A novel ensemble method based on the SBLMD-ANN-MOPSO approach for predicting milling stability regimes

Abstract

In recent decades, a lot of work has been done to mitigate self-excited vibration effects in milling operations. Still, a robust methodology is yet to be developed that can suggest stability bounds pertaining to higher metal removal rates (MRRs). In the present work, experimentally-acquired acoustic signals in milling operations have been computed using a modified spline-based local mean decomposition technique in order to cite tool chatter features. Further, three artificial neural network training algorithms; resilient propagation, conjugate gradient-based and Levenberg–Marquardt algorithms, and two activation functions; hyperbolic tangent sigmoid and log sigmoid, have been used to train the acquired chatter vibration and MRR data set. Over-fitting or under-fitting issues may arise from the random selection of a number of hidden neurons. The solution to these problems is also proposed in this paper. Among these training algorithms and activation functions, a suitable one has been selected and further invoked to develop prediction models of chatter severity and MRR. Finally, multi-objective particle swarm optimization has been invoked to optimize the developed prediction models to obtain the most favourable range of input parameters pertaining to stable milling with higher productivity.