Isaac Sim Simulator Data

该数据集由使用Isaac Sim模拟器进行的实验组成，研究对象为两款机器人模型：KYON和Unitree Go2-W。研究重点在于动态奖励塑造以及轮腿式机器人在跌倒后的恢复能力。数据集涵盖了包括恢复成功率、关节扭矩消耗以及在不同配置和随机扰动下动作序列变化等性能指标。该数据集对2048种随机生成的初始条件进行了测试，旨在评估轮腿式机器人在采用不同配置和扰动情况下跌倒后的恢复成功率。

This dataset comprises experiments conducted using the Isaac Sim simulator, with two robot models as the research subjects: KYON and Unitree Go2-W. The research focuses on dynamic reward shaping and the post-fall recovery capability of wheel-legged robots. The dataset covers performance metrics including recovery success rate, joint torque consumption, and changes in action sequences under different configurations and random disturbances. A total of 2048 randomly generated initial conditions were tested, aiming to evaluate the fall recovery success rate of wheel-legged robots under various configurations and disturbance conditions.