GUI design to allow for easy use of complex multi-objective optimization algorithms

Level of Education of Students Involved

Undergraduate

Faculty Sponsor

Jesse M Sestito

College

Engineering

Discipline(s)

Mechanical Engineering

ORCID Identifier(s)

0000-0002-4489-7054

Presentation Type

Poster Presentation

Symposium Date

Spring 4-25-2024

Abstract

Multi-objective optimization is a type of decision-making algorithm that recommends the ideal input parameters to optimize multiple output objectives congruently. These algorithms input existing parameter samples and their corresponding objective solutions to predict possible unknown combinations of parameter values that lie on the Pareto front, a non-dominated front of optimal solutions. Optimization is valuable to many disciplines, but particularly impacts engineering design considerations. However, current multi-objective optimization algorithms require intimate knowledge of programming and of the algorithm itself in order to use. This hinders the ability of engineers from all disciplines to use multi-objective optimization in their design process. To make this tool more accessible, a graphical user interface (GUI) was created that allows engineers to input their existing samples and solutions, and select an optimization function to recommend new parameter values that are most likely to be optimal. The GUI is designed to be intuitive and user-friendly, allowing people with a wide variety of expertise to use this program. It also allows users to customize the parameter boundaries, output batch size, objective minimization and maximization, and allows for different ways to import and enter data. This interface is a convenient powerful tool enabling many more engineers to take advantage of the power of multi-objective optimization.

Biographical Information about Author(s)

Kyle Morrison is a junior mechanical engineering major with a computer science minor from Lake Zurich, Illinois. He started research with Prof. Sestito in Spring 2023.

This document is currently not available here.

Share

COinS