摘要: 将小波变换、多重分形和R/S分析法相结合,选择db4小波函数进行小波变换,利用分解与重构算法,从分形维数、非周期循环长度、奇异性指数、多重分形谱等方面探讨合肥市近54年来气温系统的动力学分形特性和演化规律,分析合肥市气温趋势的持续性、长期记忆性、统计自相似性和多重分形性。分析结果表明,合肥市气温变化过程是一个层次分明的过程,具有多标度结构且1月、8月、11月的分形维数较高,气温不规则程度高,复杂性高。
关键词:
小波变换,
分形特征,
Hurst指数,
多重分形
Abstract: This paper analyses the dynamic characteristics and evolutionary principle of Hefei’s temperature over the past fifty-four years, it combine R/S analysis with wavelet transform, choosing db4 wavelet function, deconstruct and reconstruct the original data. It analyses the temperature based on multi-fractal analysis, discusses the stability, long-term memory, statistical self-similarity and multi-fractal of the temperature trend in Hefei from fractal dimension, non-periodic length of circulation, singular index, and multi-fractal spectrum. Analysis results show that the temperature change in Hefei whose level is clear has a multiple-scale structure, the fractal dimension of Janmary, August, November shows that the irregular degrees and the complexities of temperatures are higher than the others.
Key words:
wavelet transform,
fractal feature,
Hurst exponent,
multi-fractal
中图分类号:
毛军军;朱良燕;张学友;吴 涛;. 基于小波变换的气温趋势和分形特征分析[J]. 计算机工程, 2010, 36(2): 18-20.
MAO Jun-jun; ZHU Liang-yan; ZHANG Xue-you; WU-tao;. Analysis of Temperature Trend and Fractal Feature Based on Wavelet Transform[J]. Computer Engineering, 2010, 36(2): 18-20.