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：
针对局部投影去噪方法的邻域选取问题,分析了选取邻域点的准确率对去噪结果的影响.结合奇异谱分析技术,提出了一种改进的邻域选取方法.该方法对含噪声的相空间进行奇异值分解,利用较大奇异值对应的主分量重构相空间,在重构后的空间中寻找邻域点,以提高选取的邻域点准确率.用改进后的方法对含噪声的Lorenz序列及太阳黑子月观测值序列进行仿真,仿真结果表明该方法能够有效地提高选取的邻域点的准确率,进而改善局部投影方法的去噪效果.
英文摘要：
The accuracy of the selected neighbors for the local projection noise reduction is analysed. An improved neighborhood selection method combing with the SSA （ singular spectrum analysis） technique is proposed. The singular value decomposition is performed for the noise corrupted phase space, and several larger principal components are chosen to reconstruct a less noisy phase space. By searching the reconstructed phase space, the accuracy of the selected neighbors can be elevated. Simulations are done with the monthly sunspots values and the time series generated by the Lorenz equations, and the results demonstrate that the improved method can elevate the accuracy of the neighborhood selection and reduce the noise effectively.
同期刊论文项目
同项目期刊论文
The hidden neurons selection of the wavelet networks using support vector machines and ridge regress
期刊信息
位置：