改进人工势场法解决局部最小值路径规划研究  

Improving the APF to Solve Local Minima Research on Path Planning

作　　者：闫为佳 杨旗[1] 黄星卓 周杨 YAN Weijia;YANG Qi;HUANG Xingzhuo;ZHOU Yang(School of Mechanical Engineering,Shenyang Ligong University,Shenyang 110000,China)

机构地区：[1]沈阳理工大学机械工程学院,沈阳110000

出　　处：《组合机床与自动化加工技术》2024年第5期36-39,共4页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家自然科学基金项目(61971118);沈阳理工大学科研创新团队建设计划资助项目(SYLUTD202106)。

摘　　要：当移动机器人在行进过程中使用传统人工势场法(artificial potential field method, APF)进行路径规划时,通常会陷入局部最优困境,无法顺利到达目标点。为解决这一问题,首先,对APF算法规划路径失败原因进行分析,其次设置情况判断条件,判断当机器人陷入局部最小值时,通过在合适位置增加临时引导点的方法,引导其跳出局部极小值点;其次,引入分数阶微积分思想方法结合APF算法,提出混合阶次的分数阶梯度下降法进行位置信息迭代,优化算法的收敛速度和收敛精度;最后,用MATLAB软件对该方法进行仿真,实验结果表明提出的方法可以有效解决局部最小值问题,而且在速度、精度上都有明显的提高,且能适应较为复杂的多障碍物环境,验证了改进方法的有效性、正确性。When mobile robots use the traditional APF for path planning during their journey,they often fall into a local optimal dilemma and cannot reach the target point smoothly.To solve this problem,first analyze the reasons for the failure of the APF to plan the path,and then set the condition judgment condition to determine that when the robot falls into the local minimum value.By adding temporary guidance points at appropriate position,it is guided to jump out of the local minimum point.Then bring in fractional calculus thought method combined with the APF,propose mixed order fractional gradient descent method for position information iteration,and optimize the rate of convergence and convergence accuracy of the algorithm.Finally,the method was simulated using MATLAB software,and the experimental results showed that the proposed method can effectively solve the local minimum problem,and has significant improvements in speed and accuracy.It can also adapt to complex multi obstacle environments,verifying the effectiveness and correctness of the improved method.

关 键 词：分数阶 人工势场法 局部最小值 路径规划 MATLAB 

分 类 号：TH112[机械工程—机械设计及理论] TG659[金属学及工艺—金属切削加工及机床]