Multi-Objective and Predictive Control - Application to the Clear Air Turbulence Issue

An algorithm was developed for switching between different maneuvering targets where several objectives are to be simultaneously taken into account. This algorithm predicts the motion of a vehicle in the near future and activates the control law that will best lead this vehicle to follow its designated objectives. The prediction is made with the aid of a model of the vehicle motion and a stochastic representation of perturbations such as wind or failures. This method is applied to the issue of an aircraft encountering unpredictable strong vertical turbulences trying to keep a safe flight level while smoothening vertical load factor for passenger comfort purposes. First, we show that a pitch-hold control law effectively damps vertical load during severe turbulence, while allowing the aircraft to diverge in altitude. We then alternate between this law and the conventional autopilot in order to keep a safe flight level and to avoid passenger injuries simultaneously. A set of simulation results is presented, in which our method has improved the aircraft’s performance with respect to both requirements.