寿林,王雷,蔡劲草,张泽宇,夏强强.基于改进遗传算法的移动机器人路径规划研究[J].井冈山大学自然版,2024,45(6):87-93 |
基于改进遗传算法的移动机器人路径规划研究 |
PATH PLANNING FOR MOBILE ROBOT BASED ON IMPROVED GENETIC ALGORITHM |
投稿时间:2024-06-11 修订日期:2024-07-28 |
DOI:10.3969/j.issn.1674-8085.2024.06.012 |
中文关键词: 移动机器人 路径规划 改进遗传算法 平滑路径 种群初始化 |
英文关键词: mobile robot path planning improved genetic algorithm smooth path population initialization |
基金项目:国家自然科学基金项目(51305001);安徽省高校优秀拔尖人才培育项目(gxbjZD2022023);安徽省高校自然科学研究重点项目(2022AH050978);机器视觉检测安徽省重点实验室开放基金项目(KLMVI-2024-HIT-15);安徽工程大学-鸠江区产业协同创新专项基金项目(2022cyxtb6) |
作者 | 单位 | E-mail | 寿林 | 安徽工程大学机械与汽车工程学院, 安徽, 芜湖 241000 | | 王雷 | 安徽工程大学机械与汽车工程学院, 安徽, 芜湖 241000 | wangdalei2000@126.com | 蔡劲草 | 安徽工程大学机械与汽车工程学院, 安徽, 芜湖 241000 | | 张泽宇 | 安徽工程大学机械与汽车工程学院, 安徽, 芜湖 241000 | | 夏强强 | 长三角哈特机器人产业技术研究院, 安徽, 芜湖 241000 | |
|
摘要点击次数: 71 |
全文下载次数: 101 |
中文摘要: |
针对传统遗传算法在移动机器人路径规划中存在路径不平滑、容易早熟等问题,提出了一种改进遗传算法。首先,采用基于终点距离和夹角的种群初始化方法以提高初始种群质量;其次,根据路径长度和转向角度设计了新的适应度函数以提高路径平滑度,采用保留较差解的改进锦标赛选择策略以增加种群多样性;最后,设计了基于节点数量筛选交叉个体的交叉算子,加快收敛速度并减少无效交叉,同时采用了多种变异策略以增加算法后期的搜索能力。对比仿真实验表明,本研究所提出的改进遗传算法能较好地避免早熟,所得路径更平滑,收敛速度更快,综合性能相对更优。 |
英文摘要: |
Addressing the challenges of path unevenness and prematurity in mobile robot path planning using traditional genetic algorithms, an enhanced genetic algorithm was introduced. Initially, a terminal distance and included angle-based population initialization method was implemented to elevate the quality of the initial population. Subsequently, a novel fitness function tailored to path length and turning angle was developed to enhance path smoothness. An improved tournament selection strategy retaining inferior solutions was employed to bolster population diversity. Furthermore, a node quantity screening-based crossover operator was designed to expedite the convergence speed and minimize the invalid crossovers. Additionally, multiple mutation strategies were implemented to enhance the search capability in the algorithm's later stages. Extensive comparative simulation experiments reveal that this improved genetic algorithm avoids prematurity, generates smoother paths, and converges faster, and the comprehensive performance is relatively better. |
查看全文
查看/发表评论 下载PDF阅读器 |
关闭 |
|
|
|