舰用发动机某关键部件故障预测方法与应用研究

发布时间：2018-05-29 07:38

本文选题：舰用发动机轴承 + 振动信号　；参考：《沈阳航空航天大学》2016年硕士论文

【摘要】：舰用发动机是军舰上的复杂动力设备,集机械、液压和电气等多种系统于一体,与普通的航空发动机相比,其工作环境较恶劣,受海洋盐雾和湿热影响较大,易于发生故障。因此,研究舰用发动机故障预测技术,可以对舰用发动机早期故障进行预报,保障舰用发动机安全运行,对保障军舰的战斗力具有重大意义。本文以某型舰用发动机关键部件——轴承作为具体研究对象,对其开展故障预测方法与应用研究。首先,对舰用发动机轴承的故障类型和故障状态演化特点进行了深入分析;其次,采用振动检测法,在某型真实舰用发动机试验平台上,经长期试验,获取轴承振动加速度数据,并对原始数据利用CEEMD方法进行降噪处理,减少原始信号中的干扰信息;再次,针对降噪后信号特点,为获取能够表征轴承健康状态的参数,提出了基于小波包分解的特征提取方法,提取信号的小波包能量特征,作为预测舰用发动机轴承健康状态的特征信息;最后,采用小波包能量信息进行建模,分别创建了基于灰色理论的故障预测模型、基于灰色神经网络的故障预测模型和基于RVM的故障预测模型,采用所创建的模型对舰用发动机轴承的健康状态进行了预测研究,并对这几种模型的预测效果进行了分析对比。研究表明,基于RVM的故障预测方法在舰用发动机轴承故障预测中效果最佳,具有很好的应用前景。
[Abstract]:Ship engine is a complex power equipment on warship, which integrates mechanical, hydraulic and electrical systems. Compared with ordinary aero-engine, its working environment is worse, which is affected by sea salt fog and moisture and heat, and is prone to malfunction. Therefore, the study of ship engine fault prediction technology can predict the early fault of ship engine, ensure the safe operation of ship engine, and have great significance to guarantee the warship combat effectiveness. In this paper, bearing, a key component of a ship engine, is taken as a specific research object, and its fault prediction method and application are studied. Firstly, the fault types and fault state evolution characteristics of marine engine bearings are deeply analyzed. Secondly, the vibration detection method is used to test the ship engine on a real ship engine test platform for a long time. Obtain the vibration acceleration data of bearing, and use the CEEMD method to reduce the noise in the original signal. Thirdly, according to the characteristics of the signal after noise reduction, to obtain the parameters that can represent the healthy state of the bearing. A feature extraction method based on wavelet packet decomposition is proposed to extract the wavelet packet energy feature of the signal, which can be used as the feature information to predict the health state of the marine engine bearing. Finally, the wavelet packet energy information is used to model the model. The fault prediction model based on grey theory, the fault prediction model based on grey neural network and the fault prediction model based on RVM are established respectively. The prediction results of these models are analyzed and compared. The research shows that the fault prediction method based on RVM has the best effect in the fault prediction of marine engine bearings and has a good application prospect.
【学位授予单位】：沈阳航空航天大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：V263.6

【相似文献】