a calculation code based on the diffusion calculation method, the neutron flow data is converted into one-scalar data only in the representative direction, and when the calculation object is a calculation code based on the transport calculation method, the direction is calculated. Is converted to multiple scalar data representing the neutron flow.

Therefore, even when the format of input parameters of a plurality of calculation objects is different depending on the conversion method, it is possible to handle one neutron flow data with a plurality of different calculation codes, and obtain a highly accurate calculation result. Can be. Then, the calculation result of the analysis calculation of one calculation object is reflected on the neutron flow object 17 in the holding object, and thereafter, is input to another calculation object to perform the analysis calculation.

By performing the calculation by the calculation method of each calculation object for each computer a predetermined number of times, the neutron flow object on which the calculation result is reflected can be obtained as follows.
It is reflected on the neutron flow object in another computer via the transmission / reception object.

That is, the computation control object 21A of the computer A activates the transmission method using the neutron flow object 22A and the address of the computer B on the network as arguments for the transmission / reception object 23A. Then, the transmission / reception object 23A converts the neutron flow object 22A into the data corresponding to the data type information.
The data is converted into one-dimensional data composed of binarized data of the binarized data and the object data, and transmitted to the computer B by a communication program via a network.

On the other hand, in the computer B, the transmission / reception object 23B in the computer B receives the one-dimensional data transmitted from the computer A by activating the reception method, analyzes the one-dimensional data, and obtains data type information. Then, the neutron flow object 22B is generated in the storage area from the binarized data of the object data. Therefore, in the computer B, a neutron flow object 22B exactly the same as the neutron flow object 22A transmitted from the computer A is generated. The transmission of the neutron flow object from the computer B to the computer A is performed in the same procedure.

Therefore, there is no inconsistency in the content of the neutron flow object for each computer, and a highly accurate calculation result can be obtained. Further, the neutron flow object is transmitted after being converted into one-dimensional data, but can be transmitted and received only by the transmission method and the reception method of the transmission / reception object, so that the concealment of the data of the neutron flow object is maintained, Reliable transmission and reception becomes possible. Such an analysis calculation is performed for all calculation objects in the held object.

It is also possible to configure so that the transmission / reception object is activated and the neutron flow data is transmitted / received between the computers A and B each time the computation by the activation of the computation method of each computation object is completed.

Next, in the convergence determination step, the calculation control object 21 passes a message for activating the determination method to the holding object 19. The holding object 19 that has received this message interprets the message by itself and activates the determination method. This allows
It is checked whether the calculation result has converged, and if not, the inner iteration step is repeated.

If the calculation results converge, the calculation control object 21 passes a message for activating the update method to the holding object 19 in the update step. Holding object 19 that received this message
Interprets the message itself and invokes the update method. By this update method, neutron flow data 1
7 is updated with the converged calculation result.

In the outer iteration step, the computation control object 21 passes a message for activating the check method to the holding object. The holding object receiving this message interprets the message itself,
Invoke the inspection method. Thus, the effective neutron multiplication factor is calculated from the updated neutron flow data, and it is checked whether or not the effective multiplication factor has converged. If the effective multiplication factor has converged, the analysis calculation is ended. If not, the inner iteration step, the determination step, and the update step are repeated until the effective multiplication factor converges. Here, in the present embodiment, the effective neutron multiplication factor is finally calculated. However, the present invention is not limited to this. For example, a neutron flux, an output distribution, and the like can be obtained.

In the calculation result transmission step in the computer B, when the analysis calculation is completed, the object holding the calculation result of the effective multiplication factor is transmitted to the computer A by the transmission / reception object. At this time, the point that the object holding the calculation result is converted into one-dimensional data
Same as above.

In the step of receiving the calculation result in the computer A, the computer B receives
Receives the one-dimensional data transmitted from, and generates an object holding the calculation result holding the calculation result.
As described above, the results of the analysis calculation in each computer are collected in the computer A.

In the output step, the computer A outputs the calculation results of the computers A and B to the printer 5.

As described above, in the present embodiment, since the reactor core analysis can be performed based on the object-oriented programming method by the distributed processing using two computers, it has not been possible to realize it until now. The analysis calculation can be performed by combining the diffusion calculation code and the transport calculation code. That is, the area of the MOX fuel assembly in the core can be calculated based on the transport calculation code, and the area of the uranium fuel assembly can be calculated based on the diffusion calculation code. As a result, highly accurate core analysis can be performed, and the calculation time can be reduced and the storage capacity of the computer can be reduced.

