Neural Network Analysis of Pilot Landing Control in Real Flight

A methodology for analysis of a pilots' landing control at the visual approach has been developed using a neural network modeling. While our previous study analyzed flight simulator operations, this paper describes the analysis of a real flight landing case. An experimental method which utilizes image processing of recorded video data is developed to obtain necessary data such as time histories of visual cues and control inputs. The effectiveness of this proposed method is confirmed by comparing values and analysis results from the video data with results obtained using GPS/INS data. It is expected that these methods can be used to reveal the characteristic of pilot control in real flight operation.