AMCPA: A Population Metaheuristic With Adaptive Crossover Probability and Multi-Crossover Mechanism for Solving Combinatorial Optimization Problems

摘要: Combinatorial optimization is a field that receives much attention in artificial intelligence. Many problems of this type can be found the literature, and large number techniques have been developed to applied them. Nowadays, population algorithms become one most successful metaheuristics for solving kind problems. Among techniques, Genetic Algorithms (GA) received due its robustness easy applicability. In paper, an Adaptive Multi-Crossover Population Algorithm (AMCPA) proposed, which variant classic GA. The presented AMCPA changes philosophy basic GAs, giving priority mutation phase providing dynamism crossover probability. To prevent premature convergence, proposed AMCPA, probability begins with low value, adapted every generation. Apart from this, as another mechanism avoid different functions are used alternatively. order prove quality technique, it six combinatorial problems, results compared ones obtained by Additionally, convergence behaviour both also compared. Furthermore, objective performing rigorous comparison, statistical study conducted compare these outcomes. during test are: Symmetric Asymmetric Traveling Salesman Problem, Capacitated Vehicle Routing Problem Backhauls, N-Queens, one-dimensional Bin Packing Problem.