Multi-objective evolutionary optimization in time-changing environments

This research is focused on the creation of evolutionary optimization techniques for the solution of time-changing multi-objective problems. Many optimization problems, ranging from the design of controllers for time-variant systems to the optimal asset allocation in financial portfolios, need to satisfy multiple conflicting objectives that change in time. Since most practical problems involve costly numerical simulations, the goal was to create algorithmic architectures that increase computational efficiency while being robust and widely applicable. A combination of two elements lies at the core of the proposed algorithm. First, there is an anticipatory population that helps the algorithm discover the new optimum when the objective landscape moves in time. Second, a preservation of the balance between convergence and diversity in the population which provides an exploration ability to the algorithm. If there is an amount of predictability in the landscape's temporal change pattern the anticipatory population increases performance by discovering each timestep's optimal solution using fewer function evaluations. It does so by estimating the optimal solution's motion with a forecasting model and then placing anticipatory individuals at the estimated future location.

Department dc:contributor.department

Massachusetts Institute of Technology. Dept. of Mechanical Engineering.

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2007

Author dc:creator

• Hatzakis, Iason

Advisor dc:contributor.advisor

• David R. Wallace.

• Mechanical Engineering.

Statement dc:rights

• M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission.

Licence dc:rights.uri

• http://dspace.mit.edu/handle/1721.1/7582

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/1721.1/39842

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/39842