Characteristics and signal processing of a proximity sensor

Abstract

SUMMARY This paper presents a fibre optic proximity sensor suitable for robotic applications. The sensor characteristics are first identified through experiments. Then the signal processing problem is studied for general applications where no prior knowledge of the sensor environment is assumed. The nonlinear filtering is carried out by applying a Kalman Filter. Particular attention is paid to the target nature dependence of the sensor output. Experimental results are reported by using the sensors on a real robot system.