Noise Reduction Algorithm Using Principal Component Analysis

摘要

The present invention is to improve the accuracy of the noise data by removing the test data obtained through the test when evaluating the vibration noise of the vehicle less than the reference value among the main components calculated through the principal component analysis, as shown in Figure 1 In order to realize a noise reduction algorithm, calculating a square matrix for eigenvalue analysis from a matrix to remove noise transmitted from a transfer function of each boundary, and performing an eigenvalue analysis from the calculated square matrix, An eigenvalue analysis step for obtaining an eigenvalue vector matrix, a transform matrix and a principal component decomposition step for calculating a transform matrix and a principal component matrix from the eigenvalue analysis step, and a principal component matrix obtained through the transformation matrix and the principal component decomposition step in a size order Principal components smaller than value are considered perturbation and removed One will be able to remove as high a by removal of a new noise reduction type separated by a step of calculating a matrix illustrating the uncorrelated components included in the data Figure 2a.