基于约束多目标优化的多区域无人机路径规划
Multi-Region UAV Path Planning Based on Constrained Multi-Objective Optimization
张猜 1黄林 1彭超达 1崔金荣1
作者信息
- 1. 华南农业大学数学与信息学院,广东 广州 510642
- 折叠
摘要
无人机路径规划常以约束多目标优化问题形式建立数学模型,且这些模型几乎只考虑三维建模空间中两点之间的路径规划;然而缺少无人机从起点出发,经过指定的多个作业区域并到达终点的路径规划问题研究.针对上述问题,基于不可行解引导种群进化提出带有地理信息指导搜索策略的约束多目标进化算法(DW-LS).首先,根据无人机飞行所受环境约束、性能约束和访问多区域任务建立新的约束多目标优化模型;其次,基于不可行解利用机制,设计地理信息指导搜索策略优化不可行解以进一步协助种群搜索最优进化方向;最后,通过仿真对比了DW-LS与具有代表性的三个约束多目标进化算法,实验结果表明DW-LS所得可行解具有更好的收敛性和多样性.
Abstract
Drone path planning often establishes mathematical models in the form of constrained multi-objective optimization problems,and these models almost only consider path planning between two points in the three-dimen-sional modeling space;However,there is a lack of research on the path planning problem of unmanned aerial vehicles starting from the starting point,passing through multiple designated work areas,and reaching the endpoint.To remedy this issue,this paper proposes a constrained multi-objective evolutionary algorithm with a geographic information-guided search strategy based on infeasible solution-guided population evolution(DW-LS).Firstly,a new constrained multi-objective UAV path planning problem was built by combining environmental constraints,performance con-straints,and multiple region visit tasks of UAV flight.Secondly,by having the utilization mechanism of the infeasible solution,a geographic information-guided search strategy was designed to optimize the infeasible solution to further assist the population in searching for the optimal evolutionary direction.Finally,DW-LS was compared with three rep-resentative constrained multi-objective evolutionary algorithms,and the experimental results show that the performance of our algorithm is superior in terms of convergence and diversity of the obtained feasible solutions.
关键词
进化算法/约束多目标优化/无人机路径规划Key words
Evolutionary algorithm/Constrained multi-objective optimization/UAV path planning引用本文复制引用
基金项目
国家自然科学基金青年项目(62202177)
广州市科技计划项目(202201010576)
出版年
2024
NETL
NSTL
万方数据