In the present embodiment, the core of the nuclear reactor in which the MOX fuel assembly and the uranium fuel assembly are loaded at the same time is analyzed, but the analysis target is not limited to this. For example, the present invention can be applied to a fuel assembly containing gadolinia and a baffle plate.

Even when a new calculation code based on another calculation method is added, it is only necessary to define a class corresponding to the new calculation code. Performance can be improved.

In this embodiment, two computers are used, but the number is not limited to this.
By increasing the number of connected computers, analysis calculations can be performed more efficiently.

In this embodiment, the distributed processing is performed by connecting two computers to the Ethernet. However, the number of connections, the connection form, and the configuration of the computers are not limited to these. Analytical calculations can also be performed by parallel processing by the system.

Further, in the present embodiment, calculation classes based on different calculation methods are defined and respective calculation objects are generated. However, calculation classes having different mesh settings based on the same calculation method are defined and each calculation object is defined. It is also possible to generate a calculation object and perform an analysis calculation. In this case, for example, if the output of the fuel rods inside a certain fuel assembly is calculated with high accuracy, while the evaluation of the average output is sufficient for the assemblies other than the fuel assembly, the analysis in the former region is performed. In the calculation, a calculation object in which a mesh is set in detail is used, and in the analysis calculation in the latter region, the analysis calculation is performed using a calculation object in which a coarse mesh is set. Therefore, by setting a calculation code as an object, it is possible to freely set a calculation mesh, which has been difficult until now.

It is also possible to define a calculation class in which different calculation meshes are set for different calculation codes. In this case, even if the optimum calculation mesh setting or the calculation method is not known in advance, a highly accurate analysis calculation result can be obtained by dynamically selecting the calculation mesh or the calculation method during the calculation. . For example, in the core analysis of critical fuel assemblies, if the system is becoming smaller and more complex and knowledge from past data on analysis cannot be obtained, the optimal selection of mesh settings and calculation methods Will be difficult. In this case, by defining a large number of calculation classes using calculation codes based on different mesh settings and different calculation methods, and inputting the prediction accuracy conditions as initial conditions, the optimum mesh settings and calculation methods can be dynamically calculated during execution. It is possible to perform the analysis calculation by changing or selecting.

[0133]

As described above, the present invention provides a calculation code based on a different calculation method by generating a calculation code and physical quantity data handled in an analysis calculation as an object and performing the analysis calculation based on an object-oriented programming method. Alternatively, different calculation meshes can be combined. Thereby, it is possible to freely combine different calculation methods or calculation mesh settings depending on the structure of the core to be analyzed, and to obtain an effect of obtaining a highly accurate calculation result. At the same time, there is an effect that the calculation time can be reduced and the storage capacity of the computer can be effectively reduced.

In the present invention, since a calculation class is defined for each calculation code based on a different calculation method, there is an effect that maintenance such as a program change operation in the future is easy and maintainability is improved.

According to the present invention, when a plurality of processors are used, the object is transmitted and received between the plurality of processors by the transmission / reception object, so that the concealment of the object can be maintained. For this reason, even when a plurality of processors are used, the analysis calculation based on the object-oriented programming method can be performed, and there is an effect that a highly accurate calculation result can be obtained. Further, since the analysis calculation can be performed by the object-oriented programming method even with a plurality of processors, the calculation time can be reduced, the storage capacity used can be reduced, and the maintainability can be improved.

[Brief description of the drawings]

FIG. 1 is a flowchart of a reactor core analysis in the present embodiment.

FIG. 2 is a schematic configuration diagram of a computer used for a reactor core analysis in the present embodiment.

FIG. 3 is a relationship diagram of objects generated in the embodiment.

FIG. 4 is a diagram showing the relationship between objects and messages in the object-oriented programming method.

