DocumentCode :
3216894
Title :
Is stochastic ranking really better than Feasibility Rules for constraint handling in Evolutionary Algorithms?
Author :
Bansal, Sulabh ; Mani, Ashish ; Patvardhan, C.
Author_Institution :
Dept. of Comput. Sci. & Eng., Anand Eng. Coll., Agra, India
fYear :
2009
fDate :
9-11 Dec. 2009
Firstpage :
1564
Lastpage :
1567
Abstract :
Evolutionary algorithms have been widely used to solve difficult constrained optimization problems. However, evolutionary algorithms are intrinsically unconstrained optimization techniques. Constraint handling is mostly incorporated additionally and its choice has great bearing on the quality of the solution. Stochastic ranking was introduced as an improvement over feasibility rules for handling constraints in evolutionary optimization. It is widely believed that stochastic ranking is currently the best-known technique for handling constraints. However, a fair comparative study has never been attempted in the literature, where by the performance of both the constraint handling technique is compared on the same evolutionary algorithm. This paper fairly compares the performance of both the constraint handling techniques on the same evolutionary algorithm over a set of parameters like feasibility rate, successful run, success rate and success performance in addition to objective function value and number of function evaluations. The results put a question mark over the belief that feasibility rules are worse than stochastic ranking.
Keywords :
constraint handling; evolutionary computation; stochastic processes; constraint handling; evolutionary algorithm; evolutionary optimization; feasibility rule; function evaluation; objective function value; stochastic ranking; unconstrained optimization; Algorithm design and analysis; Computer science; Constraint optimization; Educational institutions; Electronic mail; Evolutionary computation; Functional programming; Search methods; Stochastic processes; Testing; Constraint handling; Evolutionary algorithms; Optimization;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Nature & Biologically Inspired Computing, 2009. NaBIC 2009. World Congress on
Conference_Location :
Coimbatore
Print_ISBN :
978-1-4244-5053-4
Type :
conf
DOI :
10.1109/NABIC.2009.5393677
Filename :
5393677
Link To Document :
بازگشت