简化蚁群算法

Simplified ant colony optimization algorithm

针对最大最小蚂蚁系统中信息素下界难以确定以及算法性能易受同构问题影响的缺点,提出一种简化蚁群算法,信息素的上下界被限制在一个固定的区间内,不随目标函数值的更新而改变;信息素的更新量是一个与具体目标函数值无关的常数,所提出的简化算法不仅具有强不变性和平移不变性,而且算法的性能不受信息素下界的影响。针对旅行商问题的仿真实验验证了改进算法的可行性和有效性。

Due to the shortcomings of max-min ant system（MMAS）, which is difficult to decide the lower bound of pheromone trail and easy to affect by isomorphic problems, a simplifiedant colony optimization（SACO） algorithm is proposed. The upper and lower bound of pheromone trail are limited to a fixed interval and can not be changed with updating the objective function value. The added pheromone trail is a constant which is independent to the function value. It is proved that the algorithm not only has the property of linear transformational invariance and translational invariance, but also its performance is not affected by the lower bound of pheromone trail. Simulation results on the traveling salesman problem show the feasibility and effectiveness of the presented algorithm.