MFA testing uses LTI analysis of the aircraft response.[^1] The MFA regulators had to be tuned to cover any zero dynamics that were not removed from the model. [^1] The equation from the MFA that must be tested is[^2]
$$y(t)=cv(t)=\int\dot{cv}_{des}(t)dt$$
Any zero dynamics are re-introduced with a regulator.[^2] To test the MFA you must remove the desired dynamic and tie the interceptor inputs to the state dot commands. Programmed test inputs provided precise command inputs. This was done to show that the dynamic inversion can produce a pure integrator. This was tested with frequency sweeps. The result was that the gain was accurate to around 10 rad/s and the phase began to degrade around 9 deg/s due to the smoothing filters for the feedback sensors.
Not all of the desired dynamic models have been tested on the MFA.
Sources
- M. C. Cotting, S. S. Stephens, J. Cole, J. Barricklow, and W. Gray, “X-62 VISTA Capabilities and Architecture,” in AIAA SCITECH 2023 Forum, in AIAA SciTech Forum. , American Institute of Aeronautics and Astronautics, 2023. doi: 10.2514/6.2023-1744.[^1]
- M. C. Cotting et al., “X-62 VISTA Simulation and Autonomy Flight Testing,” in AIAA SCITECH 2023 Forum, in AIAA SciTech Forum. , American Institute of Aeronautics and Astronautics, 2023. doi: 10.2514/6.2023-1928.[^2]
Backlinks
[[Bode Plot]]
[[Bode Plot of an Integrator]]
[[Filters for Sensor Conditioning]]
[[LTI Systems]]
MFA Control Law
Stitched Learjet-25D model
Transfer function of an Integrator
[[Voltage Sweep]]
[[VSS Testing]]