| Paper | Title | Page |
|---|---|---|
| THPOST020 | Visualisation of Pareto Optimal Spaces and Optimisation Solution Selection Using Parallel Coordinate Plots | 2487 |
| SUSPMF017 | use link to see paper's listing under its alternate paper code | |
|
||
| In this paper, we build on previous work where multi-objective genetic algorithms were used to optimise RF cavities using non-uniform rational basis splines (NURBS) to improve the cavity geometries and reduce peak fields. These optimisations can produce thousands of Pareto optimal solutions, from which a final cavity solution must be selected based on design criteria, such as accelerating gradient and power requirements. As all points are considered equally optimal, this can prove difficult without further analysis. Here we focus on the visualisation of the Pareto optimal points and the final solution selection process. We have found that the use of clustering algorithms and parallel coordinate plots (PCPs) provide the best way to represent the data and perform the necessary trade-offs between the peak fields and shunt impedance required to pick a final design. | ||
| DOI • | reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-THPOST020 | |
| About • | Received ※ 08 June 2022 — Revised ※ 13 June 2022 — Accepted ※ 13 June 2022 — Issue date ※ 29 June 2022 | |
| Cite • | reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml) | |