Simulated annealing approach with an exhaustive greedy search for the optimal machining region decomposition in CNC roughing tool path generation

Abstract

For machining various parts, it is important to generate efficient tool paths for productivity. At the stage of roughing, the workspace is sliced and the paths are generated outside the part in each sliced plane. One of important decision variables in planar tool path generation is the direction in which paths are aligned. This paper is about decomposing the machining region and deciding the path direction in each subregion so that machining time can be reduced. The machining region is decomposed into discretized meshes and the path planning problem is formulated as a combinatorial optimization problem. The formulated problem is solved by the simulated annealing technique. To speed up the overall optimization procedure, the annealing procedure is rather prematurely terminated and a clearing algorithm is applied to remove small subregions. This is needed since the cost function of this problem has many ‘flat’ areas and provides ‘narrow’ passages to its optimum.