Approach of pilot modeling for predicting pilot control behavior

During the process of aircraft design, the mathematical model of pilot control behavior characteristics is always used to predict aircraft flying qualities. This is one of the important methods to avoid pilot-aircraft adverse coupling. The pilot neural networks (NN) model was derived from experimental data obtained during pilot-aided tests using a special workstation. For a set of different aircraft dynamics configurations, which had different amplifications and different short period mode natural frequency, the pilot control behavior characteristics in the single-loop longitudinal pitch tracking task were studied based on neural networks model. The result shows that pilots have similar control behavior for similar aircraft dynamics configurations. Therefore, an approach used pilot control behavior of similar aircraft configurations to build a pilot model to predict pilot control behavior of new configuration was suggested. This pilot model's precision was further calculated. And the result verifies applicability of this new approach for pilot predicting modeling.