Adaptive Nonlinear Flight Control and Control Allocation for Failure Resilience

Abstract

In this publication, reconfiguring control is implemented by making use of Adaptive Nonlinear Dynamic Inversion (ANDI) for autopilot control. The adaptivity of the control setup is achieved by making use of a real time identified physical model of the damaged aircraft. In failure situations, the damaged aircraft model is identified by the so-called two step method in real time and this model is then provided to the model-based adaptive NDI routine in a modular structure, which allows flight control reconfiguration on-line. Three important modules of this control setup are discussed in this publication, namely aerodynamic model identification, adaptive nonlinear control, and control allocation. Control allocation is especially important when some dynamic distribution of the control commands is needed towards the different input channels. After discussing this modular adaptive controller setup, reconfiguration test results are shown for damaged aircraft models which indicate satisfactory failure handling capabilities of this fault tolerant control setup.