Fault Diagnosis of Rolling Bearing Based on a Novel Adaptive High-Order Local Projection Denoising Method
Table 1
The algorithm of standard LP denoising method.
(1) Select appropriate embedding dimension and delay time to reconstruct noisy time series into dimensional phase space.
(2) Determine neighborhood for one phase point.
(3) Calculate centroid of its neighborhood by fixed neighborhood numbers.
(4) Calculate covariance matrix C of its phase point neighborhood and then conduct eigenvalue decomposition to obtain eigenvector aq corresponding to smaller eigenvalues. .
(5) Subtract the projection of the phase point in noise subspace by
where R is a diagonal weight matrix, aiming at inhibiting the distortion of ending elements of phase points and reserving the middle stable elements (R11 and were set as 103; values were set as 1 in our research). is the centroid of neighborhood; denotes the denoised signal.
(6) Get back to step (2) and repeat subsequent steps until all phase points are processed.
