基于自组织特征映射神经网络的汽轮机回热系统故障诊断

Fault diagnosis of steam turbine regenerative system based on self-organizing neural network

提出应用自组织特征映射神经网络对汽轮机回热系统进行故障诊断,并利用MATLAB神经网络工具箱来实现.介绍了在实现过程中故障样本的提取、征兆参数的处理以及离线利用MATLAB进行网络训练等相关技术.实例仿真表明,该方法是一种可行有效的回热系统故障诊断方法.

A fault diagnosis method of steam turbine regenerative system is proposed. It is based on self\|organizing neural network and implemented by the neural network toolbox in MATLAB. Some correlative techniques, such as the selection of samples, the treatment of symptom parameters and off\|line network training by MATLAB, are expounded. The simulation results indicate that the method is effective and feasible.