基于流形的蒸汽轮机系统异常检测技术研究

Abnormal Detection Method Based on Manifold Learning for Steam Turbine System

随着应用于工业过程的蒸汽轮机系统的发展,系统中的功能关系和非线性特性也愈发明显,因此考虑这些变化趋势的故障检测方法具有越来越重要的意义。文章从蒸汽轮机系统中的形状特征的角度对系统数据中的复杂关系进行解读和表征,提出了一种利用流形学习方法和流形空间统计分析方法实现的系统级早期异常检测方法,弥补了传统故障检测方法的不足。设计了相关实验验证了该方法的有效性和优越性。

With the progressing of steam turbine system for industrial sake,the functional interrelationship and nonlinear characteristics in the system is becoming more and more evident.In such case,the fault detection method considering such trend is becoming more and more significant.A system-level fault detection method is proposed utilizing manifold learning and statistical analysis method in manifold space,which is based on the understanding of the complex relationship in the system data from aspect of the shape pattern in the steam turbine system.The shortcomings of traditional fault detection methods are made up and experiments are designed to verify the effectiveness and advantages of the method.