A 24-GHz Wireless Locating System for Human–Robot Interaction

摘要

Due to technological advances in various fields and the rising demand for automation, the popularity of service robots is expected to grow rapidly in the upcoming years. One of the core functions of such robots is their interaction with people. One type of interaction, which is studied in this paper, is when the robot leads or follows a person. This presupposes that the position of the person is known relative to the mobile robot. In this paper, we present a 24-GHz wireless locating system to enable this kind of human–robot interaction. The system consists of a radar node attached to the robot and an active portable wireless beacon that is carried by the person. Using a secondary radar approach to measure the distance and angle to the beacon, the 2-D position of the person relative to the mobile robot can be accurately determined by a single radar node on the robot. This is unfeasible with standard multilateration localization approaches. An extended Kalman filter is used to increase the accuracy and reliability of the positioning system. The measurements presented in this paper, performed in a scenario that is both challenging and realistic, demonstrate the effectiveness of the proposed approach.