robot_sim配置

本文档详细介绍了robot_sim的配置过程,包括基于深度学习和几何方法的抓取实验。遇到的问题如banana不显示、abort问题、geometry_grasp图像加载不出来等,通过参考github issue和csdn博客找到了解决方案。成功执行roslaunch和rosrun命令,但有时会出现路径规划问题导致抓取失败。验证gpd安装成功的方法和一些额外的参考资料也在文中提及。
Python3.8

Python 是一种高级、解释型、通用的编程语言,以其简洁易读的语法而闻名,适用于广泛的应用,包括Web开发、数据分析、人工智能和自动化脚本

robot_sim
https://github.com/Suyixiu/robot_sim 的issue
https://www.bilibili.com/video/BV19f4y1h73E/ 下评论
文件来源
在这里插入图片描述

其余克隆

基于几何的抓取

banana不显示
在这里插入图片描述

解决参考
https://github.com/Suyixiu/robot_sim/issues/2 (项目配置主要参考)

abort问题解决同样参考该网址
在这里插入图片描述

如果还无再试https://blog.csdn.net/qq_44642372/article/details/123935071
来自上github网址

双击第一个单选框无反应
解决参考https://blog.csdn.net/qq_38620941/article/details/125321347

(里面所有链接所涉及的安装配置基本都试了)

下评论
在这里插入图片描述

几何抓取需要改得应该只有main和recongnize

在这里插入图片描述

基于几何学习的抓取实验
roslaunch robot_sim grasp_world.launch

rosrun robot_sim geometric_method_grasp

对wooden_peg ERROR
https://blog.csdn.net/qq_38620941/article/details/125321347

不要忘了把循环下次数改为4

在这里插入图片描述

catkin_make 生效!

点击两下第一个按钮自动执行,成功!

现有问题geometry_grasp图相加载不出来
c++标定已做,python未
深度图配准失败

基于深度学习的抓取

关于编译
在这里插入图片描述

问题,将ROS_ERROR(“lookupTransform world to yixiuge_ee_link faild”, ex.what()

改为ROS_ERROR(“lookupTransform world to yixiuge_ee_link failed: %s”, ex.what());

参考gpt

在这里插入图片描述

优化代码部分为(可能没什么用)

tf::TransformListener listener;
tf::StampedTransform world2tool;

ros::Duration timeout(1.0); // 设置等待超时时间

try {
    listener.waitForTransform("/world", "/yixiuge_ee_link", ros::Time(0), timeout);
    listener.lookupTransform("/world", "/yixiuge_ee_link", ros::Time(0), world2tool);

    printf("world2tool %f  %f  %f  %f  %f  %f %f\n", 
        world2tool.getOrigin().x(), world2tool.getOrigin().y(), world2tool.getOrigin().z(),
        world2tool.getRotation().getX(), world2tool.getRotation().getY(), world2tool.getRotation().getZ(), world2tool.getRotation().getW());
} catch (tf::TransformException &ex) {
    ROS_ERROR("%s", ex.what());
    ROS_ERROR("lookupTransform world to yixiuge_ee_link failed", ex.what());
}
roslaunch robot_sim grasp_world.launch
roslaunch robot_sim gpd_run.launch type:=2 topic:=/cloud_sample
rosrun robot_sim GPD_method_grasp
没有进conda环境,开了梯子,用vscode终端运行的

在这里插入图片描述

参考https://blog.csdn.net/qq_38620941/article/details/125321347

在这里插入图片描述

更改gpd_run.launch 中的config_file为ros_eigen_params.cfg ,不改会出现world问题

在这里插入图片描述

更改里面的参数按照eigen_param.cfg ,每一个相同的参数,

更改ros_eigen_params.cfg中的weights_file路径为你的gpd中的

参考https://github.com/atenpas/gpd_ros/issues/12

在这里插入图片描述

在新弹出的Normal框中按q,再新弹出的grasp candidates中按q,在selected grasps中按q,看到rosrun robot_sim GPD_method_grasp这个窗口出现receive grasp,就可以按照视频过程点击两下good grasp实现抓取!!!!!!!!

依然会出现

在这里插入图片描述

有时候可能是路径规划有问题,有动作但没有抓取成功

验证你的gpd是否安装成功,

在这里插入图片描述

在新弹出的Normal框中按q,再新弹出的grasp candidates中按q,在selected grasps中按q,结束,不报错。

相关链接,及一些可能没有用的步骤

https://blog.csdn.net/qq_44876051/article/details/112370423

https://blog.csdn.net/qq_33328642/article/details/122267004

sudo -H pip3 install -U transforms3d(参考https://blog.csdn.net/qq_38620941/article/details/125321347下评论)

在这里插入图片描述

在这里插入图片描述

您可能感兴趣的与本文相关的镜像

Python3.8

Python3.8

Conda
Python

Python 是一种高级、解释型、通用的编程语言,以其简洁易读的语法而闻名,适用于广泛的应用,包括Web开发、数据分析、人工智能和自动化脚本

The MRSim (Multi-Robot Simulator) is an extension of the Autonomous mobile robotics toolbox SIMROBOT (SIMulated ROBOTs) created for MatLab 5 in 2001. MRSim allows the user to simulate the behavior of multiple mobile robots in virtual environments. When compared to its predecessor SIMROBOT, MRSim presents two key contributions: 1) It is fitted to the new MatLab versions - Previously, users were unable to work with SIMROBOT functions since most of them were incompatible with the new MatLab versions. MatLab significantly evolved over the last 10 years, making SIMROBOT obsolete. In this extension, all features of SIMROBOT were updated and improved based on the new MatLab functions; 2)It is also fitted to suit multi-robot applications - although SIMROBOT was endowed with various interesting features for mobile robotics, it presented several limitations for multi-robot applications. Therefore, MRSim was created primarily to allow users to develop multi-robot applications, which would benefit working with some specific requirements such as multi-hop communication. Moreover, most of the functionalities in MRSim have an integrated help (which can be accessed just by typing help function) that allows to easily understand the dynamics of how to create and run simulations. In sum, just like SIMROBOT, each robot in MRSim can be equipped with several virtual sensors and can be driven by its own control algorithm. The toolbox includes two independent applications. The first one is the EDITOR (simedit), which allows the user to create and modify the virtual environment,to edit the control algorithms of robots, to load and save simulation, and others. The second application, SIMULATOR (simview), can be run directly from the EDITOR or separately from MatLab Command Window and serves as a simulation viewer. A MatLab help file is being currently created and will be add in the future.
评论 2
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包

打赏作者

灞気水笵er

你的鼓励将是我创作的最大动力

¥1 ¥2 ¥4 ¥6 ¥10 ¥20
扫码支付:¥1
获取中
扫码支付

您的余额不足,请更换扫码支付或充值

打赏作者

实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值