摘要
传统蚁群算法在进行路径规划时存在收敛速度较慢,容易出现局部最优化,并且在复杂的环境中算法效率大幅下降的问题,提出一种优化的蚁群算法。该改进算法在路径搜寻的每一次迭代过程中,信息素会按改进规则重新分布。改进的算法通过动态调整状态转移概率,使算法避免停滞,避免产生局部最优化问题。在解决蚂蚁死锁问题方面提出清除策略,能够有效加快算法收敛速度,提高路径规划算法的鲁棒性。从仿真结果可以看出,相同环境下,本文改进算法搜索的最优路径长度,对比基本蚁群算法减少8.53%,对比已发表文献算法减少5.60%,搜索效率对比基本蚁群算法提升65.38%,对比已发表同篇文献中改进算法效率提升53.63%。
The traditional ant colony algorithm has the problems of slow convergence speed, being prone to local optimization and sharp decline in algorithm efficiency in complex environment. To deal with the problems, An optimized ant colony algorithm is proposed. In each iteration of path search, the pheromone could be redistributed according to the improved rules. The improved algorithm avoids stagnation and local optimization problems by dynamically adjusting the state transition probability. In solving the ant deadlock problem, a clearing strategy is proposed, which can effectively accelerate the convergence speed of the algorithm and improve the robustness of the path planning algorithm. The analysis of the simulation results could be concluded that under the same environment, the optimal path length of the improved algorithm in this paper is reduced by 8.53% compared with the basic ant colony algorithm, 5.60% compared with the algorithm in the published reference, and the search efficiency is increased by 65.38% compared with the basic ant colony algorithm, 53.63% compared with the improved algorithm in the same published reference.
作者
于莲芝
秦天
YU Lianzhi;QIN Tian(School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)
出处
《智能计算机与应用》
2022年第10期62-67,共6页
Intelligent Computer and Applications
基金
国家重点研发计划项目(2018YFB1700902)。
关键词
蚁群算法
路径规划
动态更新
清除策略
ant colony algorithm
path planning
dynamic update
clearance strategy
