Multi-Objective Optimal Design of Variably Constrained 2D Path Network Layouts with Application to Ascent Assembly Engineering

We present an effective optimization strategy that is capable of discovering high-quality cost-optimal solution for two-dimensional (2D) path network layouts (i.e., groups of obstacle-avoiding Euclidean Steiner trees) that, among other applications, can serve as templates for complete ascent assembly structures (CAA-structures). The main innovative aspect of our approach is that our aim is not restricted to simply synthesizing optimal assembly designs with regard to a given goal, but we also strive to discover the best tradeoffs between geometric and domain-dependent optimal designs. As such, the proposed approach is centered on a variably constrained multi-objective formulation of the optimal design task and on an efficient coevolutionary solver. The results we obtained on both artificial problems and realistic design scenarios based on an industrial test case empirically support the value of our contribution to the fields of optimal obstacleavoiding path generation in particular and design automation in general.