Axial compressor rotor optimization using a novel ensemble of surrogates-based infill criterion

Abstract

A new infill criterion for so-called ensemble of surrogates based optimization is proposed and put to the test in practice to aerodynamically optimize a compressor rotor. The optimization method takes advantage of five radial basis functions using a weighted linear combination. This set of radial basis functions
making up the ensemble across the design space is dynamically changed using the maximization of a custom metric based on the local surrogate’s agreement and the global accuracy of each ensemble combination. The selected ensemble is then searched using a hybrid optimizer, viz., a global genetic algorithm and local SQP searches, and finally the optimal points are evaluated using the rotor’s black box function. The results are compared with established optimization approaches. The best design is analysed further in terms of the flow physics. Similar to a realistic industrial setting, the constrained, single-objective shape optimization of NASA Rotor 37 involves 25 design variables with a limited computational budget.

More information

Identifiers

ORCID for David J.J. Toal: orcid.org/0000-0002-2203-0302

ORCID for Andy J. Keane: orcid.org/0000-0001-7993-1569

Catalogue record

Export record