Uncertainty Measurement for Ultrasonic Sensor Fusion Using Generalized Aggregated Uncertainty Measure 1.

In this paper, target differentiation based on the pattern of data which are obtained by a set of two ultrasonic sensors is considered. A neural network based target classifier is applied to these data to categorize the data of each sensor. Then the results are fused together by Dempster–Shafer theory (DST) and Dezert–Smarandache theory (DSmT) to make a final decision. The Generalized Aggregated Uncertainty measure named GAU1, as an extension to the Aggregated Uncertainty (AU), is used to evaluate DSmT. Then the GAU1 and AU as the uncertainty measures are applied to the obtained results of the decision makers to evaluate DSmT and DST accordingly. The introduced configuration for decision making has enough flexibility and robustness to use as a distributed sensor network.