一类自适应蚁群算法及其收敛性分析

为了克服蚁群算法易陷入局部最小点的缺点,同时提高算法的收敛速度,提出一类自适应蚁群算法.该算法利用自适应改变信息激素的挥发系数改善传统蚁群算法的全局搜索能力和收敛速度.通过马尔科夫过程对算法的全局收敛性进行分析,得出该类蚁群算法全局收敛性条件.并构造出该类算法的一种信息激素更新策略,证明了这种算法全局收敛性.利用提出的算法对典型的TSP问题进行仿真研究,结果表明比典型蚁群算法在收敛速度和解的性能上都有较大改善.

Adaptive ant colony optimization algorithms and its convergence

A class of adaptive ant colony optimization(ACO) algorithms is presented to avoid the deficiency of typical ACO that often runs into local optimum.Global searching and convergence abilities are improved by adaptively changing the pheromone trails evaporation parameters.Some convergence properties for the algorithms are analyzed with the Markov process approach.Further more,an algorithm with guaranteed convergence to the optimal solution is developed.The simulation results for typical TSP problems demonstrate that the proposed algorithms are more effective than those for other modified ant systems.