河北水利电力学院学报

河北水利电力学院学报 ›› 2025, Vol. 35 ›› Issue (2): 28-34.DOI: 10.16046/j.cnki.issn2096-5680.2025.02.006

• 人工智能与机器人 • 上一篇    下一篇

基于正交试验法的机器人最优路径问题研究

李海涛1, 张鹏程2,3,4, 张铁壁2,3,4, 王文成2,3,4   

  1. 1.河北科技师范学院机电工程学院,河北省秦皇岛市海港区河北大街360号 066004;
    2.河北水利电力学院机械工程系,河北省沧州市黄河西路49号 061001;
    3.河北省工业机械手控制与可靠性技术创新中心,河北省沧州市黄河西路49号 061001;
    4.沧州市工业机械手控制与可靠性技术创新中心,河北省沧州市黄河西路49号 061001
  • 收稿日期:2024-09-11 修回日期:2024-11-21 出版日期:2025-06-30 发布日期:2025-07-04
  • 通讯作者: 张鹏程(1976-),男,河北泊头人,硕士,高级实验师,研究方向为计算机辅助设计、智能布局。E-mail:zpc-rjy@163.com
  • 作者简介:李海涛(2000-),男,山东日照人,硕士研究生在读,研究方向为机器人的路径规划。E-mail:614667877@qq.com
  • 基金资助:
    河北省教育厅科学研究项目资助(ZD2022081),河北省高等学校基本科研业务费研究项目(SYKY2403)

Research on the Optimal Path Problem of Robot Based on Orthogonal Test Method

LI Haitao1, ZHANG Pengcheng2,3,4, ZHANG Tiebi2,3,4, WANG Wencheng2,3,4   

  1. 1. Collage of Mechanical and Electrical Engineering, Hebei Normal University of Science & Technology, 066004, Qinhuangdao, Hebei, China;
    2. Department of Mechanical Engineering, Hebei University of Water Resources and Electric Engineering, 061001, Cangzhou, Hebei, China;
    3. Hebei Industrial Manipulator Control and Reliability Technology Innovation Center, 061001, Cangzhou, Hebei, China;
    4. Cangzhou Industrial Manipulator Control and Reliability Technology Innovation Center, 061001, Cangzhou, Hebei, China
  • Received:2024-09-11 Revised:2024-11-21 Online:2025-06-30 Published:2025-07-04

PDF

摘要： 为了探究机器人运动过程中的自主路径规划问题,以TSP问题为模型,在遗传算法的基础上,采用5因素4水平正交试验法,通过极差、方差分析不同初始化参数对于最终路径结果的影响。通过实例验证,结果表明:种群数量和迭代次数的取值对应的路径结果具有较大的偏差平方和,对应较大的均方,对结果的影响显著,而其他参数的影响较小。各参数对于优化结果的影响从大到小的顺序为:迭代次数、种群数量、选择因子、变异概率,交叉概率。利用正交试验法对遗传算法的参数进行优化,可以避免单一参数选优时不同参数之间耦合效应的影响,从而找到一组适应不同类型和规模问题的最优化参数组合,利用该参数初始化遗传算法求解此类问题,可以迅速获得最优解。

关键词: 机器人路径规划, 遗传算法, MATLAB, 正交试验

Abstract: To explore the autonomous path planning problem in the process of robot motion.Taking TSP problem as the model, on the basis of genetic algorithm, the orthogonal test method of 5 factors and 4 levels was used to explore the influence of different initialization parameters on the final path results through range and variance analysis. Through the example verification, the results show that the path results corresponding to the number of populations and the number of iterations have a large sum of squared deviations, corresponding to a large mean square. They have a significant impact on the results, while other parameters have less impact. The order of the influence of each parameter on the optimization results from large to small is: the number of iterations, the number of populations, the selection factor, the mutation probability, and the crossover probability. Using the orthogonal test method to optimize the parameters of the genetic algorithm can avoid the influence of the coupling effect between different parameters when the single parameter is optimized, so as to find a set of optimal parameter combinations suitable for different types and scales of problems. Using this parameter to initialize the genetic algorithm to solve such problems can quickly obtain the optimal solution.

Key words: robot path planning, genetic algorithm, MATLAB, orthogonal test

中图分类号: