基于概率神经网络的柴油机故障诊断与预测研究

基金项目:

贵州省科技基金项目（黔科合J字LKS［2011］5号）

详细信息

Research of fault diagnosis and prediction for diesel engine based on probabilistic neural network

1.
College of Mechanical and Electronic Engineering, Guizhou Normal University, Guiyang 550059, China
2.
Guiyang Aluminum and Magnesium Design Institute Co., Ltd., Guiyang 550081, China

摘要: 针对柴油机故障诊断、预测难的问题，分析了柴油机常见故障及影响因素, 介绍了柴油机故障数据的提取、分析和处理方法，建立了一种基于概率神经网络的故障诊断与预测模型。仿真结果表明，该模型能够有效地对柴油机等复杂机械系统故障进行诊断和预测，可以快速准确地给出诊断结果，其故障诊断和预测准确率达到94.84%。
- 柴油机 /
- 故障诊断 /
- 预测 /
- 概率神经网络 /
- PN
Abstract: In view of problem of difficult fault diagnosis and prediction for diesel engine, the paper analyzed common faults and influencing factors of diesel engine, introduced method of extraction, analysis and processing of the failure data, and built a model of fault diagnosis and prediction based on probabilistic neural network. The simulation results show that the model can diagnose and forecast fault of complex mechanical system such as diesel engine and give diagnosis results accurately, and its accuracy rate of fault diagnosis and prediction is up to 94.84 percent.
- diesel engine /
- fault diagnosis /
- prediction /
- probabilistic neural network /
- PN