Title :
Art-enhanced modified binary differential evolution algorithm for optimization
Author :
Wu, Chun-yin ; Kuan-Shien Nu
Author_Institution :
Dept. of Mech. Eng., Tatung Univ., Taipei, Taiwan
Abstract :
Differential evolution (DE) is a heuristic optimization method with a relatively simple and efficient form of mutation and crossover and it has been applied to solve many real world optimization problems in real-valued search space. Modified binary differential evolution (MBDE) with a simple binary mutation mechanism based on a logical operation is suitable for dealing with binary and continuous optimization problems. In this study, the modified binary differential evolution is enhanced by using adaptive resonance theory (ART) to classify binary image pattern of population into groups for balancing exploration and exploitation in optimization search. The diversity and convergence of search are both enhanced by applying ART clustering strategy. Different types of optimization problems consisting of test function optimization and topology optimization of structure are used to illustrate the high viability of the proposed algorithm in optimization.
Keywords :
differential equations; evolutionary computation; ART; MBDE; adaptive resonance theory; art enhanced modified binary differential evolution algorithm; binary mutation mechanism; image pattern; logical operation; optimization method; Abstracts; Algorithm design and analysis; Clustering algorithms; Optimization; Adaptive resonance theory; Exploration and exploitation; Modified binary differential evolution;
Conference_Titel :
Machine Learning and Cybernetics (ICMLC), 2012 International Conference on
Conference_Location :
Xian
Print_ISBN :
978-1-4673-1484-8
DOI :
10.1109/ICMLC.2012.6359655