基于过程支持向量机的航空发动机润滑油消耗率预测  

Aeroengine Lubrication Oil Consumption Rate Prediction Based on the Process Support Vector Machine

作　　者：钟诗胜[1] 雷达[1] 

机构地区：[1]哈尔滨工业大学机电工程学院,黑龙江哈尔滨150001

出　　处：《润滑与密封》2009年第10期6-9,共4页Lubrication Engineering

基　　金：国家863计划项目(2008AA04Z401);武器装备预研基金项目(9140A17030708HT01);中国民航科技项目(MHRD200816)

摘　　要：分析了目前航空发动机润滑油消耗率监控现状及其不足,提出了一种基于过程支持向量机的润滑油消耗率预测方法,针对润滑油消耗率实际数据有较大波动性的特点,使用了样条插值增加样本点的方法来提高预测精度。通过润滑油消耗率预测实例对提出的方法进行了检验,实例结果表明过程支持向量机能够比较准确地对润滑油消耗率进行预测。After analyzing the current situation and the disadvantage of the lubrication oil consumption rate monitoring of the aero engine, a lubrication oil consumption rate prediction method based on the process support vector machine （PSVM） was presented. Aimed at the problem of the prominent fluctuation of the actual oil consumption rate data, the spline interpolation method was employed to add sample points in order to improve the prediction precision. The prediction effect was tested by a practical lubrication oil consumption rate prediction instance. According to the test result, PSVM can precisely predict the lubrication oil consumption rate.

关 键 词：润滑油消耗率 过程支持向量机 样条插值 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] V263.6[自动化与计算机技术—控制科学与工程]