Enhance Upset Prediction and Recovery Training Motion Cueing Using Switched Model Predictive Control and Pilot Model

Upset prediction and recovery training (UPRT) has been required to perform by airline pilots in Level-D flight simulators. However, hexapod-type motion platform now used in simulators, which has limited workspace and motion capability, cannot faithfully reproduce the motion feeling pilots experience in real aircrafts under upset scenarios. In this paper, an enhanced motion cueing algorithm (MCA) for UPRT, which combined switched model predictive control and pilot model, is proposed. The UPRT scenario can be divided into five stages according to aircraft upset states. In order to ensure motion fidelity in all stages, a MCA based on model predictive control (MPC) is designed to switch between these stages with different performance indexes and constraints. The constraints of MPC were estimated by automatically executing the upset recovery process using a probabilistic pilot model. Simulation results showed that the proposed enhanced MCA has better motion simulation fidelity and also make better usage for the motion system's workspace than the classical MCA.