基于分层思想的无人飞行器航路规划研究

Unmanned Aerial Vehicle Path Planning Based on Hierarchical Thinking

在无人飞行器航路规划问题的研究中,为提高航路规划的效率和精度,针对传统遗传算法收敛速度慢、易陷入局部最优、寻优精度较差的问题,提出了一种分层思想的解决方法。首先用链接图法描述规划环境,通过采用Dijkstra算法寻找初始最优航路,并利用航路编码技术对初始航路进行优化;然后在已有的研究成果上,提出一种集混沌优化、模拟退火、遗传算法为一体的改进遗传算法(CGASA),在解决多目标多约束优化问题时取得了较好的结果;最后综合考虑飞行器的机动性能、威胁因素、飞越目标进入角度等代价的选取,利用改进遗传算法调整导航点的位置得出了满足性能要求的航路。

In the research on route planning of unmanned aerial vechicle, the basic algorithm may always have problelms such as falling in local optimum, slow convergence speed leading to low efficiency. In this paper, the hierarchical thinking was considered in unmanned aerial vehicle route planning. Firstly, the planning environment was described with link graph, and Dijkstra algorithm was used to find the initial optimal route; then on the findings of other researchers, an improved genetic algorithm（CGASA） was advocated; at last, considering the selection costs of the motor performance of the aircraft, threats, attacks entry angle and so on, the improved genetic algorithm was used to adjust the waypoint location to come to meet the performance requirements of the path.