多无人艇协同围捕智能逃跑目标方法研究

针对敌方船舶采用合围战术,研究我方无人艇（USV）被敌方船舶包围情况下的逃跑策略规划问题。

提出一种混合采样深度Q网络（HS-DQN）强化学习算法,逐步增加重要样本的回放频率,并保留一定的探索性,防止算法陷入局部最优。设计状态空间、动作空间和奖励函数,通过训练获得最优的USV逃跑策略,并从奖励值和逃脱成功率方面与DQN算法进行对比。

仿真结果表明,使用HS-DQN算法进行训练,逃脱成功率提高2%,算法的收敛速度提高了20%。

HS-DQN算法可以减少USV无效探索的次数,并加快算法的收敛速度,仿真实验验证了USV逃跑策略的有效性。

     

Path planning method for unmanned surface vehicle with controllable distance from obstacles北大核心CSCD

[目的]为了解决水面无人艇(USV)路径规划中安全性和平滑性方面的问题,提出一种与障碍物距离可控的USV路径规划方法。[方法]首先,结合雷达图像生成栅格化环境信息,利用维诺场算法(VFA)为每个栅格添加危险势场并建立航行界限;其次,建立与航行界限关联的危险度函数对A^(*)算法的评价函数进行改进,利用改进的A^(*)算法进行路径规划;最后,针对航行路径转向角较大的问题,采用梯度下降法(GDM)进行航行路径的平滑处理,得到满足USV实际航行要求的连续平滑路径。[结果]仿真结果表明,所提路径规划方法通过设置不同的航行界限可以实现路径与障碍物之间距离的控制且平滑性符合航行要求。[结论]该方法在USV路径规划过程中具有一定的合理性和有效性,可为USV自主避障决策提供参考。     

Cooperative path following control of UAV and USV cluster for maritime search and rescue北大核心CSCD

［Objectives ］ This paper studies a three-dimensional (3D) cooperative path-following control problem in the process of maritime search and rescue for a heterogeneous unmanned cluster system composed of unmanned aerial vehicles (UAVs) and unmanned surface vehicles (USVs).［Methods ］ First, kinematic models of the UAVs and USVs are established under a fixed coordinate system and body coordinate system. In order to design a 3D path-following controller suitable for motion control, an air coordinate system is established, and the path tracking error models of the UAVs and USVs are established in the Serret-Frenet coordinate system. Next, a 3D line-of-sight (LOS) guidance law is designed at the kinematic level, and a cooperative path-following control method suitable for heterogeneous clusters of marine vehicles is proposed, allowing the UAVs and USVs to track the preset parameterized path. Finally, the stability of the control system is analyzed based on the Lyapunov stability theory.［Results］The simulation results verify the effectiveness of the proposed cooperative path-following control method for heterogeneous clusters of marine vehicles.［Conclusions］The results of this study can provide references for maritime search and rescue by using the proposed cooperative path-following control method. © 2022 Journal of Clinical Hepatology. All rights reserved.