面向AGV全局路径规划的蚁群算法多策略改进研究  

Research on Multi-Strategy Improvement of Ant Colony Algorithm for Global Path Planning for AGVs

作　　者：侯书增 孙巍峰 罗程远 伍志明 李轩 HOU Shu-zeng;SUN Wei-feng;LUO Cheng-yuan;WU Zhi-ming;LI Xuan(School of Mechanical Engineering,Sichuan University of Science and Engineering,Yibin 644005)

机构地区：[1]四川轻化工大学机械工程学院,四川宜宾644005

出　　处：《制造业自动化》2025年第3期1-8,共8页Manufacturing Automation

基　　金：国家自然科学基金(51961003);四川省科技厅重点研发项目(2023YFG0239)。

摘　　要：传统蚁群算法在应对大规模复杂场景时,普遍存在收敛效率偏低、全局搜索能力不足以及易陷入局部最优等缺陷。提出了一种多策略改进的蚁群算法。通过增强启发函数的导向性和自适应性,提高算法的寻优能力;利用伪随机转移策略改进状态转移规则并对信息素启发因子和期望启发因子进行动态调节,增强全局搜索能力,避免局部最优问题。同时,采用多策略优化信息素更新机制,并结合自适应调节挥发系数,显著提升收敛效率和搜索性能。此外,通过冗余点去除的二次优化,进一步优化路径质量。通过栅格地图的仿真实验表明,蚁群算法经多策略改进后在搜索能力和收敛速度上获得了较大提升,可为AGV解决复杂环境下的路径规划提供理论支持。Traditional ant colony algorithms generally suffer from low convergence efficiency,insufficient global search capabilities,and a tendency to fall into local optima when dealing with large-scale complex scenarios.The method enhances the guidance and adaptability of the heuristic function to improve the algorithm's optimization ability.By utilizing a pseudo-random transfer strategy to improve state transition rules and dynamically adjusting the pheromone heuristic factor and the expected heuristic factor,the global search capability is enhanced,and the issue of local optima is avoided.At the same time,a multi-strategy approach is adopted to optimize the pheromone update mechanism,combined with an adaptive adjustment of the evaporation coefficient,significantly improving convergence efficiency and search performance.Additionally,through secondary optimization by removing redundant points,the path quality is further optimized.The simulation experiments on grid maps demonstrate that the ant colony algorithm,after multiple strategy improvements,achieves significant improvements in search ability and convergence speed,providing theoretical support for solving path planning problems for AGVs in complex environments.

关 键 词：蚁群算法 启发函数 信息素挥发系数 路径二次优化 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]