engleski

Localization, Motion Planning and Control of Mobile Robots in Intelligent Spaces

This thesis investigates application of mobile robots in environments provided with ambient intelligence to form the so-called Intelligent Spaces (iSpaces). The purpose of the iSpace is providing various services to its users, where mobile robots provide numerous possibilities, such as load delivery, visitor guidance etc. Thus the focus of the thesis is set to developing the capability of the space to fully utilize mobile robots, with the emphasis on development of low-level algorithms that can benefit by using intelligent space advantages, such as precise localization and mapping and increased computational power. In this way a method for fast and precise mobile robot localization using distributed cameras is developed. The developed algorithm enables high framerates and at the same time high measurement precision and accuracy as it works in subpixel precision. The algorithm was tested in multiple experiments and performed with very good results. Second, a fast and flexible robot mobile robot motion planning method appropriate for application in iSpaces is developed. The developed motion planner is decoupled and consists of four stages: path planning, path smoothing, trajectory planning and trajectory tracking. The method can replan the path in real time, which enables planning in dynamic environments.

Intelligent Spaces; mobile robotics; global vision; motion planning; trajectory tracking

nije evidentirano