基于BP神经网络的燃气轮机冷热部件故障诊断  

Fault Diagnosis of Cooling and Thermal Parts of Gas Turbine based on BP Neural Network

作　　者：姚杰[1,2] 李红伟[1] 马欣[3] 陈贵[2] 

机构地区：[1]西南石油大学研究生院,成都610500 [2]西南石油大学电气信息学院,成都610500 [3]西南石油大学机电工程学院,成都610500

出　　处：《燃气轮机技术》2012年第4期39-43,共5页Gas Turbine Technology

摘　　要：为解决燃气轮机故障诊断中可测参数难以直接反映机组故障状态的问题,在分析燃气轮机冷热部件组成和故障判据的基础上,提出利用小偏差方程建立并求解燃气轮机故障数学模型,并结合故障判据寻找可测参数与性能参数的关系,确立从可测参数到性能参数,再到故障原因的诊断思路。另外,将BP神经网络作为故障诊断工具,通过向网络中输入8个机组可测参数来实现智能故障诊断,为燃气轮机的快速精确诊断提供了可参考的思路和方法。In order to solve the problem that the measurable parameters is difficult to reflect the state of unit fault, this paper presents a diagnostic thinking based on analyzing the gas turbine components and faults criterion for gas turbine diagnosis, it contains using small deviation equation for establishment and solving of the gas turbine faults mathematical model, using fimlts criterion for search of the relation of measurable parameters and performance parameters, and locating the fault causes through this relation. In addition, this paper takes the BP neural network as a troubleshooting tool, it realizes the gas turbine intelligent faults diagnosis by inputting 8 unit measurable parameters into network and provides some ideas and methods for gas turbine fast real-time diagnosis.

关 键 词：BP算法 神经网络 燃气轮机 冷热部件故障诊断 

分 类 号：TK478[动力工程及工程热物理—动力机械及工程]