| 基于神经网络的汽车发动机建模和故障诊断 |
| |
| 引用本文： | 苏蓉,张增年.基于神经网络的汽车发动机建模和故障诊断[J].上海电机学院学报,2005,8(3):40-43. |
| |
| 作者姓名： | 苏蓉 张增年 |
| |
| 作者单位： | 1. 钱江集团环球机械电器分公司,温岭,317500
2. 浙江万里学院,电子信息学院,宁波,315100 |
| |
| 摘 要： | 采用神经网络来辨识汽车发动机的有关歧气管压力环节的非线性动态模型,为实现系统的非线性动态映射,引入了外部回归项,然后用动态Levenberg-marquardt算法来对动态模型进行参数估计。仿真结果表明,基于神经网络的模型具有较高的精度和较强的通用性。在此基础上,用动态增益矩阵法的故障诊断方法和神经网络实现汽车发动机的在线故障诊断。
|
| 关 键 词： | 神经网络 汽车发动机 非线性系统 模型辨识 故障诊断 |
| 文章编号： | 1671-2730(2005)03-0040-03 |
| 修稿时间： | 2005年4月13日 |
Automobile Engine Modeling and Malfunction Diagnosis Based on Neural Network |
| |
| Su Rong,Zhang Zengnian.Automobile Engine Modeling and Malfunction Diagnosis Based on Neural Network[J].JOurnal of Shanghai Dianji University,2005,8(3):40-43. |
| |
| Authors: | Su Rong Zhang Zengnian |
| |
| Affiliation: | Su Rong~1,Zhang Zengnian~2 |
| |
| Abstract: | Neural network is adopted to distinguish non-linear dynamic model of fork windpipe pressure link of automobile engine. In order to realize non-linear dynamic mapping of the system, the external regression terms are introduced, and then the dynamic Levenberg-marquardt algorithm is used to estimate parameters of the dynamic model. The simulation shows that the model based on neural network has higher accuracy and wider general purpose. Based on this, the failure diagnosis, using dynamic gain matrix method and neural Network, is used to realize on-line failure diagnosis of automobile engine. |
| |
| Keywords: | neural network automobile engine non-linear system model identifiction fault diagnosis |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