[Explanation of symbols]

1, 7: Hard disk drive 3: Floppy disk drive 5: Printer 6, 8: Ethernet board 9: Network cable 11, 13: Storage area 15A, 15B: Diffusion calculation object 16A, 16B: Transport calculation object 17A, 17B, 22A , 22B: Neutron flow object 19A, 19B: Holding object 21A, 21B: Calculation control object 23A, 23B: Transmission / reception object

Continuation of the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G21C 17/00 G06F 9/44

Claims (6)

(57) [Claims] In a reactor core analysis method for performing a core analysis of a reactor by a computer based on a program code based on a predetermined calculation code, an input procedure for inputting analysis information and an area corresponding to an area of a core structure. A plurality of diffusion calculation objects having a calculation method for performing a core analysis calculation based on a diffusion calculation method, and a plurality of transportation calculation objects having a calculation method for performing a core analysis calculation based on a transport calculation method according to a region of the core structure; and Displays a single calculation system with a physical quantity object having physical quantity data as data
An object generation procedure to be generated on the storage area to be displayed; and a storage object having the calculation object and the physical quantity object as object data, and having a calculation start method for starting the calculation method for the calculation object, on the storage area. The calculation system construction procedure to be generated, and the holding object selectively selects the calculation start method .
By launching on the basis of the selected calculation method
Each calculation object selectively activates the calculation method, inputs and exchanges the physical quantity objects with each other, performs an analysis calculation procedure, and outputs a calculation result by the analysis calculation procedure. A reactor core analysis method comprising: 2. The physical quantity object further comprises a conversion method for converting the physical quantity data into an input format of the calculation object, wherein the analysis calculation procedure requests the calculation object to input to the physical quantity object. When the physical quantity object further comprises a conversion step of invoking the conversion method, converting the physical quantity data into the input format of the calculation object requested to be input, and inputting the converted physical quantity data into the calculation object. The reactor core analysis method according to claim 1, wherein: 3. A plurality of diffusion media having a calculation method for performing a core analysis calculation based on a diffusion calculation method in accordance with a region of a core structure in a medium in which a computer records a reactor core analysis program for performing a core analysis of a reactor by a computer. Calculation classes, a plurality of transport calculation classes having a calculation method for performing core analysis calculation based on the transport calculation method according to the area of the core structure, and a conversion method for converting object data into the input format of the calculation code of the calculation class. A holding class having a calculation start method for starting the calculation method, an input procedure for inputting analysis information, a calculation object based on the calculation class, and a physical amount based on the physical amount class. single computing a physical quantity objects having data as object data System
An object materialization procedure for materializing on a storage area that expresses a calculation system, and a calculation system construction procedure for materializing a holding object having the calculation object and the physical quantity object as object data on the storage area based on the holding class. The holding object selects the calculation activation method
Manner by invoking, based on the calculation method selected
The calculation object selectively invokes the calculation method, exchanges the physical quantity objects with each other to input and perform an analysis calculation, and the calculation object requests the physical quantity object for an input. The physical quantity object activates the conversion method, converts the physical quantity data into the input format of the calculation object requested to be input, and inputs the converted physical quantity data to the calculation object. A medium recording a reactor core analysis program, comprising: a calculation procedure; and an output procedure for outputting a calculation result by the analysis calculation procedure. 4. A reactor core analysis method in which a plurality of processors perform a core analysis of a reactor by a program process based on a predetermined calculation code, wherein an input procedure for inputting analysis information in a first processor is provided. in all the processors, and a plurality of diffusion calculation object having a calculation method for performing core analysis calculated based on expanding <br/> dispersion calculation method according to the region of the core structure, in the region of the core structure
A plurality of transport calculation objects having a calculation method for performing core analysis calculation based on the transport calculation method, a physical quantity object having physical quantity data as object data, a construction method for building a calculation system, and a calculation execution method for controlling analysis calculation And a transmission method for converting an arbitrary object into one-dimensional data and transmitting the object to another processor, and converting the object from the one-dimensional data received from the other processor.
An object generation procedure for generating, on a storage area, a transmission / reception object having a reception method for generating on a storage area of its own processor that represents a single computation system; and To generate, on the storage area, a holding object having the calculation object and the physical quantity object as object data and having a calculation start method for starting the calculation method of the calculation object based on the analysis information. and calculus build procedure, all of the processors, by the calculation control object activates the calculation execution method, within the holding objects in each processor, each calculation of
