The disadvantage of the DI controllers are that they have trouble operating close to a singularity. This requires a high-fidelity model of the inverse aircraft dynamics to be used. This takes up lots of valuable computer time for an embedded controller. It also requires that the state transition matrix be invertible. For large control inputs, saturation could cause problems with the controller. Typically actuators and sensor noise are not included in the inverted model. The model requires full-state feedback, which means that unmeasured states must be estimated. At high frequencies the low-order inverse model may be unable to cancel some higher-order elements.[3] For landing aircraft, because the aircraft response is expected to match the internal model, mismatches are error signals.[4] During landing, large unexpected forces are applied which can cause the integrators to run away and lose the aircraft.[4]
[[Weight on Ground Switch]] – can be used to stop the forward loop integrators
Sources
- [1] C. Lubas and P. Paglione, “NONLINEAR DYNAMIC INVERSION APPLIED TO A F-16 AIRCRAFT,” 2009.
- [2] _20100033140
- [3] C. Frost, W. Hindson, E. Moralez, G. Tucker, and J. Dryfoos, “Design and Testing of Flight Control Laws on the RASCAL Research Helicopter,” in AIAA Modeling and Simulation Technologies Conference and Exhibit, Monterey, California: American Institute of Aeronautics and Astronautics, Aug. 2002. doi: 10.2514/6.2002-4869.
- [4] R. Colgren, D. Enns, R. Colgren, and D. Enns, “Dynamic inversion applied to the F-117A,” in Modeling and Simulation Technologies Conference, New Orleans,LA,U.S.A.: American Institute of Aeronautics and Astronautics, Aug. 1997. doi: 10.2514/6.1997-3786.
Backlinks
Dynamic Inversion Control Law Structure
[[Integral Windup]]
Reduced Order Models
State Feedback
State-Space Estimators