摘要
针对传统蚁群算法在林业机器人路径规划中未能考虑林区地形与地表状态,且收敛速度慢,易陷入局部最优的问题,提出一种基于林地综合代价地图与开拓优化机制的改进蚁群算法。基于林区地形地表构建林地综合代价地图;在蚁群搜索阶段中引入开拓优化机制;在轮盘赌中引入随机游走机制,并将其应用到林业机器人路径规划问题中。实验表明:改进蚁群算法能搜寻到更佳路径,具有更好的全局寻优能力。
An improved ant colony algorithm based on the comprehensive cost map of forest land and the research-optimization mechanism is proposed to address the problem of the traditional ant colony algorithms which is absent of considerationof forest terrain and surface conditions in forestry robot path planning,slow at convergence speed and prone to falling into local optima.On the terrain and surface of the forest area,a comprehensive forest land cost map is constructed;a pioneering optimization mechanism is introduced in the ant colony search stage;and a random walk mechanism is led into the roulette wheel mechanism,which is applied the path planning problem of forestry robots.The experiments show that the proposed algorithm proposed can search for better paths and has better exploration and optimization capabilities.
作者
何浩天
彭富明
方斌
相福磊
张少杰
HE Haotian;PENG Fuming;FANG Bin;XIANG Fulei;ZHANG Shaojie(School of Automation,Nanjing University of Science and Technology,Nanjing 210094,China)
出处
《机械制造与自动化》
2024年第4期146-150,共5页
Machine Building & Automation
基金
国家重点研发计划项目(2021YFE0194600)
江苏省科技计划项目(BZ2023023)。
关键词
蚁群算法
精英奖励
双层蚁群算法
林间路径规划
ant colony
elite rewards
double layer ant colony algorithm
forest path planning
