小波时间序列对非平稳信号中突变点的辨识与处理

Identification and processing of wavelet time series models to mutation point in non-stationary signal

本文利用小波时间序列模型找出非平稳信号中突变的位置,通过分析把突变点分为两类;若对两类突变点作适当的插值处理,既可保留原始数据发展趋势的真实性也可提高预测模型的精度。文中把迭代模型的预测值去逼近突变点值和单纯地用Pchip插值、spline插值、linear插值作为内插值进行了对比研究,得出了各自的优缺点。

This paper used wavelet time series models to identify the location of the mutation point in non-stationary signal, and the mutation points were divided into two categories through analysis. Two types of mutation points were processed by appropriate inter- polation approach to retain the authenticity of the raw data and improve the prediction accuracy of the model. The paper compared the predictive value of the iterative model to approximate the mutation point value with the simple uses of Pchip, spline and linear interpo- lation as the interpolations, and finally obtained the advantages and disadvantages respectively.