欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
在总体平均经验模态分解（Ensemble Empirical Mode Decomposition,EEMD）降噪过程中,对本征模态分量（Intrinsic Mode Function,IMF）的有效处理一直是影响降噪效果的关键。为此,提出一种基于改进EEMD的去噪方法。基于＂3σ＂法则和奇异值分解（Singular Value Decomposition,SVD）提取第一个IMF分量中有用信号细节。利用连续均方误差准则对剩余IMF分量进行高低频区分,分别使用SVD和S-G算法提取高低频分量的有用信号,可以有效避免了高频部分有用信号的流失,同时剔除低频分量中的部分噪声,克服了EEMD去噪时IMFs难以有效处理的不足。为了验证该方法的有效性,进行了数字仿真与双势阱混沌振动试验,结果表明,该方法的降噪效果优于小波加权和EEMD去噪方法。
英文摘要：
In de-noising process using the ensemble empirical mode decomposition（ EEMD）,the effective treatment of intrinsic mode functions（ IMFs） is a key affecting noise reduction effect. Here,an improved EEMD denoising method was proposed. Firstly,the useful signal details of the first IMF were extracted based on the ＂3σ＂criterion and the singular value decomposition（ SVD）. Then the remaining IMFs were divided into higher frequency components and lower ones based on the consecutive mean square error（ CMSE）. Secondly,useful signals in higher frequency components and lower ones were extracted based on SVD and Savitzky-Golay（ S-G） filtering method,respectively. Thus,the loss of useful signals in higher frequency components was avoided while parts of noise in lower frequency components were removed effectively to overcome the shortcoming of IMFs being difficult to treat with EEMD to do de-noising. In order to evaluate the effectiveness of the proposed method,the test rig of leaf spring based on the double-potential well theory was made and the test results showed that the proposed method is better than the wavelet weighted parameters method and the EEMD de-noising one.
同期刊论文项目
同项目期刊论文
期刊信息
位置：