基于短时分形维数的模糊控制滤波

Fuzzy Control Filtering Based on Short-Time Fractal Dimension

中文关键词  模糊控制  分形  混沌  短时分形维数

英文关键词  fuzzy control  fractal  chaos  short-time fractal dimension

中文摘要

      维数是系统的拓扑不变量，从理论上说，Hausdorff维数是最为重要的维数，但实际应用中几乎无法计算，而分形维数的计算机实现是很方便的. 本文基于分形理论中的分形维数概念，给出了复杂时闻序列的短时分形维数概念，提出了一种新的滤波方法即基于短时分形维数的模糊控制滤波方法，仿真结果表明，该方法滤波效果良好，可应用于复杂信号的滤波.

英文摘要

      Along with the development of computer science and technology， the researches in chaos and fractal theories and applications cause public attentions and many important results are obtained. It is well known that fractal dimension is t. he important dimension but difficult to compute. In order to overcome this defect， the short-time fractal dimension， which can exactly depict the local features of complex signals， is giv-en. Based on this concept， a new fuzzy control signal filtering method (SFDF) is put forward. The advantage of SFDF is that the fuzzy control parameter can be dynamically revised according to the complex signal's fea-ture. We apply SFDF to voice signal filtering and the simulation results show that the filtering effect is satis-factory.