Crocoddyl Optimal Control Problems

该数据集包含了一系列最优控制问题，涵盖了从欠驱动双摆到四旋翼飞行器导航场景，再到四足机器人步态以及人形机器人操作任务等多样化内容。此外，数据集还展示了四足机器人和人形机器人在跳跃、蹦跳和引体向上等任务上的解决成果，这些成果彰显了BOX-FDDP算法的有效性。该数据集的任务主要针对腿部机器人的控制问题以及操作任务。

This dataset comprises a series of optimal control problems, covering a diverse set of scenarios including underactuated double pendulums, quadrotor navigation tasks, quadruped robot gait generation tasks and humanoid robot manipulation tasks. Furthermore, the dataset also displays the successful solutions of quadruped and humanoid robots for tasks such as jumping, bounding and pull-ups, which demonstrates the effectiveness of the BOX-FDDP algorithm. The primary tasks of this dataset center on legged robot control problems and manipulation tasks.