PROBLEMS USING DIFFERENTIAL EVOLUTION Roger, L.S. Tan, Mekapati Srinivas and G.P. Rangaiah* Department of Chemical and Biomolecular Engineering National University of Singapore, 4 Engineering Drive 4, SINGAPORE 117576 *Author for correspondence; Email: chegpr@nus.edu.sg Fax: (65) 6779 1936; Phone: (65) 6874 2187 Abstract The computational efficiency and reliability of one recent stochastic method for minimizing nonlinear and non-differentiable continuous space functions, Differential Evolution (DE) proposed by Storn and Price [1], is examined by application to many benchmark problems. The DE algorithm is modified to improve its computational efficiency while maintaining its reliability. The application of DE is then extended to phase equilibrium calculations where it has yet to be tested. The results show that DE is reliable in locating the global minimum for most problems but its computational efficiency is less than Tabu Search and Genetic Algorithm. Furthermore, DE is easy to use, involves a few control variables and is highly reliable in locating the global minimum in consecutive independent trials, even when little information of the problem known.
[1]
Rainer Storn,et al.
Differential Evolution – A Simple and Efficient Heuristic for global Optimization over Continuous Spaces
,
1997,
J. Glob. Optim..
[2]
Xueguang Shao,et al.
A fast annealing evolutionary algorithm for global optimization
,
2002,
J. Comput. Chem..
[3]
R. Storn,et al.
Differential Evolution
,
2004
.
[4]
Mangesh D. Kapadi,et al.
Optimal control of fed-batch fermentation involving multiple feeds using Differential Evolution
,
2004
.
[5]
Patrick Siarry,et al.
Tabu Search applied to global optimization
,
2000,
Eur. J. Oper. Res..
[6]
Gade Pandu Rangaiah,et al.
Tabu search for global optimization of continuous functions with application to phase equilibrium calculations
,
2003,
Comput. Chem. Eng..