改进背景值的非等间距阶跃灰色模型及其在油液光谱分析中的应用  

Improvement of Background Value in Non-equidistant Step Grey Model and Its Application on Spectral Analysis of Hydraulic Oil

作　　者：王海涛[1] 龚烈航[1] 王林[2] 姜玉颖[1] 

机构地区：[1]解放军理工大学工程兵工程学院江苏南京210007 [2]中国人民解放军74126部队60分队河南焦作454000

出　　处：《润滑与密封》2012年第12期92-95,共4页Lubrication Engineering

基　　金：国家自然科学基金项目(51175511)

摘　　要：针对非等间距、具有阶跃跳变的原始数据序列建立Simpson求积公式构造背景值的非等间距阶跃灰色模型。将该模型用于某型军用挖掘机液压系统油液光谱分析数据的拟合与预测,得到较高的拟合预测精度。利用阶跃参数对预测数据进行修正,根据修正数据结合线性回归理论制定的阈值,可以有效判断液压系统的运行状态,在一定程度上消除了换油对光谱分析数据的影响。Non-equidistant jump step grey model was built, which structured the background value by Simpson formula for original data sequence with non-equidistant and step jump changed. The model was used to fit and predict spectrometric data of a military excavator hydraulic system, and a high fitting and prediction precision was obtained. Through correcting the prediction data using step parameters, using the revised data and the threshold formulated by linear regression theory, the operating status of the hydraulic system can be determined, and the impact of oil change on the spectral analysis data can be eliminated to some extent.

关 键 词：灰色模型 非等间距 背景值 光谱分析 

分 类 号：TH117.1[机械工程—机械设计及理论]