摘要
支撑向量机因其适用于多参数、小样本的非线形问题,是一种极具潜力的航空发动机故障诊断方法。在分析基于支撑向量机的航空发动机故障诊断研究进展的基础上,对若干需要进一步研究的问题,如核函数的参数选择、核函数的选择、多类故障的分类策略以及SVM算法的优化进行探讨。
Support vector machines are suitable for multiparameter, high dimensional, nonlinear problems, with limited samples. This paper presents the research progress in the Aero-Engine fault diagnosis based on the support vector machine analysis, such as parameters of the kernel function, selection of the kernel function model, classification strategy of multi-class faults and the SVM algorithm optimization.
出处
《自动化技术与应用》
2009年第3期91-92,97,共3页
Techniques of Automation and Applications
关键词
支撑向量机
航空发动机
故障诊断
Support Vector Machine
aero-engine
fault diagnosis
