Date of Award

9-20-2019

Publication Type

Master Thesis

Degree Name

M.Sc.

Department

Computer Science

First Advisor

Kobti, Z.

Keywords

Cultural Algorithms, Decomposition Methods, Evolutionary Computation, Optimization Problems

Rights

info:eu-repo/semantics/openAccess

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Abstract

Decomposition is used to solve optimization problems by introducing many simple scalar optimization subproblems and optimizing them simultaneously. Dynamic Multi-Objective Optimization Problems (DMOP) have several objective functions and constraints that vary over time. As a consequence of such dynamic changes, the optimal solutions may vary over time, affecting the performance of convergence. In this thesis, we propose a new Cultural Algorithm (CA) based on decomposition (CA/D). The objective of the CA/D algorithm is to decompose DMOP into a number of subproblems that can be optimized using the information shared by neighboring problems. The proposed CA/D approach is evaluated using a number of CEC 2015 optimization benchmark functions. When compared to CA, Multi-population CA (MPCA), and MPCA incorporating game strategies (MPCA-GS), the results obtained showed that CA/D outperformed them in 7 out of the 15 benchmark functions.

Share

COinS