A systematic methodology including a computational pilot model and a pattern recognition method is presented to identify the boundary of the flight performance margin for quantifying the human factors. The pilot model...A systematic methodology including a computational pilot model and a pattern recognition method is presented to identify the boundary of the flight performance margin for quantifying the human factors. The pilot model is proposed to correlate a set of quantitative human factors which represent the attributes and characteristics of a group of pilots. Three information processing components which are influenced by human factors are modeled： information perception, decision making, and action execution. By treating the human factors as stochastic variables that follow appropriate probability density functions, the effects of human factors on flight performance can be investigated through Monte Carlo（MC） simulation. Kernel density estimation algorithm is selected to find and rank the influential human factors. Subsequently, human factors are quantified through identifying the boundary of the flight performance margin by the k-nearest neighbor（k-NN） classifier. Simulation-based analysis shows that flight performance can be dramatically improved with the quantitative human factors.展开更多
基金supported by the National Basic Research Program of China(No.2010CB734103)
文摘A systematic methodology including a computational pilot model and a pattern recognition method is presented to identify the boundary of the flight performance margin for quantifying the human factors. The pilot model is proposed to correlate a set of quantitative human factors which represent the attributes and characteristics of a group of pilots. Three information processing components which are influenced by human factors are modeled： information perception, decision making, and action execution. By treating the human factors as stochastic variables that follow appropriate probability density functions, the effects of human factors on flight performance can be investigated through Monte Carlo（MC） simulation. Kernel density estimation algorithm is selected to find and rank the influential human factors. Subsequently, human factors are quantified through identifying the boundary of the flight performance margin by the k-nearest neighbor（k-NN） classifier. Simulation-based analysis shows that flight performance can be dramatically improved with the quantitative human factors.
