QD-Gym Environments
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/rwickman/NEAT_RL
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了模拟的机器人环境，包括QDWalker、QDHalfCheetah、QDAnt和QDHopper等，在这些环境中，机器人学习节能的行走行为。每个机器人的状态由多个参数定义，包括重心高度和关节位置，而动作则表现为连续值的扭矩。该数据集涵盖了多个环境，任务是对机器人进行模拟以及技能发现。
This dataset contains simulated robotic environments, including QDWalker, QDHalfCheetah, QDAnt, QDHopper, and others. In these environments, robots learn energy-efficient walking behaviors. The state of each robot is defined by multiple parameters, such as center-of-mass height and joint positions, while actions are represented as continuous-valued torques. This dataset covers multiple environments, with the tasks focusing on robot simulation and skill discovery.
提供机构：
QD-Gym
