Omni-Directional Locomotion Task

该数据集是对全方位移动任务的一个改编版本，其中一款拥有18个自由度的六足机器人从初始位置开始学习向各个方向行走。机器人采用经进化调整的正弦控制信号参数进行移动，并且在无重置设置下，评估是依次进行的。规模上，每个运行周期对单一机器人进行10,000次评估。任务目标是让机器人在开放环境中有自主地学习行走。

This dataset is an adapted variant of the omnidirectional mobile locomotion task. It involves a hexapod robot with 18 degrees of freedom (DoF) learning to walk in all directions starting from its initial position. The robot utilizes evolutionarily tuned sine control signal parameters for locomotion, and all evaluations are conducted sequentially without any reset settings. Each run cycle carries out 10,000 evaluations on a single robot. The core goal of this task is to enable the robot to autonomously learn to walk in an open environment.