Multi-objective Bayesian optimization for design of Pareto-optimal current drive profiles in STEP

TL;DR

This paper introduces a multi-objective Bayesian optimization method for designing electron-cyclotron heating profiles in tokamaks, producing diverse Pareto-optimal solutions that outperform genetic algorithms in quality and variety.

Contribution

The paper presents a novel application of multi-objective Bayesian optimization to plasma profile design, offering more diverse and higher-quality solutions than previous genetic algorithm approaches.

Findings

Generated solutions with higher scores than genetic algorithms.

Produced a diverse set of Pareto-optimal solutions.

Achieved this with no additional computational cost.

Abstract

The safety factor profile is a key property in determining the stability of tokamak plasmas. To design the safety factor profile in the United Kingdom's proposed Spherical Tokamak for Energy Production (STEP), we apply multi-objective Bayesian optimisation to design electron-cyclotron heating profiles. Bayesian optimisation is an iterative machine learning technique that uses an uncertainty-aware predictive model to choose the next designs to evaluate based on the data gathered during optimisation. By taking a multi-objective approach, the optimiser generates sets of solutions that represent optimal tradeoffs between objectives, enabling decision makers to understand the compromises made in each design. The solutions from our method score higher than those generated in previous work by a genetic algorithm; however, the key result is that our method returns a purposefully diverse range of optimal solutions, providing more information to tokamak designers without incurring additional computational cost.