# step 1 安装Isaac Sim ## step 1-1: 创建虚拟环境conda create-nenv_isaaclabpython3.11conda activate env_isaaclab pipinstall--upgradepip# step 1-2: 安装依赖项# 安装isaac simpipinstallisaacsim[all,extscache]5.1.0--extra-index-url https://pypi.nvidia.com# 安装pytorchpipinstall-Utorch2.7.0torchvision0.22.0 --index-url https://download.pytorch.org/whl/cu128# step 1-3: 验证是否安装成功isaacsim# step 2 安装Isaac Lab ## step 2-1: 下载源码gitclone gitgithub.com:isaac-sim/IsaacLab.git# 如果需要安装其他版本请执行如下指令以v.2.3.2为例cdIsaacLabgitcheckout f73c331738# step 2-2: 安装系统依赖sudoaptinstallcmake build-essential# step 2-3: 运行Isaac Lab安装脚本cdIsaacLab ./isaaclab.sh--install# 该命令会安装rl_gamesrsl_rlsb3skrlrobomimic# step 2-4: 验证安装# 方法1: 使用 isaaclab.sh 脚本./isaaclab.sh-pscripts/tutorials/00_sim/create_empty.py# 方法2: 直接使用 Pythonpython scripts/tutorials/00_sim/create_empty.py# 成功标志 启动一个黑色视口的模拟器窗口按 CtrlC 退出# step 2-5: 训练机器人# 训练 Ant 机器人./isaaclab.sh-pscripts/reinforcement_learning/rsl_rl/train.py--taskIsaac-Ant-v0--headless# 训练好之后查看训练结果./isaaclab.sh-pscripts/reinforcement_learning/rsl_rl/play.py--taskIsaac-Ant-v0# 训练四足机器人./isaaclab.sh-pscripts/reinforcement_learning/rsl_rl/train.py--taskIsaac-Velocity-Rough-Anymal-C-v0--headless# 训练好之后查看训练结果./isaaclab.sh-pscripts/reinforcement_learning/rsl_rl/play.py--taskIsaac-Velocity-Rough-Anymal-C-v0