摘要
针对传统遗传算法在路径规划中存在收敛速度慢和易陷入局部最优等缺陷,提出一种融合Bezier遗传算法.首先,对传统遗传算法进行改进,采用启发式中值插入法建立初始种群,基于路径长度、路径安全性和路径能耗3个指标生成多目标适应度函数,再分别利用分层法、单点交叉法和八邻域单点变异法设计选择、交叉、变异算子;其次,引入Bezier曲线的概念,以改进遗传算法运行得出的路径序列点作为Bezier曲线的控制点生成最优曲线路径,确保移动机器人的高效运行;最后,在栅格环境下进行仿真实验.结果表明,该算法是有效的,且在路径长度和运行时间方面具有显著优势.
Aiming at the shortcomings of traditional genetic algorithm,such as slow convergence speed and easy to fall into local optimization,a fusion Bezier-genetic algorithm is proposed.Firstly,the traditional genetic algorithm is improved by using heuristic median insertion method to establish the initial population,and multi-objective fitness function is generated based on path length,path security and path energy consumption.The selection,crossover and mutation operators are designed by using hierarchical method,single point crossover method and eight neighborhood single point mutation method.Then the concept of Bezier curve is introduced,and the path sequence points obtained from the operation of improved genetic algorithm are used as the control points of Bezier curve to generate the optimal curve path,which ensures the efficient operation of mobile robot.Finally,a simulation experiment is carried out in a grid environment.The results show that the algorithm is effective and has significant advantages in terms of path length and running time.
作者
李开荣
胡倩倩
LI Kairong;HU Qianqian(School of Information Engineering,Yangzhou University,Yangzhou 225127,China)
出处
《扬州大学学报（自然科学版）》
CAS
北大核心
2021年第5期58-64,共7页
Journal of Yangzhou University：Natural Science Edition
基金
国家自然科学基金资助项目(61872313)
江苏省应急管理科技资助项目(YJGL-YF-2020-17).
