Application of Neuro-Fuzzy Systems to Behavioral Representation in Computer Generated Forces

摘要

A new mathematical modeling technique known as soft computing is showing promise in modeling complex human behaviors. This paper will discuss the theoretical concepts of a soft computing technique, neuro-fuzzy systems. This technique has direct application to behavioral modeling of synthetic forces. Furthermore, the paper will present a specific experiment and modeling of the human operator's behavior. In the compensatory task the human has the error displayed and he attempts to null it to zero in the presence of random disturbances in the control process. Results will be presented demonstrating the validity of the hybrid modeling approach. This type of human behavior requires the modeling of perceptual, cognitive and motor processing.