Modularization Design And Research Of Load Balancing Of Networked Fire Control Simulation System

Networked fire control architecture is the product of modern warfare, which combinescomputer science, network and information technology. It is the trend of the current firecontrol architecture. Simulation is an effective way to verify the effectiveness of the firecontrol architecture and it can save quantities of resources. Meanwhile, some seriousproblems about simulation have also emerged. As the current distributed simulation systemdevelopment technology, HLA (High Level Architecture) has poor reusability and can’tsolve load imbalance. To solve these problems, the reusability and load optimal allocationof simulation system have been researched. The development method and optimizationalgorithm were designed having HLA based networked fire control simulation system as abackground. The specific contents are as follows.Firstly, a hierarchical and distributed networked fire control architecture which is morereasonable was selected on the basis of the analysis of some typical fire control systems.According to its characteristics and simulation requirements, a networked fire controlsimulation system based on HLA was designed and the functions of its federates andinteractions among them were determined. Then, the solutions that developing the federatesby component technology and balancing the system loads by multi-objective particleswarm optimization algorithm were proposed.Secondly, a component-based design idea was introduced into the developmentprocess of the simulation system. Smaller functional modules were divided. A multi-levelsimulation system architecture from the application layer to the support layer was formedby the analysis of internal information flow in each federate. According to the horizontaland vertical characteristics on modules’ reusability, three types of components includingcommon components, special components and communication components were designedby COM technology. And a component library which has good reusability andencapsulation was formed.Thirdly, some load description models were established and quantization method ofparameters in models was proposed for solving the load imbalance. Several improvementstrategies for PSO(Particle Swarm Optimization) and multi-objective algorithm wereadopted and multi-objective PSO load balancing algorithm was designed according to themodel’s characteristics. The designed algorithm was verified and satisfying load distribution schemes were obtained by MATLAB. Then, the problem that the federates orcomponents join the federation execution dynamically was analyzed and dynamic loadbalancing method without load migration was formed.Finally, the development process of networked fire control architecture simulationplatform was elaborated according to the component-based development method offederates combined with HLA specification, and the running state test of simulation systemwas carried out. Test results show that the interactions among federates meet designedrequirements and the simulation system is operating normally, which can provide a reliableverification platform for networked fire control architecture. Meanwhile, the load balancingalgorithm proposed in this paper was verified on this platform.In summary, the research of this thesis is in line with the development trend of HLAspecification and provide a new development idea for the simulation of military objects.During the development of simulation system, interfaces of functional modules wereseparated from their implementation codes. During the simulation execution, thecalculation load and communication load on simulation nodes were well configured. As aresult, the related work in this paper is significant for improving the development andexecution efficiency of simulation systems based on HLA.