作 者: ;
机构地区: 广东交通职业技术学院
出 处: 《装备制造技术》 2012年第7期247-248,250,共3页
摘 要: 为了简化船舶主机故障诊断,提高诊断效率,文章采用了支持向量机的故障诊断原理,通过小波包分解提取信号的特征参数,再将特征量送入故障分类器中进行训练,即可得出诊断结果。当数据样本较少时,采用支持向量机与采用神经网络诊断相比,具有算法简单、故障分类能力强的优点。 In order to simplify the ship host fauh diagnosis, and to improve the diagnostic efficiency, a support vector machine principle of fault diagnosis in this paper, through the wavelet packet decomposition extraction of signal characteristic parameters, the characteristic features into fault classifier in training Again , can draw a diagnosis. When data sample is small, the SVM (support vector machine) and using neural network diagnosis, it is characterized by the algorithm is simple, fault classification ability strong advantages.
领 域: [电气工程]