Fuzzy inference engine provides opportunity for testbed

Abstract

In the development of a fuzzy logic implementation of a high-order, nonlinear, dynamic muscle model, opportunities continue to arise for the development and verification of other fuzzy implementations, including selected differential calculus and other classical mathematical relationships. The availability of a reliable, accessible, and powerful Generalized Fuzzy Inference Engine (GFIE), developed as part of the fuzzy modeling effort, facilitates the realization of fuzzy implementations such as low-pass filtering. This tool has also facilitated evaluation of fuzzy mathematical operations published in the literature but not always included in fuzzy systems software packages (e.g. MATLAB). Advantages of using fuzzy implementations described herein include robustness to noise and signal uncertainty (e.g. temporary signal drop-out) as well as potential reduction of sampling rates. Noise and parameter sensitivity studies are planned for quantifying these advantages in the setting of the fuzzy muscle model.