摘要
提出一种基于粗糙集和径向基函数思想的网络层故障检测算法—RRBFNN.该方法具有简化样本、适应性强、容错性高等特点,能有效处理网络层故障诊断中噪声和不相容的信息。由于检测问题的实质是一种映射,该方法用一种前馈型网络来逼近这种映射关系,实现对故障的有效分类。同时,RRBFNN结构可以随着网络层中各种服务和应用的变化而构造。仿真表明,利用该方法实现的系统与同类的其他方法相比,提高了检测准确率和诊断速度。
Based on rough sets and radial basis function, a RRBFNN algorithm was proposed for the design of network fault diagnosis system. Reduced information table is obtained implying that the number of evaluation criteria is reduced with no information loss through the rough set approach. This reduced information is used to develop classification rules and train neural network to infer appropriate parameters. The rules developed by RA-Neural network analysis show the best prediction accuracy if a case does match any of the rules. It is capable of overcoming several shortcomings in the existing diagnosis systems, such as a dilemma between stability and redundancy, and provides the functions such as gathering of data, analysis, storing and response. Since the essence of fault diagnosis is a kind of mapping , an artificial neural network model is adopted to deal with the mapping relations, categorizing the network faults. The experimental system implemented by this method shows fine diagnostic ability.
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2006年第3期422-427,共6页
Acta Armamentarii
基金
国家自然科学基金资助项目(60273035)
关键词
计算机系统结构
粗糙集
径向基函数
故障诊断
computer system architecture
rough set
radial basis function
fault diagnosis
