Tabu efficient global optimization with applications in additive manufacturing

Abstract Methods based on Gaussian stochastic process (GSP) models and expected improvement (EI) functions have been promising for box-constrained expensive optimization problems. These include robust design problems with environmental variables having set-type constraints. However, the methods that combine GSP EI sub-optimizations suffer from following problem, which limits their computational performance. Efficient global (EGO) often repeat same or nearly experimental points. We present a novel EGO-type constraint-handling method maintains so-called tabu list to avoid past Our includes two types of penalties key “infill” optimization, selects next test runs. benchmark our EGO algorithm five alternative approaches, including DIRECT using nine engineering examples. The examples are additive manufacturing parameter informed point-based thermal simulations robust-type quality span unconstrained, simply constrained, constrained comparative results imply offers very performance all black-box in terms convergence speed final solution.