一、项目地址
git clone https://gitee.com/mrqiguan/lerobot_so100_sim.git

二、环境配置
conda create -n lerobot_so100_sim --clone lerobot # 在lerobot环境里安装下面的包即可
conda activate lerobot_so100_sim
pip install rl_zoo3
pip install git+https://github.com/perezjln/gym-lowcostrobot.git
三、仿真操控
conda activate lerobot_so100_sim
sudo chmod 777 /dev/ttyACM0
1、在mujoco里控制so100模型
python 1_so100_feetech_mujoco_control_sim.py
2、利用实机控制mujoco的so100模型
python 2_so100_feetech_mujoco_control_realTosim.py
注意:根据实机,修改2_so100_feetech_mujoco_control_realTosim.py,改变机械臂当前运动状态,观察符号(转向)和偏移值(置0),修改参数直到和初始位置符合
initial post [0 -3.14 3.14 0.817 0 -0.157]
Adjust the plus-minus sign and the offsets
target_pos[i] = ± target_pos[i] + 0
target_pos[0] = -target_pos[0] - 0.838 # =0
target_pos[1] = target_pos[1] - 1.81 # =-3.14
target_pos[2] = target_pos[2] + 1.1 # =3.14
target_pos[3] = target_pos[3] - 1.413 # =0.817
target_pos[4] = target_pos[4] - 2.23 # =0
target_pos[5] = target_pos[5] - 0.626 # =-0.157
四、参考链接
jpata/gym-so100 at integrationhttps://github.com/jpata/gym-so100/tree/integrationmichel-aractingi/Sim-LeRobotHackathon: This simulation environment that will be used for the LeRobot Hackathon 2024.
https://github.com/michel-aractingi/Sim-LeRobotHackathon?tab=readme-ov-file