| 摘要: |
| 本文以一个典型的泵压式液体火箭发动机(LRE)为对象,考虑发动机的几种主要故障,建立于描述发动机静态工作过程的非线性方程组。用Hopfield神经网络模拟这些非线性方程组,用ART2网络识别故障模式。对该发动机,特别是涡轮泵故障的识别方法进行了研究。 |
| 关键词: 液体推进剂火箭发动机 非线性系统 神经元件 网络分析 故障识别 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金资助项目 |
|
| FAULT DIAGNOSIS OF A LIQUID ROCKET ENGINE BASED ON NEURAL NETWORKS |
|
Huang Minchao, Feng Xin, Zhang Yulin
|
|
National University of Defence Technology
|
| Abstract: |
| A method of fault diagnosis of rockets, especially of turbine pump rocket is studied in this paper. Accounting for several main faults of rockets, nonlinear equations describing rocket still process are set up. The Hopfield neural networks are used for simulating nonlinear equations and ART2 networks for diagnosing the fault patterns. |
| Key words: Liquid propellant rocket engine Nonlinear system Neural component Networks analysis Fault diagnosis |
