Local measurement validation for an intelligent chirped-FM ultrasonic range sensor

Abstract

An important intelligent functionality for improving measurement systems is the validation of the returned information. To obtain this validation at the sensor level, it is essential that the sensor produces additional information for this purpose. This aspect follows directly from the measurement method. In this article, we apply these ideas to an ultrasonic range finding sensor dedicated to the navigation of a mobile robot. The original measurement method proposed, based on a modulated emission, provides many configuration parameters influencing the accuracy and detection performances, and two measured parameters linked to the distance to the obstacle. The redundancy between these two parameters allow building a local index of validity of the measured distance. The analog-digital implementation of this measurement method using a microprocessor board 80C196 provides easy observation and adjustment of all the considered parameters. In particular, the theoretical and experimental study of the measurement method configuration parameters leads to a sensor configuration model, whose input is the expected range, and whose output is the set of optimal values of configuration parameters in the sense of a maximal validity index.