论文标题
具有新颖的动态全球路径计划方法的设计和开发的面向研究的低成本机器人平台
Design and Development of a Research Oriented Low Cost Robotics Platform with a Novel Dynamic Global Path Planning Approach
论文作者
论文摘要
基于计算机视觉传感器的自主导航系统通常需要非常昂贵的复杂机器人平台。这为在机器人应用程序中至关重要的复杂本地化，映射和导航算法的实施和测试构成了障碍。在这项工作中，比较了机器人操作系统（ROS）支持的移动机器人平台，并根据低成本的教育机器人平台的自主导航系统的端到端实施，同时介绍了Intel realsense D435相机，同时介绍了Alphabot2。此外，提出了一种新颖的方法，用于实施动态路径计划者作为ROS框架中的全球路径计划者。我们评估了这种方法的性能，并突出了通过动态的全球路径计划者可以实现的改进。这种低成本修改的Alphabot2机器人平台以及拟议的动态全球路径计划方法将对研究人员和学生获得基于计算机视觉的导航系统的动手体验。
Autonomous navigation systems based on computer vision sensors often require sophisticated robotics platforms which are very expensive. This poses a barrier for the implementation and testing of complex localization, mapping, and navigation algorithms that are vital in robotics applications. Addressing this issue, in this work, Robot Operating System (ROS) supported mobile robotics platforms are compared and an end-to-end implementation of an autonomous navigation system based on a low-cost educational robotics platform, AlphaBot2 is presented, while integrating the Intel RealSense D435 camera. Furthermore, a novel approach to implement dynamic path planners as global path planners in the ROS framework is presented. We evaluate the performance of this approach and highlight the improvements that could be achieved through a dynamic global path planner. This low-cost modified AlphaBot2 robotics platform along with the proposed dynamic global path planning approach will be useful for researchers and students for getting hands-on experience with computer vision-based navigation systems.
