CAJ | 학술논문

针对蚁群算法在解决大规模优化问题中易陷入局部最优及收敛速度慢等缺陷，提出了一种基于云模型的自适应蚁群算法。通过对信息素分布状况进行评价，自适应地调整云模型中各参数，并根据云模型参数，确定全局最优及次优路径，进行全局信息素更新，以改善算法的全局搜索能力。同时，为了避免算法陷入停滞，将信息素大小限制在一个最大最小区间。仿真实验结果验证了提出的算法的高效性和稳定性。
침대의군산법재해결대규모우화문제중역함입국부최우급수렴속도만등결함，제출료일충기우운모형적자괄응의군산법。통과대신식소분포상황진행평개，자괄응지조정운모형중각삼수，병근거운모형삼수，학정전국최우급차우로경，진행전국신식소경신，이개선산법적전국수색능력。동시，위료피면산법함입정체，장신식소대소한제재일개최대최소구간。방진실험결과험증료제출적산법적고효성화은정성。
Ant colony algorithm is easy to fall in local best, and its convergent speed is slow in solving large-scale optimiza-tion problems. In this paper, a self-adaptive ant colony algorithm based on cloud model is proposed. Through the evaluation of pher-omone distribution, the parameters of the cloud model are adjusted adaptively. According to the parameters, the global optimal and suboptimal paths are determined, and the global pheromone is updated to improve the global search ability of the algorithm. Mean-while, in order to avoid stagnation, the range of pheromone is limited to a maximum-minimum interval. Simulation results validate the efficiency and stability of the proposed algorithm.