基于柯西-高斯分布ACO-DWA导盲路径优化  

Cauchy-Gaussian Distribution Based ACO-DWA Guided Blind Path Optimization

作　　者：李静 黄明 李志峰 LI Jing;HUANG Ming;LI Zhifeng(School of Mechanical Engineering,Jiangsu University of Technology,Changzhou 213000,China)

机构地区：[1]江苏理工学院机械工程学院,常州213000

出　　处：《自动化与仪表》2025年第3期48-53,共6页Automation & Instrumentation

基　　金：江苏省输配电装备技术重点实验室自主科研课题(2022JSSPD13);江苏高校“青蓝工程”资助项目。

摘　　要：为提升导盲机器狗在复杂环境中的导航性能,提出了一种基于柯西-高斯分布优化的蚁群算法与动态窗口评估(DWA)算法融合的路径规划方法。通过柯西-高斯分布优化改进蚁群算法,加速收敛并避免局部最优。结合三次B样条曲线平滑轨迹,增强了轨迹的光滑性。DWA算法中新增障碍物评价和全局路径追踪功能,通过加权系数平衡避障与路径追踪。仿真实验表明,改进融合算法相较于传统蚁群算法在路径长度上缩短了13.5%和10.7%,在迭代次数上减少了63.5%和61.0%,在转折次数上减少了33.3%和44.4%,并经ROS平台验证其动态规划的可行性,为导盲机器狗提供最优路径规划。In order to enhance the navigation performance of guide dogs for the blind in complex environments,a path planning method that integrates an ant colony optimization(ACO)algorithm optimized by Cauchy-Gaussian distribution with a dynamic window approach(DWA)algorithm was proposed.The ACO is refined using Cauchy-Gaussian optimiza-tion to speed up convergence and mitigate local optima issues.To improve trajectory smoothness,cubic B-spline curve smoothing is applied.Additionally,the DWA algorithm incorporates obstacle evaluation and global path tracking fea-tures,utilizing weighted coefficients to balance between avoiding obstacles and following the path accurately.Simula-tion results indicate that this enhanced fusion method achieves a reduction of 13.5%and 10.7%in path length com-pared to traditional AcO methods,along with decreases of 63.5%and 61.0%in iteration counts,as well as reductions of 33.3%and 44.4%in turning points.The effectiveness of dynamic planning has been validated on the ROS plat-form,offering optimal path solutions for guide dog robots.

关 键 词：导盲机器狗 路径规划 优化蚁群算法 动态避障 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]