Study Of Modeling The Pilots' Control Process Based On Predictive Control

Flight safety is one of the most important issues in the field of aviation.Statistics show that human factors have become the main cause of flight accidents,and the research of human factors is significant to improve flight safety.Human-machine-Environment integrated simulation is an effective method for quantitative study of human factors.The core issue is to build a pilot model that includes a variety of human factors,which can reasonably and accurately reflect pilots' perception,decision-making and control.The research contents of this paper include: The advantages and disadvantages of the existing pilot model are compared and analyzed,and the direction of improvement is proposed.The nonlinear dynamic model of the aircraft is realized in the simulation program,and some controllers are designed to simulate the automatic and auxiliary control function of the aircraft.A cognitive process model for pilots is established based on the information processing theory.Based on the cognitive model,a predictive control model is established to describe the control behavior of pilots.The model can simulate the pilot's recognition of aircraft system and the prediction of the fight path.The model is integrated with human factors such as human time delay,human errors,experience and habit,physiological state and so on to simulate the characteristics of the pilot's behavior.The advantages and disadvantages of the predictive control model and the PID control model are compared with the real flight data.The influence of human factors on flight performance is studied based on the human-aircraft joint simulation experimental results.The results of the study show that: The proposed pilot cognitive model can reflect the cognitive process of the pilot and provide theoretical support for the quantification of human factors.Compared with the pilot's PID control model,the proposed pilot predictive control model has a better comprehensive control performance and a higher fitting degree to the real flight data.The predictive control model can reflect the effect of a variety of human factors on flight performance.