Compared with traditional structures, lattice structures show more excellent mechanical properties, e.g., higher rigidity and lighter weight. The additive manufacturing (AM) technology enables complex lattice structure to be realized. The path planning problem of AM for lattice structures faces new challenges that have been barely explored before now, because the lattice structures have unique characteristics of complex geometrical features and high-precision processing requirements. Despite many efforts toward path generation, the printing efficiency problem of the lattice structures is more than one single printing path can resolve. Using machine learning method based on a support vector machine (SVM) system, the goal of this paper is to automatically determine a suitable filling path for each sub-domain of the slicing layer of the lattice structures. Moreover, a route is found to traverse every scattered sub-domain by solving a traveling salesman problem (TSP). We demonstrate various testing examples and the experimental results to show the superiority and effectiveness of our method in terms of classification accuracy, filling effects, and linking path length.

This work is published on International Journal of Advanced Manufacturing Technology 116.5-6(2021):1467-1490.