JP2006075529A - Distributed simulation system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a distributed simulation system constituted of a plurality of federations for allowing a federation to freely attend or leave a simulation under performance. <P>SOLUTION: When a certain federation among the respective federations attends/leaves a simulation in course of the simulation, a processing to report attendance/withdrawal is added to the federation through the use of an instruction. When "attendance" is received in another previously attending federation, whole kinds of information to be transmitted only in the case of updating are re-transmitted, so as to enable the federation which attends the simulation in course of the simulation have necessary information. When "withdrawal" is received, information related to the federation to withdraw is consistently deleted from an internally managed table. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a distributed simulation system in which a plurality of objects existing in a distributed environment exchange information via a common infrastructure such as RTI (Run-Time Infrastructure).

  In recent years, demand for simulation systems has increased in various fields. Against this background, HLA (High Level Architecture) has been formulated for the purpose of improving reusability and interoperability of simulation systems developed in the past and reducing future development and maintenance costs.

  The HLA is a specification for connecting different simulation systems, and includes rules, an object model template (OMT), and interface specification elements. Of these elements, the interface specification is implemented by RTI. In particular, a simulation object (simulator) having an interface specification that can be connected to an RTI is referred to as federation, and the set, that is, the entire simulation is referred to as federation. RTI is positioned as middleware for realizing simulation.

  By forming a simulation system based on such a framework, resources developed in the past can be used effectively, and a large-scale system in a distributed environment can be built, resulting in significant benefits. be able to. Patent Document 1 below discloses a game system constructed using this type of technology.

By the way, this type of simulation system is generally composed of a plurality of federations (simulators), but its constituent elements (members) are fixed during the simulation. That is, the model to be simulated (participating federate) is determined before the execution of the simulation, and the state (participating federation) cannot be changed during the execution of the simulation. In other words, the federation constituting the federation cannot be dynamically changed from the start to the end of the simulation. For this reason, if it is desired to change the federation as a member during the simulation, the simulation must be temporarily suspended, which is very inconvenient.
JP 2000-350867 A

As described above, in the existing distributed simulation system, the simulator cannot freely participate in or leave the running simulation. For this reason, in order to realize a simulation in a more realistic manner, a very undesirable procedure of temporarily interrupting the simulation is required.
The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a distributed simulation system in which a federation can freely participate in and leave a running simulation in a distributed simulation system including a plurality of federates. It is to provide.

  In order to achieve the above object, according to an aspect of the present invention, there is provided a distributed simulation system using a plurality of computers connected to each other via a network, wherein the computer behaves as a simulation model participating in a simulation. A storage unit that stores a management table for determining the simulation, a simulation progress control unit that autonomously determines the behavior of each simulation model based on the contents of the management table, and advances the simulation, and When a simulation model requesting participation in the simulation occurs during the progress, notification means for notifying the simulation model of information for updating the contents of the management table; When the simulation model to request the withdrawal from the simulation occurs, the distributed simulation system, characterized by comprising a deleting means for deleting the information relating to the simulation model from the contents of the management table is provided.

  By taking such a means, it becomes possible to provide each federation constituting the simulation system with information regarding a newly participating federation and information regarding a federation leaving the simulation without omission. Therefore, it is possible to easily change the scenario data even while the simulation is in progress, thereby making it possible to realize the participation / leaving during the federation. As a result, simulation study and execution in a form close to reality can be performed.

  According to the present invention, it is possible to provide a distributed simulation system in which a federation can freely participate in and leave a running simulation in a distributed simulation system including a plurality of federates.

  FIG. 1 is a system diagram showing an embodiment of a distributed simulation system according to the present invention. In this embodiment, so-called simulation is referred to as federation. Federation is a term defined in the HLA interface specification realized based on RTI, and various objects generated during federation are called federation.

  In FIG. 1, a plurality of computers PC1 to PC3 are connected to each other via a network 100 to form a distributed processing environment. The computers PC1 to PC3 include an interface unit (I / F) 21, a display unit 22, a storage unit 23, a control unit 24, and a user interface unit that serve as an intermediary for exchanging information with other computers via the network 100. 25. The storage unit 23 stores scenario data 23a for defining the behavior of each federation in a predetermined storage area. The user interface unit 25 includes a keyboard, a mouse (not shown), and the like, and accepts a user operation using a GUI (Graphical User Interface) environment on the display unit 22.

  The control unit 24 of the PC 1 receives the RTI. an exe file (reference numeral 24a) and a federation application 24b. RTI. The exe file 24a is a control program for causing the control unit 24 to operate as an execution body for providing the RTI environment. The federation application 24b is a control program for realizing federation according to various specifications requested by the user. Of these, the federation application 24b is also loaded into the control unit 24 of the PC2 and PC3.

  The federation application 24b in each of the PC1 to PC3 is connected to the RTI. Federation is realized by performing processing such as call, generation, and disappearance of an object under the management of the exe 24a. That is, the federation application 24b and the RTI. The exe 24a forms an environment in which federation proceeds by autonomous determination of federation based on the scenario data 23a.

The control unit 24 of the PC 3 includes a notification processing unit 24c and a deletion processing unit 24d as software processing functions according to the present embodiment, in addition to the federation application 24b.
When a federation requesting participation in the federation occurs during the federation, the notification processing unit 24c notifies the federation of information necessary for updating the contents of the scenario data 23a. That is, when a certain federation requests mid-participation / withdrawal, the notification processing unit 24c notifies the other federation of participation / leaving using the interaction. As a result, the federation that has already participated in the ongoing federation retransmits all types of information to be transmitted only when it is updated when “participation” is notified. As a result, all the necessary information can be acquired by the federate who participates on the way.

