Abstract:

Service Robots are built for making people more convenient and more comfortable. Robots provide services to people wherever humans go. Before that, robot should be capable to interact with people appropriately and move in the surroundings smoothly.

This thesis attempts to develop a system which combines mobile robot techniques of understanding knowledge of surrounding environments and algorithms that deal with human robot interaction. SLAMMOT is used on understanding environments and localizing itself. By using SLAMMOT and DAO* planning algorithm, the robot can provide environment information to users anytime and guide for them. The Color-based Particle Filter algorithm is implemented for recognizing the user’s dressing, so that it can follow specific user anywhere.

This thesis also proposes an effective following algorithm which can filter high-frequency noises. Impedance control is used to give velocity commands to robots. This algorithm is sustainable. It can be improved by adding new terms into it when the environment conditions are different.