An artificial neural network model of certain aspects of fighter pilot cognition
Jeffrey A. Smith, Rick Walters, Walt Goodson, R. Stites · 2002
The purpose of the research described was to model the representation of the major cognitive states and decision making processes of a fighter pilot during the intercept phase of a two-versus-two air combat engagement against a single group of adversary aircraft. A highlight of the research was the integration of an artificial neural network (ANN) model into a hybrid structure containing conventional symbolic logic and algorithmic elements. A conceptual framework was developed within which to define the situational awareness (SA) construct. The key design goal for the ANN was to achieve a depth of learning that transcends associative mapping and internalizes the multivariate, nonlinear relationships characteristic of an air combat engagement. An air combat simulation was created to illustrate the model's performance.>