Accelerating adaptive trade-off model using shrinking space technique for constrained evolutionary optimization

Adaptive trade-off model (ATM) is a constraint-handling mechanism proposed recently. The main advantages of this model are it, simplicity and adaptation, Moreover, it can be easily embedded into evolutionary algorithms for solving constrained optimization problems. This paper proposes a novel method for constrained optimization, which aims at accelerating the ATM using shrinking space technique. Eighteen benchmark test functions and five engineering design problems are used to test the performance of the method proposed. Experimental results suggest that combining the ATM with the shrinking space technique is very beneficial. The method proposed can promptly converge to competitive results without loss of the quality and the precision of the final results. Performance comparisons with some other state-of-the-art approaches from the literature are also presented. Copyright (C) 2008 John Wiley & Sons, Ltd.