Research On Intelligent Social Robot

Due to the fact that the robot requirements in academia and services is rapidly increasing, many universities and research institutes are seeking high cost-effective robots to carry out robotics-related education and research. Even in many other fields such as child education, health, brain science, cognitive science and sociology, the demand for robots is also rapidly increasing. Moreover, many robotic applications were fast growing in the past few year, and recent advances in robotics-related technologies such as mobility, artificial intelligence, wearable devices, robotic bases, sensors, RGB-D cameras, ROS (Robot Operating System) platforms, etc., have made it possible to quickly build an affordable robot.In this case, ROS-based framework, we design an easy to build social robot, which is mainly used for research and education. This thesis focuses on the research below:1. Designing a new social robot architecture based on the ROS framework, and divided into visual system, voice interaction system and control system.2. Constructing a prototype of the social robot, including the mobile mechanism and robot body. Then the motion model of mobile mechanism was analyzed and derived.3. The key algorithms such as autonomous obstacle avoidance, head detection, voice interaction, people following and social interaction are implemented.The results have been verified using the social robot prototype. This social robot has the capability of autonomously navigating in an unstructured and dynamic environment, avoiding collisions with stationary or dynamic obstacles and generating rich social behaviors such as head detection, taking pictures, interaction with social networks, people following, voice interaction...