摘要
在导航设备智能诊断系统中,提出了一种基于案例(CBR)故障诊断的技术解决方案,采用循环匹配法和相似度偏差补偿法改进了案例匹配计算,克服了案例匹配中对用户征兆向量的分量次序敏感的问题,在用户征兆向量与案例征兆向量维数不相等时对相似度的偏差进行了有效补偿。实验数据表明,该案例匹配方法切实可行,能有效提高匹配成功率,从而提高设备故障诊断效率。
A kind of techno- solving scheme based on CBR（Cases Based Reasoning）in Navigation Devices Intelligent Diagnosis System is advanced. The scheme which uses ways of cycle matching and of similarity-error-compensation improves the computation of case matching and overcomes the problem of being sensitive to the order of the sub-vector of the user symptom vector in the case- matching. The error of similarity is effectively compensated when the dimensions of the user symptom vector and the case symptom vector are not equal.The experiment data indicate that the way of case matching, which can improve probability of success matching , is feasible, and good fault-diagnose efficiency will be achieved.
出处
《微计算机信息》
北大核心
2007年第03S期216-218,共3页
Control & Automation
关键词
故障诊断
案例推理
案例匹配
相似度
fault-diagnosis,ase reasoning,case matching,similarity
