An Approach to Information Fusion using Interval-Valued Probability

This paper discusses the issues of combining disparate sources of information (information fusion) in the context of planning and decision making for intelligent vehicles. Here interval-valued probability is used to capture uncertainty due to stochastic processes (objective uncertainty) and due to a lack of knowledge about the processes (subjective uncertainty). This paper discusses the representation problem for information fusion, an example of its use and the interactions between the information fusion and planning functions.