| 摘要: |
| 针对传统意义的BP（Back-Propagated）神经网络在水轮机故障诊断中的不足，提出了一种基于径向基RBF（Radial basis function）神经网络的水轮机组故障诊断方法。实例应用表明，该方法克服了BP神经网络的不足，具有精度高、收敛快、可以避免局部极小值的优点；RBF神经网络收敛速度约是BP神经网络的40倍，并能准确地诊断出水轮机组的故障。 |
| 关键词: 水轮机组 故障诊断 神经网络 径向基 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| Research on fault diagnosis based on RBFNN for hydropower units |
|
|
| Abstract: |
| For the system of faults diagnosis of hydropower sets,the deficiency of faults diagnosis using BP Neural Network is analyzed and a RBF Neural Network algorithm is presented,which has advantage of high precision,avoiding local minima and fast convergence rate.In real diagnosis system convergence rate of RBFNN is nearly 40 times faster than BPNN,and it can diagnose the faults of hydropower sets exactly. |
| Key words: hydropower unit fault diagnosis Neural Network RBF |
