自适应学习率的 BP 网络算法及其在汽轮发电机组故障模糊诊断中的应用

BP Network Algorithm with Self-adaptive Learning Rate and Its Use in the Fuzzy Diagnosis of Turbogenerator Failures

针对ＢＰ网络的不足，提出了自适应学习率的ＢＰ网络算法，从根本上解决了ＢＰ网络学习率的取值问题和收敛速度慢的问题，并有效地克服了ＢＰ网络易陷入局部最小点的缺点，采用这种改进算法成功地实现了对汽轮发电机组故障的模糊诊断。

To overcome the shortcomings of a BP network,this paper proposes a BP algorithm with self adaptive learning rate,thus once for all overcoming the demerits of BP networkds,i.e.,slow convergency and inability to determine the value of learning rate.In addition,the disadvantage of BP network liable to fall into a local minimum point can also be effectively eliminated.Through the use of such an improved algorithm successfully realized is the fuzzy diagnosis of turbogenerator failures.