Simulation Modeling Research Based On Agent In Maritime Aerostat Search

Maritime search and rescue as an important component of modern maritime security planning are concerned about by more and more people. How to carry out maritime rescue simulation training to improve the ability and level of maritime search scientific and effectively is the main task of maritime security planning.System simulation technology can not only make analysis and evaluation on the existing system, as well as the feasibility of the new system and make a correct evaluation of economic efficiency. It cost small, high efficiency and can function in analyzing and comparing the different policy programmes according to established objective function. Choose the optimal choice from the programmes and support optimal management decisions. System simulation now is toward intelligent, realistic direction, it is to be an intelligent simulation, and the agent technology is the hot of intelligent computer field currently. Their own initiative and the autonomy of reaction make them the important issues in the field of intelligent simulation such as artificial intelligence.This paper used system simulation technology, combined with agent technology in intelligent modeling and simulation training in search and rescue at sea as the backdrop, carried out simulation modeling research based on agent in maritime aerostat search. The system includes three subsystems, search and rescue center, on-site command and aircraft in distress, used discrete-system simulation and simulation clock. The system focused on the collaboration and echo control between the subsystems, the produce of stochastic model and characteristics of realistic reappearance. In the end, achieved the search system prototype based on the agent through .NET platform.This paper focus on the achieved of search and rescue center subsystem, the research emphases as follows: (1) Proposed BAT (Basic Agent Template) architecture and detailed design of the BAT structure, the knowledge representation and design of the control process. Research on its instruction identification, the relevant knowledge database, design of the rules and intelligent search algorithm, then applied to the realization of the ship agent; (2) Research on the management of distributed simulation clock, formulate corresponding promotion mechanisms and the simulation clock synchronization issues between subsystem, as well as socket communication mechanism and solving the synchronization of distributed simulation; (3) Establish witnesses information randomly generated model, stochastic simulate the witnesses appearing time and the deviation alarm information. In this paper, the agent simulation modeling study not only has practical value, the results of research on pivotal technical problems of the system has the theoretical depth and versatility; Distributed simulation mode of management of the clock resolve a number of coordination and cooperation adjustment to guarantee synchronization system plays an important role.