数学算法
dckwin
世界最不著名程序员之一
展开
-
IROS2020:全局位置量测与基于优化的VIO紧耦合融合
原创 2020-12-03 16:41:16 · 457 阅读 · 0 评论 -
轻小型无人机路径规划算法学习(一)
自主无人机系统框图附上链接:https://www.bilibili.com/video/BV1W4411E7jq原创 2020-08-28 17:59:17 · 1381 阅读 · 0 评论 -
DDPG:基于深度强化学习的连续控制(ICLR 2016)
摘要我们将“深度Q学习(Deep Q-learning)”成功的基础思想应用于连续动作领域。我们基于确定性策略梯度(Deterministic Policy Gradient,DPG),提出了一种基于演员评论家(Actor-critic),无模型(model-free)的算法,该算法能应用在连续动作空间。我们进一步证明,对于许多任务,算法可以学习“端到端”策略:直接从原始像素输入。引言D...原创 2020-04-01 17:44:53 · 2796 阅读 · 1 评论 -
IROS2019多传感器融合SLAM方向
Stereo Visual Inertial LiDAR Simultaneous Localization and MappingAbstractSimultaneous Localization and Mapping (SLAM) is a fundamental task to mobile and aerial robotics. LiDAR based systems have ...原创 2020-02-17 17:36:09 · 3576 阅读 · 0 评论 -
Comparison of DRL Policies to Formal Methods for Moving Obstacle Avoidance
AbstractDeep Reinforcement Learning (RL) has recently emerged as a solution for moving obstacle avoidance. Deep RL learns to simultaneously predict obstacle motions and corresponding avoidance action...原创 2019-11-29 20:52:03 · 400 阅读 · 4 评论 -
Self-supervised DRL with Generalized Computation Graphs for Robot Navigation
AbstractLearning-based methods improve as the robot acts in the environment, but are difficult to deploy in the real-world due to their high sample complexity. To address the need to learn complex po...原创 2019-11-16 11:32:14 · 475 阅读 · 0 评论 -
Virtual-to-real DRL: Continuous Control of Mobile Robots for Mapless Navigation
AbstractWe present a learning-based mapless motion planner (基于学习方法的无地图运动规划器) by taking the sparse 10-dimensional range findings(稀疏的雷达scan) and the target position with respect to the mobile robot coo...原创 2019-11-13 14:01:43 · 1003 阅读 · 0 评论 -
FollowNet: Robot Navigation by Following Natural Language Directions with DRL
AbstractWe present FollowNet, an end-to-end differentiable neural architecture for learning multi-modal navigation policies. FollowNet maps natural language instructions as well as visual and depth ...原创 2019-11-12 17:07:53 · 387 阅读 · 0 评论 -
PX4飞控Avoidance功能包2016论文分享
PX4飞控Avoidance功能包2016论文基于风险的无人机路径寻找摘要简介相关工作风险最安全路径结合距离和风险风险有界的最短路径在部分可观测的环境中的路径寻找路径寻找最短路径Dijkstra算法A*算法最安全路径转换成Dijkstra算法最优路径选择代价函数概率 vs 期望值联合边界、期望作为近似探索的风险最平滑路径动态搜索算法双向搜索仿真环境软件组成ROSGazeboPX4OctoMap交互...原创 2019-05-11 19:42:49 · 2030 阅读 · 1 评论 -
PX4飞控Avoidance功能包2018论文分享
PX4飞控Avoidance功能包2018论文分享PX4官方Avoidance-2018论文引言(略)摘要简介相关工作全局障碍规避局部障碍规避对本文启发方法$3DVFH$算法建立记忆将直方图重投影到3D点从3D点建立直方图结合记忆直方图和新直方图适应安全和拐角情况在FOV内部方向运动安全措施:后退直方图中的安全裕度取决于障碍距离取决于进程的代价参数自适应球体避障-离障碍的最小距离地面检测建立内部高...原创 2019-05-13 23:32:51 · 3973 阅读 · 3 评论 -
结合无人机室内导航谈谈3DVFH+算法
无人机室内导航路径规划——障碍规避在浏览PX4开发手册过程中看见里面提到的一种障碍规避算法——3DVFH+。该算法在之前做本科毕设时候有所接触,不过当时是进行了2D的Matlab仿真,down个代码改改障碍物形状,飞机当做质点考虑,没有深入了解这个算法在实际中该怎么应用。今天找到了论文的原文《3DVFH+:Real-Time Three_Dimensional Obstacle Avoidanc...原创 2019-03-18 17:30:41 · 3442 阅读 · 1 评论 -
三次轨迹多项式在无人机上实现的一点小想法
无人机路径规划中的三次轨迹多项式IntroductionMethodoctree-based可行轨迹连接飞行长廊约束下的轨迹生成最近实验涉及到了无人机的室内飞行,在调试飞机的同时,对路径规划算法的文献进行了一些查阅。Shen的一篇文章中提到了一些无人机室内感知与路径规划算法。文章发表在2016年的ICRA,名为"online generation of collision-free trajec...原创 2019-03-09 19:41:40 · 855 阅读 · 0 评论