Solving Constrained Optimization Problems with Hybrid Evolutionary Algorithms

摘要: The foundations for evolutionary algorithms (EAs) were established in the end of 60’s [1, 2] and strengthened beginning 70’s [3, 4]. EAs appeared as an alternative to exact or approximate optimization methods whose application many real problems not acceptable terms performance. When applied problems, provide a valuable relation between quality solution efficiency obtain it; this reason these techniques attracted immediately attention researchers became what they nowadays represent: cutting-edge approach real-world optimization. Certainly, has also been case other related techniques, such simulated annealing [5] (SA), tabu search [6] (TS), etc. term metaheuristics coined denote them. hybrid algorithm (HEAs) (resp. metaheuristics) refers combination technique with another (perhaps approximate) aim is combine best both worlds objective producing better results than each involved components working alone. HEAs have proved be very successful practical (e.g., [7, 8]) and, consequence, currently there exist increasing interest community kind techniques. One crucial point development (and general) need exploiting problem knowledge was clearly exposed formulation No Free Lunch Theorem (NFL) by Wolpert Macready [9] (a performs strict accordance amount incorporate). Quite interestingly, line thinking had