Download PDFOpen PDF in browser
EN
The title and the abstract of this preprint are also available
in English

航母甲板上舰载机的混合避障和仿真

EasyChair Preprint 5372

11 pagesDate: April 25, 2021

Abstract

航母甲板上舰载机的混合避障是路径规划中一个重要的问题. 现有方法在处理动态障碍物时存在效果不稳定且决策空间离散的问题. 为了解决这些问题, 我们提出了一种结合模型预测控制(MPC)与深度确定性策略梯度(DDPG)的新算法. 首先, 用MPC算法预测动态障碍物的短期轨迹. 其次, 根据MPC预测的动态障碍物的短期轨迹, DDPG为机器人提供在连续空间里学习和决策行为的能力. 最后, 人工势场的概念被引入来设置奖励函数, 以提高收敛速度和准确率. 我们使用Unity 3D在高动态复杂场景中进行仿真实验, 结果表明, 我们的方法与其他方法相比, 在精度上提高了7%-30%. 且与DQN(Deep Q Network)相比, 在路径长度和转弯角度上分别减少了100个单位和400-450度.

Keyphrases: MPC, 人工势场, 路径规划, 避障

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@booklet{EasyChair:5372,
  author    = {Xue Junxiao and Kong Xiangyan and Dong Bowei and Guo Yibo and Xu Mingliang},
  title     = {Hybrid Obstacle Avoidance and Simulation of Carrier-Based Aircraft on the Deck of an Aircraft Carrier},
  howpublished = {EasyChair Preprint 5372},
  year      = {EasyChair, 2021}}
Download PDFOpen PDF in browser