engleski

Control Allocation with Actuator Dynamics for Aircraft Flight Controls

This paper addresses the control allocation to several aircraft flight controls to produce required body axis angular accelerations. Control law is designed to produce the virtual control effort signals, which are then distributed by solving a sequential least squares problem using active set method to the flight control surfaces to generate this effort. Two cases are described: in the first case the control law and allocation for the healthy aircraft is implemented, and in the second case, jamming of one control surface is introduced at time zero. In this case, it was shown how the controller and allocation compensate for this failure without changing the control law. To implement this system it was assumed that there is a good fault identification system onboard. Normally aircraft are over-actuated and in the case of a control failure this over actuation is more pronounced due to coupling of aircraft dynamics. Instead of using one-to-one mapping between control allocator and control surfaces, actuator dynamics was included in the system. The discrepancy in the optimal signal from control allocation due to this additional dynamics was compensated using the scheme mentioned in this paper. Each gain corresponding to the actuator is tuned using genetic algorithms (GA). The controller and allocation design are implemented on a nonlinear B747 model with actuator dynamics.

filght control; virtual control effort method; sequential least squares; active set method; fault-tolerance

nije evidentirano