| 张泽宇,王雷,寿林,夏强强.改进遗传算法在路径规划中的应用研究[J].井冈山大学自然版,2025,(5):80-90 |
| 改进遗传算法在路径规划中的应用研究 |
| THE APPLICATION RESEARCH OF IMPROVED GENETIC ALGORITHM IN PATH PLANNING |
| 投稿时间：2024-12-15 修订日期：2025-03-03 |
| DOI：10.3969/j.issn.1674-8085.2025.05.010 |
| 中文关键词: 路径规划 遗传算法 自适应交叉策略 自适应变异策略 模拟退火算法 |
| 英文关键词: path planning genetic algorithm adaptive crossover strategy adaptive mutation strategy simulated annealing algorithm |
| 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(2023AH050935); 安徽省机器视觉检测与感知重点实验室开放基金项目(KLMVI-2024-HIT-15) |
|
| 摘要点击次数: 81 |
| 全文下载次数: 90 |
| 中文摘要: |
| 针对传统遗传算法在路径规划中存在偏转次数过多、收敛速度慢、易陷入局部最优等问题，提出一种新的改进遗传算法。通过中位过渡点将两条路径首尾相连为一条路径的方法来改进种群初始化，生成优秀初始种群以提高前期搜索效率；采用改进锦标赛选择策略，结合模拟退火方法防止陷入局部最优并提高算法搜索能力；设计自适应交叉和变异概率函数，提高其收敛速度和种群多样性；改进多点交叉策略和多种变异策略，提高路径规划求解的质量和稳定性。路径规划仿真结果表明，相比传统遗传算法、改进领航跟随遗传算法、改进自适应遗传算法、多种群自适应蚁群算法、改进灾变遗传算法，本研究所提出的改进遗传算法能够提高收敛速度、减少路径偏转次数和长度，从而搜索到更优路径。 |
| 英文摘要: |
| Aiming at the problems of excessive path deviation, slow convergence speed, and susceptibility to local optima in traditional genetic algorithms for path planning, an improved genetic algorithm is proposed in this paper.By initializing an improved population by connecting two paths end-to-end through a median transition point, an excellent initial population is generated to improve the search efficiency in the early stages. Adopting an improved tournament selection strategy and proposing a simulated annealing method is done to prevent falling into local optima and improve search capability. Adaptive crossover and mutation probability functions are designed to enhance convergence speed and population diversity. Additionally, improved multi-point crossover and multiple mutation strategies are employed to improve the quality and stability of path planning solutions. Simulation results demonstrate that, compared with the traditional genetic algorithms, the improved guided-following genetic algorithm, improved adaptive genetic algorithm, multi-population adaptive ant algorithm, and improved catastrophic mutation genetic algorithm, the proposed improved genetic algorithm can enhance the convergence speed, reduce the number of path deflections, and path length, so as to search the more optimal paths. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
