Reinforcement Learning for Accelerated Aerodynamic Shape Optimisation

TL;DR

This paper presents an RL-based adaptive optimization method for aerodynamic shape design that reduces computational effort and enhances interpretability of the optimization process.

Contribution

It introduces a surrogate-based, actor-critic RL approach with temporal parameter freezing for efficient aerodynamic shape optimization.

Findings

Speeds up global optimization via local parameter adjustments.

Allows interpretation of flow-field extrema through feature importance.

Demonstrates effectiveness on a fluid-dynamical problem.

Abstract

We introduce a reinforcement learning (RL) based adaptive optimization algorithm for aerodynamic shape optimization focused on dimensionality reduction. The form in which RL is applied here is that of a surrogate-based, actor-critic policy evaluation MCMC approach allowing for temporal 'freezing' of some of the parameters to be optimized. The goals are to minimize computational effort, and to use the observed optimization results for interpretation of the discovered extrema in terms of their role in achieving the desired flow-field. By a sequence of local optimized parameter changes around intermediate CFD simulations acting as ground truth, it is possible to speed up the global optimization if (a) the local neighbourhoods of the parameters in which the changed parameters must reside are sufficiently large to compete with the grid-sized steps and its large number of simulations, and (b) the estimates of the rewards and costs on these neighbourhoods necessary for a good step-wise parameter adaption are sufficiently accurate. We give an example of a simple fluid-dynamical problem on which the method allows interpretation in the sense of a feature importance scoring.