Point Mass Navigation Environment

该数据集旨在创建一个二维世界环境，其中智能体控制一个点质量物体，需要穿越特定大小和位置的闸门，以到达期望的目标。此外，该环境空间可以改变闸门的宽度和位置以及动态摩擦系数，为强化学习提供了多样化的任务。任务是通过强化学习实现导航与目标到达。

This dataset is designed to construct a 2D world environment, where an AI Agent controls a point-mass object and is required to pass through gates with specific sizes and positions to reach a predefined target. Furthermore, the environment allows adjustment of parameters including gate width, gate position, and dynamic friction coefficient, providing diversified tasks for reinforcement learning. The core task of this environment is to achieve navigation and target arrival via reinforcement learning.