【摘要】 借鉴蚁群优化算法和粒子群优化算法的思想,提出了一种用于求解约束优化问题的连续域蚁群算法。将搜索域中的任意一点看成食物源,使用多组蚁群进行寻优,每一组蚁群代表问题的一个解,在每一迭代中首先在所有蚁群中选则一组种子蚁群,然后在该组蚁群的信息素密度分布函数下进行采样,生成子代蚁群,最后进行蚁群选择,从而使各组蚁群不断向适应度值较高的搜索区域移动,最终收敛到最优解。对基准测试函数G01-G12的求解结果表明,该方法具有较快的收敛速度和较好的全局寻优能力。更多还原

【Abstract】 Using the idea of both ant colony optimization algorithm and particle swam optimization algorithm,a continuous domains ant colony algorithm for solving constrained optimization problem was proposed. Regarding any point within the search space as the food,the optimal solution was searched using several ant colonies,so every ant colony represented a solution. In any iteration,a seed ant colony was selected from all the ant colonies,and the pheromone density function was sampled with,the child ant colonies were made out. Finally,by ant colony selections,the groups of ant colonies continually move to the search space of higher fitness value,and converge to the optimal solution in the end. The solving results of the function G01-G12 show that,this method has better convergence speed and greater global optimization ability.更多还原