基于免疫多Agent的省级气象信息网络监控系统建模

MODELLING MONITORING SYSTEM FOR PROVINCIAL METEOROLOGICAL INFORMATION NETWORK BASED ON IMMUNE MULTI-AGENT

针对传统气象信息网络监控系统故障检测速度慢、误报率高、自适应差等问题,结合人工免疫原理和多Agent技术,构建省级气象信息网络监控系统模型。该模型融合了两者的优势,能同时进行多层次的监测和不同级别的响应。在模型设计过程中,充分考虑了生物免疫特性的算法和模型,将免疫Agent做为构建Multi-Agent系统的模型基础,增强了整个气象信息网络监控系统的新抗体生成能力、记忆性及耐受性。实验仿真证明这种新型的网络监控模型能实现较好的检测率和较低的误测率,为建立网络监控系统提供了理论模型。

In view of the problems in traditional meteorological information network monitoring system including slow in fault detection, high in false alarm rate and poor in self-adaption, etc. , we build the model of provincial meteorological information network monitoring system in combination with artificial immune and multi-Agent technology. The model integrates the advantages of the both,it can carry out multi-level monitoring and response at different levels simultaneously. In the design process the algorithm and model of biological immunity characteristics has been taken into full account. The immune Agent is employed as the model basis to build the multi-Agent system, this enhances the new antibody generation capability, memory property and tolerance of the whole system. Experimental simulation proves that the new network monitoring model has good detection rate and lower false detection rate, it provides a theoretical model for the establishment of network monitoring system.