基于Q学习的无人机航迹规划
Path Planning of UAV Based on Q-learning
本文使用Q学习算法来解决实时无人机航迹规划问题. 最优的航迹规划问题实际上是根据几何问题对路线进行求解,Q学习本质上来说是一种增强学习算法, 不仅能够充分利用距离信息来计算基于几何距离信息的航迹, 同时也融合了复杂环境中的危险信息,从实用和理论角度都很好的解决了无人机航迹规划问题. 最后通过分析几种不同的仿真结果, 表明该方法的可行性及有效性。
In this paper, we used a classic learning algorithm, the Q-Learning, and its application on the UAV (unmanned aircraft vehicles) planning problem. The route planning problem is considered as a geometric problem in this paper and is solved by Q-learning. Essentially the Geometric learning is a kind of reinforce learning methods and is able to utilize the distance and danger information from the map at the same time which can lead to a comprehensive solution to the planning problem. At the end of paper, some representative results are shown to prove Q-Learning's effectiveness and feasibility.
毛治力、徐彬、张宝昌
航空航天技术自动化技术、自动化技术设备计算技术、计算机技术
无人机航迹规划Q学习
UAVpath planningQ-learning
毛治力,徐彬,张宝昌.基于Q学习的无人机航迹规划[EB/OL].(2012-03-29)[2025-07-25].http://www.paper.edu.cn/releasepaper/content/201203-807.点此复制
评论