A fuzzy approach to the weighted region problem for autonomous vehicles

An approach is proposed to represent and reason about the uncertainty in terrain description by using fuzzy set theory. The whole terrain is assumed to be classified into a set of fuzzy regions, and weight of region is interpreted as a linguistic variable whose values are linguistic labels of fuzzy sets. Based on this representation, a decision function is established whose minus gradient information is used to lead the vehicle to travel to its goal with an acceptable cost.<>