基于经验模态分解的并列双圆柱绕流信号处理  

Data Disposing in Flow Around Parallel Circular Cylinders Based on EMD

作　　者：黄晖[1] 杨松[2] 印荣[3] 麻伟巍[1] 谢锡麟[4] 

机构地区：[1]东华大学理学院,上海201620 [2]东华大学信息科学与技术学院,上海201620 [3]东华大学机械工程学院,上海201620 [4]复旦大学力学与工程科学系,上海200433

出　　处：《实验室研究与探索》2012年第6期5-9,共5页Research and Exploration In Laboratory

基　　金：国家自然科学基金项目(10872051)

摘　　要：基于磨光算子处理后空间N-S方程解的存在唯一性,针对一定雷诺数下并列双圆柱绕流具有附着于原大尺度结构上的小尺度结构的特性,提出利用经验模态分解(EMD)这一非线性、非平稳信号处理方法,对风洞实验中一定雷诺数下并列双圆柱绕流实测所得的原始信号进行多尺度分解,得到了去除骑行波之后的有效信号。实验结果表明,该方法磨光效果好,磨光后的信号标准偏差减小25%且无有效信息的损失,同时这种方法还具有自适应等特点。In the certain Reynolds number,flow around parallel circular cylinders has a structure where small-scale vortexes attach to large-scale vortexes.By disposing this typical non-linear,non-stationary signal by empirical mode decomposition（EMD）,the effective signal can be obtained.The method is based on the solution of the N-S equation existence and uniqueness after mollifier.The experimental results show that,this method not only has a good polishing effect,but also has adaptive characteristics.Besides the standard deviation of the new signal is reduced by 25% without loss.

关 键 词：经验模态分解 双圆柱绕流 磨光算子 内模函数 

分 类 号：O357.1[理学—流体力学]