2D Particle Motion Dataset

该数据集包含了一个2D粒子在一个布满障碍物的环境中的移动情况，旨在从任意起点无碰撞地到达距离目标点指定距离的位置。此外，该数据集用于评估拟合SGTDP算法，其中包含了使用K=5的K近邻函数近似方法。该数据集的规模为125K个元组，每个元组包含状态s、动作u、成本c和下一个状态s′。其任务是进行强化学习以实现轨迹规划。

This dataset comprises trajectory samples of a 2D particle moving in an obstacle-cluttered environment, with the objective of traveling collision-free from any arbitrary starting point to a position at a specified distance from the target. Furthermore, this dataset is utilized to evaluate the fitted SGTDP algorithm, which adopts the K-nearest neighbor function approximation method with K=5. The dataset contains a total of 125K tuples, where each tuple encompasses the state s, action u, cost c, and next state s′. The downstream task of this dataset is to perform reinforcement learning for trajectory planning.