Surrogate-based optimization with adaptive parallel infill strategy enhanced by inaccurate multi-objective search

In recent decades, surrogate-based optimization (SBO) has been developed to replace costly models with cheap surrogates improve efficiency. this article, an adaptive parallel infill strategy is proposed balance exploration and exploitation over the design space during process of SBO. Within method, inaccurate search adopted optimize surrogate models, thereby helping locate point. An elite archive exploited store superior sampling points for batch sampling, while a customized size determination introduced. The SBO method its tested on six unconstrained five constrained analytical cases results compared state-of-the-art algorithms. 582-bar tower truss system also performed utilized verify method. shows excellent performance better stability.