Piecewise linear representation of time series based on mean trend in sliding window

摘要

Seismic data show some important characteristics,such as big volume and strong timeliness.Specific to the time series data of earthquake precursory observations,a piecewise linear representation based on the sliding window mean value(PLR MTSW)algorithm is proposed.With this algorithm,the mutation points can be identified accurately according to the rate of mean value change,while the main features of time series are maintained well.This algorithm can also smooth the noise and improve the compression accuracy with sliding window.Meanwhile the local extreme points can be identified effectively according to the change of mean value trend within window.