摘要
提出了多功能室外智能移动机器人避障轨迹自动规划方法。利用遗传算法平行求解不同空间搜索特征,使其不趋于局部化,同时对接多个可行解,通过将路径的二维码变成一维码的简化方式,选择合适的路边约束、动态避障、路径最短适应度函数,确保其稳定收敛,并在算法中加入自适应调整方法,优化自适应参数,避免路径规划时搜索范围大、易陷入局部最小、收敛速度慢问题,提升机器人避障轨迹规划效果。实验结果表明:所提方法的种群平均适应值及最优个体进化次数少,在第56代时便可快速且清晰地找到避障规划最优路径。对比传统规划方法,所研究方法规划所耗时间短,可以提高移动机器人的作业效率。良好的避障功能,在实际的应用过程中,具有节省时间,提高工作效率。
The traditional programming method can not solve the problems of local minimum and slow convergence in the process of obstacle avoidance. An automatic obstacle avoidance trajectory planning method for multifunctional outdoor intelligent mobile robot was proposed. Genetic algorithm was used to solve different spatial search features in parallel, so that it does not tend to be localized, and multiple feasible solutions can be connected. By changing the two-dimensional code of the path into the simplified coding length of one-dimensional code, appropriate roadside constraints, dynamic obstacle avoidance and the shortest fitness function of the path were selected to ensure its stable convergence.An adaptive adjustment method was added to the algorithm, to optimize its adaptive parameters to avoid the problems of large search range, being easy to fall into local minimum and slow convergence speed in path planning, and to improve the effect of Obstacle Avoidance Trajectory Planning. The experimental results show that the proposed method has less population average fitness and optimal individual evolution times, and can quickly and clearly find the optimal path of obstacle avoidance planning in the 56 th generation. Compared with the traditional planning methods, the planning time of the proposed method is short, which can improve the operation efficiency of the mobile robot. It has good obstacle avoidance function. In the process of practical application, it is of great significance to save time and improve work efficiency.
作者
刘学文
任兴贵
许诺
徐定杰
LIU Xuewen;REN Xinggui;XU Nuo;XU Dingjie(Harbin Institute of Technology,Harbin 150000,China;Guangzhou Huaxia Vocational College,Guangzhou 510935,China)
出处
《兵器装备工程学报》
CAS
CSCD
北大核心
2022年第10期201-206,共6页
Journal of Ordnance Equipment Engineering
基金
国家自然科学基金项目(61573117)
广东省青年人才创新基金项目(2019GKQNCX092)。
关键词
多功能
机器人
智能避障
轨迹规划
适应度函数
遗传编码
multi function
robot
intelligent obstacle avoidance
trajectory planning
fitness function
genetic code
