摘要
根据模拟系统故障检测原理,提出一种新型的基于支持向量分类机的故障检测与诊断方法。采用多传感器信息融合技术,利用故障特征进行建模,采用序列最小最优化算法求解.实验结果表明,该方法能准确、快速地判断出发生故障的元件.
According to the principle of fault detection and diagnosis,a new method of fault detection and diagnosis based on support vector machine is proposed. The way is to design a model based on data fusion of multisensor. The characteristics of faults are used to model which can be solved by Sequential Minimal Optimization. And the model has been used to diagnose an analog circuit. The result of experiment shows the new method can quickly identify the components with faults.
出处
《大庆石油学院学报》
CAS
北大核心
2007年第3期102-103,135,共3页
Journal of Daqing Petroleum Institute
关键词
支持向量机
信息融合
故障诊断
支持向量分类机
support vector machine
information fusion
fault diagnosis
support vector classification