GM优化方法在机械系统寿命预测中的应用

Application of GM( 1,1, ω * )methods in the life prediction of mechanical systems

针对机械系统失效单元寿命预测所涉及的时间序列 ,介绍了灰色模型GM(1,1) (GM :GreyModels)方法在预测中的特长及缺陷 .同时为了提高预测精度 ,将灰色模型GM(1,1)法拓广为GM(1,1,ω)预测模型法 .因为新模型中参数ω与预测误差之间存在着明显的非线性特性 ,而且ω数值离散化 ,所以采用优化逼近方法优化ω 值 ,最优GM(1,1,ω )预测精度高于GM(1,1) ,工程实例也证实了此方法的效果是显著的 .

s:Aimed at the time series involved in life prediction of the failure elements in mechanical systems, the advantages and shortcomings of GM(1,1)in prediction are introduced. Meanwhile, the grey model GM(1,1)has been converted into a new predictive model of the GM(1,1, ω )to improve prediction precision. Because of the nonlinear correlation between ω and the prediction errors, as well as the discretization of ω values, the optimal approaching method is taken in this article to optimize the parameter ω *. A conclusion can be drawn that the accuracy of GM(1,1, ω *)is much higher than that of unoptimizable GM(1,1). Some practical examples show that the effect of the method is remarkable.