Equiprobable symbolization pattern entropy for time series complexity measurement

Abstract In order to effectively mine the structural features in time series and simplify complexity of analysis, equiprobable symbolization pattern entropy (EPSPE) is proposed this paper. The original are implemented through symbolic processing according an equal probability distribution. Then, sliding window technique used obtain a finite number different patterns, pairs determined by calculating conversion between patterns. Next, frequency symbolized patterns counted calculate pairs, thus estimating measurement complex signals. Finally, we conduct extensive experiments based on Logistic system under parameters natural wind field. experimental results show our EPSPE increases from 5 7.5 as increase, which makes distinction periodic with varying degrees intuitive. Meanwhile, it can more concisely reflect characteristics interrelationships field (8.8–10 for outdoor 7.8–8.3 indoor). contrast, several state-of-the-art schemes irregular cannot distinguish well accurately predict spatial deployment relationship nine 2D ultrasonic anemometers.