2010 | OriginalPaper | Buchkapitel
Multiobjective Optimization
Erschienen in: Introduction to Evolutionary Algorithms
Verlag: Springer London
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
EAs are developed to solve real-world problems, such as designing and scheduling. In real conditions, there are many requirements to fulfill. In previous chapters, we sometimes wanted to model them into constraints because it is hard to compare two objectives simultaneously. Pareto gave us the idea of dominance, so we can divide the relationship between two vectors into three categories: one is better than the other, the converse is true, or they are incomparable or incommensurable. For problems with multiple objectives, there exist many “good” points that are no worse than any other point in the objective space. EAs contain a group of individuals. So if we can distribute the individuals evenly on these “good” points, it will be very helpful for designers and decision makers. This chapter discusses how to use EAs to solve such problems. It is a fascinating and hot research area. You will experience the shining wisdom of other researchers, which will deepen considerably your understanding of EAs.