Distributed multi-agent system-based combat simulation engine

Abstract

The invention provides a distributed multi-agent system-based combat simulation engine. According to the combat simulation engine, a hierarchical blackboard structure is constructed on the basis of a blackboard mechanism; and by applying a distributed parallel computing theory, the reasonable distribution and full utilization of the computing resources of simulation nodes in the distributed simulation are realized, and the multi-process parallel computing management and scheduling of a large number of intelligent agent models in the simulation nodes are realized. The combat simulation engine comprises a universal blackboard frame, a blackboard manager, a serial network communication component, a blackboard browser, a recording / playback component, a CGF behavior modeling frame, a graphical CGF behavior modeling tool, an equipment performance parameter library component and a behavior knowledge base.

Description

technical field [0001] The invention belongs to the field of system simulation, and further relates to the field of combat simulation based on a distributed multi-Agent system. The present invention is mainly used for military simulation training. Developers can construct various equipment physical models, behavior models, etc., and if necessary, they can connect to the human-in-the-loop simulation training system to simulate and generate realistic battlefields, so as to achieve tactical deduction, combat legal research purposes. Background technique [0002] The agent-based simulation structure simulates the thinking of human beings to solve problems, and decomposes complex problems into multiple easy-to-solve sub-problems, and the solution to each sub-problem is the input of another sub-problem solution. Agent technology is considered to be the inheritance of object-oriented pattern. [0003] Multi-Agent Systems (Multi-Agent Systems, MAS) refers to a system in which mult...

AI Technical Summary

Problems solved by technology

Most commercial software on the market adopts the modeling method of single-process sequential execution, which can be used when the number of simulation entities is small, the model is simple, and the real-time requirements are not high. When the number of entities increases, the model complexity increases, and the real-time requirements are high. , the operating efficiency of the simulation platform is reduced, and real-time simulation cannot be performed

Images