Abstract: The positive feedback is one of main characteristics of Ant Colony Optimization（ACO），it works through the cumulation of the pheromone by which the ant’s behavior is biased，then attracts more ants to choose this path.The positive feedback benefits for the rapid discovery of good solutions，but it makes the local optima possible，so，we propose the negative feedback which works through the repulsion pheromone.Experimental results show that the application of negative feedback improves the algorithm’s ability to find the good solutions and it is feasible.

Key words: Ant Colony Optimization（ACO）, positive feedback, negative feedback, attraction pheromone, repulsion pheromone

摘要： 正反馈机制是蚁群算法的一个重要特征，它通过信息素的累积作用对蚂蚁的寻径产生诱导，从而吸引更多的蚂蚁，加快了发现较优解的速度，但是同时也为陷入局部最优埋下了隐患，在此基础上，引入了负反馈机制，通过排斥信息素来实现。实验表明，负反馈机制的应用增强了算法发现最优解的能力，是实际可行的。

关键词: 蚁群算法, 正反馈, 负反馈, 吸引信息素, 排斥信息素