A technique for predicting pilot model parameters using inverse simulation

The practice of using mathematical models to simulate pilot behaviour in on e-axis stabilisation tasks is a well-known conventional simulation problem. In this paper an accepted mathematical model of a pilot is used as the con troller of a rudimentary helicopter model. Flight test manoeuvre data from inverse simulation runs is used to provide time-histories which represent i nput forcing functions to the pilot-helicopter system, and a constrained op timisation routine is utilised to obtain values for the pilot gain and lead /lag equalisation parameters. It will be shown that as the theoretical pilo t is required to 'fly' different manoeuvres, or indeed if the level of mano euvre aggression is varied, the 'pilot' adjusts these parameters to perform a tracking task in compensatory control. The paper considers initially the pilot and helicopter models and subsequently analyses the whole system, il lustrating how the pilot model changes with different situations.