基于小波和分形的时间序列压缩

Time Series Compression Based on Wavelet and IFS

引入离散小波变换的金字塔模型，将小波与分形结合起来，提出一种新的算法用于时间序列压缩，在信噪比保持不变的条件下，其压缩比为单纯的ＩＦＳ建模方法所获得的压缩比的２倍多．给出了ＩＦＳ以及ＩＦＳ与小波结合的算法。

Iterated Function System can be used to a kind of compression method for time series. But when compressed time series is complex. IFS modeling method can not obtain high signal to noise ratio. It is implied that pyramid model of discrete wavelet transformated and gave a new algorithm to compress time series based on connection it with IFS. This Compression algorithm can obtain three times compression ratio as IFS modeling method with signal to noise ratio unchanged. IFS algorithm and DWT&IFS algorithm are presented. Their experimental results of compression and comparison are also given.