2272. Multiobjective optimisation of a multifidelity simulation
Invited abstract in session WA-35: Recent trends in zeroth order and simulation-based optimization: 2, stream Continuous and mixed-integer nonlinear programming: theory and algorithms.
Wednesday, 8:30-10:00Room: Michael Sadler LG15
Authors (first author is the speaker)
| 1. | Warren Hare
|
| University of British Columbia |
Abstract
Optical dosimetry uses light to measure radiation dosage. It is a critical step in ensuring radiation therapy is delivered safely and accurately. This talk explores the design of a solid-state tank for fan-beam optical dosimetry. We begin with an overview of optical dosimetry, its applications, and solid-state tank design. We then discuss a novel simulation for solid-state tank design that measures image uniformity and scanning coverage. The simulation is used to power a multiobjective optimisation problem, which we solve using NSGA-II and simple multifidelity soft restart techniques. Combined the research provides several useful results: a new test problem for multifidelity multiobjective optimisation, and an improved solid-state tank design for optical dosimetry.
Keywords
- Programming, Multi-Objective
- Health Care
- Robust Optimization
Status: accepted
Back to the list of papers