Human Strategy Modeling via Critical Coupling of Multiple Characteristic Patterns

Abstract

Serious lags have been exposed in determining possible causes of air accidents because implications on flight safety lie deep in flight data that cannot be gained ahead of time. A human strategy modeling method is proposed in this paper to generate very large sample sizes of flight data for safety analysis. Three kinds of characteristic patterns are formulated and critically coupled to implement fast-time simulations of a man–aircraft–environment complex system with a human strategy model as the core. An instantiation of this human strategy model for the final approach and landing mission is developed and integrated into a digital flight deck. The results of five controlled experiments manifest that the human strategy model is flexible in testing different conditions to obtain flight data that may be valuable for safety recommendations. With stochastic optimization factors and recursive feedback dynamics in piloting problems having been addressed, the human strategy model is potentially capable for the extension of automation and the enhancement of safety.