  The deletion processing unit 24d deletes information related to the federation from the contents of the scenario data 23a when a federation requesting the withdrawal from the federation occurs during the progress of the federation. As a result, the federate notified of “leave” deletes information related to the leaving federation from the internally managed scenario data 23a so that there is no contradiction.

  FIG. 2 is a schematic diagram showing a state in which the RTI manages the passage of time in federation. The federations participating in the federation are indicated by federations AC. Each federation A-C periodically queries the RTI for time. The RTI grants permission for this inquiry, and each federation A to C can proceed with its own processing only after receiving this permission.

  FIG. 3 is a diagram showing an outline of RTI time management. The federation requesting midway participation or withdrawal from the federation notifies the RTI of midway participation or midway withdrawal. Then, by receiving information on the simulation execution time from the RTI, it is possible to perform midway participation or midway withdrawal. As described above, the RTI includes a mechanism (service) for managing time by executing a query and a permission for the query as a pair. In this embodiment, this is used to realize the participation / withdrawal during the federation.

  In FIG. 3, the federation A inquires the RTI to proceed to the federation time “15s”. In response to this, the RTI permits the federation A to proceed to the federation time “15s”. The same applies to the federation B, and each time the time required for each federation changes, an inquiry to the RTI is executed one by one. In this way, the time progress in the federation is managed by executing the inquiry and permission for the RTI as a pair.

  FIG. 4 is a schematic diagram showing in more detail the federation RTI participation during the RTI. In FIG. 4, when the federate D requests midway participation, the fact is notified to the federates A to C via the network (step S1). Then, each of the federations A to C adds information on the federation D to the information table (scenario data 23a) managed internally (step S2). Next, each of the federations A to C retransmits information about itself to the federation D (step S3). Here, in particular, only information to be transmitted at the time of information update is retransmitted.

  FIG. 5 is a schematic diagram showing in more detail the midway departure of the federate from the RTI. In FIG. 5, when the federate D requests a midway withdrawal, the fact is notified to the federates A to C via the network (step S11). Then, each of the federations A to C deletes information related to the federation D from the information table (scenario data 23a) managed internally (step S12).

  Existing simulation systems have been constructed to send and receive a variety of information, including information that is not often updated between federates. For this reason, always disclosing such information as an attribute of an object is to be transmitted / received only when updated by interaction from the viewpoint of increasing cost (processing time, communication bandwidth).

  For this reason, when a certain federation participates in the middle of the simulation, the federated who participated in the middle cannot know these information at the time of mid-participation. It was. Further, when a certain federation leaves during the simulation, other federates constituting the simulation system cannot know the federation information of the midway withdrawal. Therefore, it is not possible to realize the middle of the federation.

  On the other hand, in this embodiment, in each federation, when a certain federation joins / leaves halfway, a process of notifying participation / leaving using interaction is added to the federate. In other already participating federates, when "participation" is received, all the types of information that are being sent are retransmitted only when updated, and there is no missing information in the federated participation I did it. Also, when "leave" is received, the information related to the federation that leaves the table managed internally is deleted so that there is no contradiction.

  Since this is done, a distributed simulation system is built that recognizes the simulation time during which the federation that participates and leaves is running, so that the federate can join and leave during the simulation. can do. Furthermore, since it becomes possible to participate in the federation at a necessary part, it is possible to reduce the load such as management of communication data and participation federation in the middleware. Therefore, it is possible to improve the processing capability in the simulation system, and it is possible to reduce the operation time and the work time for the user.

  In other words, even during the execution of the simulation, the federation that the user wants to add or change can freely join or leave the way without ending or temporarily stopping the running simulation. Therefore, it is possible to execute a simulation that is more flexible and more realistic than existing systems. For these reasons, it is possible to provide a distributed simulation system in which a federation can freely participate in and leave a running simulation in a distributed simulation system including a plurality of federates.

  In addition, this invention is not limited to the said embodiment, In the implementation stage, it can embody by modifying a component in the range which does not deviate from the summary. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment.

1 is a system diagram showing an embodiment of a distributed simulation system according to the present invention. The schematic diagram which shows the state which RTI manages the time passage in federation. The figure which shows the outline | summary of the time management of RTI. Schematic diagram showing in more detail the federation RTI participation on the way. FIG. 5 is a schematic diagram showing in more detail the midway departure from federation RTI.

Explanation of symbols

  A to D ... federate, PC1 to PC3 ... computer, 21 ... interface unit, 22 ... display unit, 23 ... storage unit, 23a ... scenario data, 24 ... control unit, 24b ... federation application, 24a ... RTI. exe file, 24c ... notification processing unit, 24d ... deletion processing unit, 25 ... user interface unit, 100 ... network

Claims (2)

A distributed simulation system using a plurality of computers connected to each other via a network,
The computer
Storage means for storing a management table for defining the behavior of the simulation model participating in the simulation;
Simulation progress control means for autonomously judging the behavior of each simulation model based on the contents of the management table and proceeding with the simulation;
A notification means for notifying the simulation model of information for updating the contents of the management table when a simulation model requesting participation in the simulation occurs during the simulation;
A distributed type comprising: deletion means for deleting information related to the simulation model from the contents of the management table when a simulation model requesting the withdrawal from the simulation occurs during the progress of the simulation Simulation system. 2. The distributed type according to claim 1, wherein the simulation is realized under an RTI (Run-Time Infrastructure) for executing each service defined in an HLA (High Level Architecture) interface specification. Simulation system.

Images