摘要
借助于仿真数据及某２００ＭＷ机组的实测数据，对ＧＭ（１，１）、ＡＲ（ｐ）及基于合成ＢＰ网络的ＮＡＲ（ｐ）模型的预测性能进行分析及比较，得出了适合于汽轮发电机组的故障预测模型，为火电机组的安全运行和预知状态维修做好了理论准备。
By the use of simulation data and operating data of a 200MW turbogenerator set,the GM(1,1),AR(p) and NAR(p) model based on combined BP network are analyzed and compared.Then the predictive model which fits for the faults of the turbogenerator set is obtained,provides ready for safe operation and realizing predictive maintenance of the turbogenerator set.
出处
《东北电力学院学报》
1996年第3期36-42,共7页
Journal of Northeast China Institute of Electric Power Engineering
关键词
汽轮发电机
故障
预测模型
GM(1,1) model,AR(p) model,BP neural network,prediction
