一种求解连续空间优化问题的改进蚁群算法

蚁群算法是近几年优化领域中新出现的一种启发式仿生类并行智能进化算法,该算法采用分布式并行计算和正反馈机制,易于与其它方法结合,目前虽然已经在离散空间优化领域中得到了广泛应用,但是在求解连续空间优化问题方面的研究相对较少。在介绍基本蚁群算法机制原理和数学模型的基础上,提出了一种用于求解连续空间优化问题的改进蚁群算法。将连续空间优化问题的解向量分解成有限个网格,同时构造了一个与蚁群转移概率相关的评价函数,并借助相遇搜索策略对蚁群算法进行了改进,将各条寻优路径上可能的残留信息素数量限制在一个最大最小区间,以提高改进后蚁群算法的全局收敛性能。仿真实验表明,提出的改进蚁群算法较文献11]所提出的自适应蚁群算法能更快地找到连续空间优化问题更优良的全局解,从而为蚁群算法求解这类问题提供了一条可行有效的新途径。

Improved Ant Colony Algorithm for Solving Continuous Space Optimization Problems

Ant colony algorithm is a novel category of bionic meta-heuristic algorithm,and parallel computation and positive feedback mechanism are adopted in this algorithm.The ant colony algorithm has strong robustness and easy to combine with other methods in optimization.Although the ant colony algorithm for the heuristic solution of discrete space optimization problems enjoys a rapidly growing popularity,but few are reported for the heuristic solution of continuous space optimization problems.Based on the introduction of the mechanism and mathematical model of basic ant colony algorithm,an improved ant colony algorithm for solving continuous space optimization problems was proposed.The solution vector of continuous space optimization problem was decomposed into finite grids.Meanwhile,the cost function related to the transition probability was constructed.In order to enhance the global convergence performance of the improved ant colony algorithm,meeting search strategy was adopted in the improved ant colony algorithm,and the range of possible pheromone trails on each solution component was limited to a maximum-minimum interval.The numerical simulation results demonstrate that the improved ant colony algorithm can find better global solution for continuous space optimization problems than the adaptive ant colony algorithm proposed in the literature11],and this new algorithm presents a feasible and effective way to solve various continuous space optimization problems.