System Identification-based Fault Detection and Dynamic Inversion Control of an Uncrewed Aerial System

As the wide-scale use of uncrewed aerial systems proceeds towards civilian airspace, guarantees of operational safety over the entire flight envelope become more critical. In this paper, the adaptation capabilities of dynamic inversion given knowledge of failures are investigated. A reduced order model is estimated from measurements based on relevant dynamic modes for a defined window of flight data. This system identification algorithm allows faults to be estimated via tracking of calculated model parameters. These failures will be parameterized and fed into the control loop to update the dynamic model and prevent inversion error. The mathematical basis of the fault detection and diagnosis module is discussed and the initial results of the system identification algorithm as well as performance of the base dynamic inversion controller is shown.