| L-M神经网络算法在汽轮发电机组故障诊断中的应用 |
| |
| 引用本文： | 程伟,李华.L-M神经网络算法在汽轮发电机组故障诊断中的应用[J].兰州铁道学院学报,2007,26(3):137-140. |
| |
| 作者姓名： | 程伟 李华 |
| |
| 作者单位： | 兰州交通大学自动化与电气工程学院 甘肃兰州730070 |
| |
| 基金项目： | 甘肃省自然科学基金，教育部重点实验室基金 |
| |
| 摘 要： | 针对汽轮发电机组的故障诊断,采用Levenberg-Marquardt算法建立多层前向人工神经网络,采用改进算法训练网络,克服了传统BP算法收敛速度慢,易陷入局部最小的缺陷.就BP网络的不足,提出了一种改进的BP神经网络模型,并使用L-M算法用于汽轮发电机组故障的诊断.经理论和实践证明:该方法有效地提高了故障诊断的精度和可靠度,为旋转机械故障诊断提供了有效方法.
|
| 关 键 词： | 神经网络 L-M算法 汽轮发电机 故障诊断 |
| 文章编号： | 1001-4373（2007）03-0137-04 |
| 修稿时间： | 2006-09-30 |
Application of Artificial Neural Network Based on Levenberg-Marquardt Algorithm to Fault Diagnosis of Turbine Generator Set |
| |
| Cheng Wei,Li Hua.Application of Artificial Neural Network Based on Levenberg-Marquardt Algorithm to Fault Diagnosis of Turbine Generator Set[J].Journal of Lanzhou Railway University,2007,26(3):137-140. |
| |
| Authors: | Cheng Wei Li Hua |
| |
| Institution: | School of Automation and Electrical Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China |
| |
| Abstract: | An artificial neural network(ANN) using Levenberg-Marquardt algorithm for network training is presented to diagnose faults in turbine generator set equipment.The ANN uses improved training method to avoid the drawbacks of BP algorithm,and diagnoses faults by L-M algorithm.Proved theoretically and practically,this method increases efficiently the precision and reliability of breakdown diagnosis,and provides efficient way for breakdown diagnosis of whirling machinery. |
| |
| Keywords: | neural network Levenberg-Marquardt algorithm turbine generator breakdown diagnosis |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
