Fuzzy Logic-Based Inferencing in the Presence of Input Data Uncertainty.

摘要

In this study, the fundamental problem of handling uncertainty in the input data to a fuzzy inference system is addressed. A novel solution to this problem is derived, based on the principle of fuzzy composition. Application of this concept to the fuzzy characterization of contact speed with uncertain platform classification information is demonstrated and is shown to provide significant improvements in tracking solution quality for the single-leg target motion analysis (TMA) problem. Follow-on efforts are underway to address the issue of modeling uncertainty in the basic structure of a fuzzy inference system.