Intelligent Processing Research for Target Fusion Recognition System Based on Multi-Agents

According to the target fusion recognition under dynamic and uncertain target information conditions, a distributed target fusion recognition model is presented based on multi-agents. As a distributed multi-agents target recognition system, the model includes several widely distributed agent workgroups. The model is mainly made up of the detection agent, the recognition agent and the decision-making agent. The detection agent mainly realizes characteristics matching and sensor conditions controlling, the recognition agent is composed of information fusion algorithms into the nesting agents, the decision-making agent carries out various decision-making and cooperation mechanism. Then, the intelligent fusion processing under dynamic and uncertain environment is realized through communication and cooperation of the three agents for the actual condition request. It combines MAS and information fusion technology in target recognition system, and incarnates the advantages such as flexibility, reliability. Finally, by analyzing a ballistic target recognition experiment, the result is shown the model efficiency is advanced by the means.