Article ID Journal Published Year Pages File Type
714337 IFAC-PapersOnLine 2015 6 Pages PDF
Abstract

Actuator jamming in flight control applications may be attributed to a temporary stall load due to large aerodynamic forces, or even a permanently stuck faulty control surface. These two root causes of actuator jamming have different consequences on available control authority and fault duration. As an important consequence, any reconfiguration strategy that may be applied to handle these types of jamming faults needs to distinguish between them in order to take appropriate measures. However, the similarity of fault phenomena between the two root causes makes this problem challenging (i.e., not possible using only passive fault diagnosis). We propose to integrate reconfigurable model predictive control (MPC) with active fault diagnosis (FD) to address this problem. After detecting actuator jamming, a sequence of reconfiguration strategies (relying on actuator constraint updates and a soft constraint formulation) is adopted to aid in the FD decision process and take appropriate measures after obtaining the FD results. The proposed reconfiguration strategies allow the FD module to distinguish the two root causes of the jamming and also detect the end of stall load. The MPC then implements a suitable controller reconfiguration depending on the FD results. A nonlinear Airbus civil aircraft simulator is used to illustrate the effectiveness of our proposed approach.

Related Topics
Physical Sciences and Engineering Engineering Computational Mechanics