.png){kind=icon}
.png){kind=icon}
Return
A two-step multi-objectivization method for improved evolutionary optimization of industrial problems
.png){kind=icon}
PRE
AI
DOI：10.1016/j.asoc.2017.12.027.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Multi-objectivization means that helper objectives are added to an optimization problem with the purpose of altering the search space in a way that improves the progress of the optimization algorithm. In this paper, a new method for multi-objectivization is proposed that is based on a two-step process. In the first step, a helper objective that conflicts with the main objective is added, and in the second step a helper objective that is in harmony with, but subservient to, the main objective is added. In contrast to existing methods for multi-objectivization, the proposed method aims at obtaining improved results in real-world optimizations by focusing on three aspects: (a) adding as little extra complexity to the problem as possible, (b) achieving an optimal balance between exploration and exploitation in order to promote an efficient search, and (c) ensuring that the main objective, which is of main interest to the user, is always prioritized. Results from evaluating the proposed method on a complex real-world scheduling problem and a theoretical benchmark problem show that the method outperforms both a traditional single-objective approach and the prevailing method for multi-objectivization. Besides describing the proposed method, the paper also outlines interesting aspects of multi-objectivization to investigate in the future. (C) 2018 Elsevier B.V. All rights reserved.
Keywords:
Multi-objectivization
Evolutionary algorithms
Optimization
Real-world scheduling problem
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.