摘要
多维可控自回归（ＣＡＲＶ）模型对模态参数的识别精度很高，但受噪声影响却很严重，现有的方法很少考虑有噪声输入的情况而使ＣＡＲＶ方法的使用受到极大限制。本文讨论了有噪声存在的情况下，ＣＡＲＶ模型对模态参数的识别，推导了ＣＡＲＶ模型在有噪声情况下的建模公式，并通过仿真验证了方法的有效性，针对ＣＡＲＶ模型在具体应用中的一些问题进行了讨论。
The precision of control average regression vector (CARV) model for parameter identification is quite high, but it is affected seriously by noise. The input noise condition are rarely considered by the existed methods, therefore the CARV method is mostly restricted in its applications. The parameter identification with CARV model is mainly discussed in this paper, and the modeling formula of CARV model in those noisy conditions are deduced. Simulations demonstrated the validity of the method. Some problems of CARV model in practical application are discussed in the paper.
出处
《振动工程学报》
EI
CSCD
1998年第4期501-506,共6页
Journal of Vibration Engineering
基金
国家自然科学基金
关键词
时间序列分析
参数识别
噪声
模态参数
CARV
autoregressive model
time series analysis
parameter identification
noise
