Decomposition of machining error for surfaces using complete ensemble empirical mode decomposition with adaptive noise

Abstract

ABSTRACT Complex surfaces are widely used in the industry, and their machining accuracies have gained growing attention. Aiming at the error sources of machining errors, this paper proposes the use of the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) to decompose machining errors into systematic and random errors. First, the machining error signal was processed by CEEMDAN and broken down into multiple intrinsic mode functions (IMFs). Based on the different correlation degrees between the system error signal and the machining error signal, the IMF of each intrinsic modal component was analyzed using autocorrelation and spectral analyses. By setting the threshold value, the systematic errors of each component were judged, and the systematic error signal was reconstructed to separate the original error signal into systematic and random errors. A simulation was used to verify that the proposed decomposition method could effectively obtain separated systematic error signals with high accuracy. Finally, using a machined part with a free-form surface, the decomposition of the machining errors for complex surfaces was achieved.