Design of Probabilistic Pilot Model with Fuzzy Controller for Microburst Escape

Abstract

Pilot should control the aircraft manually when encountering low altitude wind shear during takeoff and landing. For wind shear escape and flight safety research, an effective human pilot model together with wind shear and flight dynamics model should be built with high fidelity. A skill-based human pilot model was built which can describe pilots’ characteristics such as experiences, skills, emotions, reaction abilities, etc. A fuzzy controller was designed for lateral and longitudinal escape control in pilot model. Since single pilot could not represent a group of pilots’ control behavior, some of the model parameters were set to be stochastic, then the Monte Carlo method was adopted to obtain a numerical approximation of safety analysis results. With the probabilistic pilot model, escape strategies and safety analysis can be studied by simulation with high fidelity.