Data-driven Harris Hawks constrained optimization for computationally expensive constrained problems

Abstract Aiming at the constrained optimization problem where function evaluation is time-consuming, this paper proposed a novel algorithm called data-driven Harris Hawks (DHHCO). In DHHCO, Kriging models are utilized to prospect potentially optimal areas by leveraging computationally expensive historical data during optimization. Three powerful strategies are, respectively, embedded into different phases of conventional (HHO) generate diverse candidate sample for exploiting around existing and exploring uncharted region. Moreover, Kriging-based strategy composed population construction individual selection presented, which fully mines utilizes potential available information in data. DHHCO inherits develops HHO's offspring updating mechanism, meanwhile exerts prediction ability Kriging, reduces number evaluations, provides new ideas constraint Comprehensive experiments have been conducted on 13 benchmark functions real-world problem. The experimental results suggest that can achieve quite competitive performance compared with six representative algorithms find near global optimum 200 evaluations most examples. applied structural internal components real underwater vehicle, final satisfactory weight reduction effect more than 18%.