作　　者： ; ; ; ;

机构地区： 浙江大学机械工程系

出　　处： 《机床与液压》 2003年第1期247-249,共3页

摘　　要： 提出了一种AR模型的盲辨识的算法,该算法与传统AR模型的辨识算法的区别在于:它可同时确定AR模型的阶次 和参数,而不需事先确定AR模型的阶次或者假定AR模型的阶次已知,其特点是计算量小,具有很好的收敛性,并在此基础 上,将AR模型的盲辨识方法应用到机械故障诊断中,进行了实验研究,取得了很好的效果。 Blind recursive identification algorithm of autoregressive models was presented, and the algorithm could determine system model order and parameters. Compared with classic identification methods of AR models, the distinctness was that the blind recursive identification method didn' t need to be prior to determine, or had known its order. The distinguish feature of the algorithm is to desire a limited calculation quantity and good convergence, and the meuhod was applied to feature extraction of fault diagnosis. Experiment result shows that this method is very efficiency.

关 键 词： 机械 盲系统辨识 故障诊断 模型

领　　域： [机械工程]