Fuzzy behavior modulation with threshold activation for autonomous vehicle navigation

Abstract

The paper describes fuzzy logic techniques used in a hierarchical behavior-based architecture for robot navigation. An architectural feature for threshold activation of fuzzy behaviors is emphasized, which is potentially useful for tuning navigation performance in real world applications. The target application is autonomous local navigation of a small planetary rover. Threshold activation of low-level navigation behaviors is the primary focus. A preliminary assessment of its impact on local navigation performance is provided based on computer simulations.