A Constraint Handling Approach with Guaranteed Feasibility for Surrogate Based Optimization

TL;DR

This paper introduces a new constraint-handling method for surrogate-based optimization that guarantees feasible solutions, improving reliability in constrained black-box optimization problems, especially in engineering applications.

Contribution

The paper presents a novel constraint-handling approach that ensures feasibility during surrogate-based optimization, addressing a key limitation of existing methods.

Findings

The proposed method guarantees strictly feasible candidates during optimization.

It outperforms other SBO, GAs, and gradient-based algorithms on test functions.

It effectively optimizes a CFD problem involving fish-like swimming motion.

Abstract

Gradient-free optimization methods, such as surrogate based optimization (SBO) methods, and genetic (GAs), or evolutionary (EAs) algorithms have gained popularity in the field of constrained optimization of expensive black-box functions. However, constraint-handling methods, by both classes of solvers, do not usually guarantee strictly feasible candidates during optimization. This can become an issue in applied engineering problems where design variables must remain feasible for simulations to not fail. We propose a constraint-handling method for computationally inexpensive constraint functions which guarantees strictly feasible candidates when using a surrogate-based optimizer. We compare our method to other SBO, GA/EA and gradient-based algorithms on two (relatively simple and relatively hard) analytical test functions, and an applied fully-resolved Computational Fluid Dynamics (CFD) problem concerned with optimization of an undulatory swimming of a fish-like body, and show that the proposed algorithm shows favorable results while guaranteeing feasible candidates.