The object selectively activates the calculation method, inputs the physical quantity object in the holding object while exchanging the physical quantity object, and performs an analysis calculation, and every time the calculation procedure for a predetermined number of physical quantity objects is completed. Invoking the transmission method of the transmission / reception object, transmitting the physical quantity object updated with the calculation result by the calculation object to another processor as one-dimensional data, and in the other processor, By activating the reception method, an analysis calculation procedure having an object update procedure for generating a physical quantity object from the received one-dimensional data and reflecting a calculation result in another processor, and in a processor other than the first processor, The transmission / reception object The object activates the transmission method, transmits the physical quantity object storing the calculation result in each processor to the first processor as one-dimensional data, and in the first processor, the transmission / reception object By activating the method, a calculation result aggregation procedure of generating a physical quantity object in which a calculation result is reflected from the received one-dimensional data in a storage area, and outputting the calculation results in all the processors in the first processor A reactor core analysis method comprising: an output procedure. 5. The physical quantity object further comprises a conversion method for converting the physical quantity data into an input format of the calculation object. The first processor converts physical quantity data based on analysis information as object data. An area object generating procedure for generating an area object corresponding to each processor in a storage area; and in the first processor, the transmitting / receiving object activates the transmission method, and the processor corresponding to each area object is activated. In the processor other than the first processor, the transmitting / receiving object activates the receiving method to generate the area object from the received one-dimensional data in a storage area. Object transmission / reception procedure, wherein the computation system construction procedure generates the holding object in a storage area based on the area object, and the analysis computation procedure includes the computation object being the physical quantity object. The physical quantity object activates the conversion method, converts the physical quantity data into the input format of the requested calculation object, and inputs the converted physical quantity data to the calculation object The reactor core analysis method according to claim 4, further comprising: 6. A medium storing a reactor core analysis program for performing a core analysis of a reactor by a plurality of processors, the medium having a calculation method for performing a core analysis calculation based on a diffusion calculation method according to an area of a core structure. Diffusion calculation class, a plurality of transport calculation classes having a calculation method for performing core analysis calculation based on the transport calculation method according to the region of the core structure, and a physical quantity class having a conversion method for converting the input format of the calculation code, A calculation control class having a construction method for constructing a calculation system and a calculation execution method for controlling analysis calculations, a transmission method for converting an arbitrary object into one-dimensional data and transmitting the data to another processor, and a reception method from the other processor Transmission / reception class having a reception method for generating the object from the obtained one-dimensional data, and an area A class and a holding class having a calculation start method for starting a calculation method of the calculation object are defined, and in a first processor, an input procedure of inputting analysis information; A single calculation of a calculation object based on the physical quantity class, a physical quantity object having the physical quantity data as object data, a calculation control object based on the calculation control class, and a transmission / reception object based on the transmission / reception class. An object materialization procedure for materializing on a storage area expressing a system; and in the first processor, an area object having physical quantity data based on the analysis information as object data based on the area class corresponds to each processor. do it An area object generation procedure to be generated on the storage area, and in the first processor, the transmission / reception object activates the transmission method, thereby transmitting each area object to the corresponding processor as one-dimensional data, In a processor other than the first processor, the transmission / reception object activates the reception method, thereby generating an area object on a storage area from the received one-dimensional data. When the calculation control object activates the construction method, the calculation control object has the calculation object and the physical quantity object as object data based on the area object, and the calculation method of the calculation object. A computation system construction procedure for generating a holding object having a calculation activation method for activating a load on a storage area, and in all processors, the computation control object activates the computation execution method to execute the computation execution method. In the holding object, each of the calculation objects
Object , selectively activates the calculation method, inputs and exchanges the physical quantity object in the holding object to perform an analysis calculation, and when the calculation object makes an input request to the physical quantity object. The physical quantity object activates the conversion method, converts the physical quantity data into the input format of the calculation object requested to be input, and converts the converted physical quantity data into the calculation object. Each time the calculation procedure for the physical quantity object is completed, the transmission / reception object activates the transmission method and transmits the physical quantity object updated with the calculation result by the calculation object as one-dimensional data to another processor. In another processor, the An analysis calculation procedure having an object update procedure in which the communication object activates the reception method to generate the physical quantity object from the received one-dimensional data and reflects a calculation result in another processor; and In a processor other than a processor, the transmission / reception object activates the transmission method, transmits the physical quantity object in which a calculation result in each processor is stored to the first processor as one-dimensional data, and In the processor, the transmission / reception object activates the reception method, and a calculation result aggregation procedure of generating a physical quantity object in which a calculation result is reflected from the received one-dimensional data of the physical quantity object on a storage area; In the processor, all processes A medium for recording a reactor core analysis program, comprising: an output procedure for outputting a calculation result in the reactor.