Authors: Yifan Dai Shanyong Chen Nianhui Kang Shengyi Li
Publish Date: 2009/11/25
Volume: 49, Issue: 5-8, Pages: 635-641
Abstract
Corrective machining is fundamental to obtain higher precision than the machine tool based on error inspection and calculation techniques Generally the error to be corrected is calculated by minimizing the material removal volume However there are often constraints for corrective machining such as nonnegative error and allowance requirements making it a constrained minimization problem The basic algorithm for error calculation combines the alternating optimization method and the successive linearization method Whereas the sequence of objective function values is not guaranteed monotonously decreasing because of the local validity of linearization Therefore in the improved algorithm the problem is reformulated as an inequalityconstrained least square problem Bound constraints are imposed on the optimization variables to keep local validity Then the line search and Palaciostype adjustment strategy are incorporated in each iteration in order to find a better point reducing the value of objective function This point is chosen as the start of the next iteration Finally better convergence of the improved algorithm is verified by numerical simulations and a case of application in optical machining
Keywords: