论文部分内容阅读
在递归损失量和递归趋势两个非平稳测度的基础上,结合主成分分析方法,提出一套实用的振动信号非平稳评价方法.首先,在一定观测时间尺度的基础上,分别建立由递归损失量和递归趋势两个测度组成的样本向量;依据工作平稳性的定义,将两个向量组装成多变量的样本矩阵,对其进行主成分分析,取第一主成分作为振动信号非平稳性综合评价的对象,构造非平稳性的统计模式差异指标,并通过几种典型的数值模拟非平稳信号的评价结果验证指标的可靠性.最后,将此方法应用于桥梁结构环境振动信号分析,以得到更加客观的分析结果.
Based on two non-stationary measures of recursive loss and recursive trend, a set of practical non-stationary vibration signal evaluation methods are proposed based on principal component analysis.Firstly, on the basis of observing time scales, Quantity and recursive trend. According to the definition of work stability, the two vectors are assembled into a multivariate sample matrix, and the principal components are analyzed. The first principal component is taken as the vibration signal non-stationary composite The object of evaluation is to construct the statistical index of non-stationary statistical model, and verify the reliability of the index through several typical numerical simulation results of non-stationary signals.Finally, this method is applied to the analysis of environmental vibration signals of bridge structure to obtain More objective analysis of the results.