Abstract

One important prerequisite for the effective behaviour of adaptive industrial robots is the ability to detect the location and orientation of surrounding surfaces. This can be done by three-dimensional (3-D) vision, for example, but the method is rather complex and costly. In this article a simpler method is presented. The idea is to equip a robot gripper with ultrasonic sensors. The measuring principle used is bandpass-to-lowpass transformed impulse response. The impulse response is chosen since it is a sensitive indicator of time delays. The method called Taylor-series estimation is used for parameter estimation to reduce the influence of extraneous noise signals on the measured signals. With one transmitter and three receivers one can measure the distance to a plane or almost plane surface of not too small area, as well as its orientation relative to the sensor plane. In this article, the mathematical model for the detection of planar surfaces is derived and some experimental results are presented.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call