| A series of Artificial Neural Networks were trained using flight test data, to identify a possible spin entry, differentiate between an incipient spin and a stabilized spin, and predict required recovery controls. These were then implemented into a simulation and tested using actual flight test data from NASA to verify that artificial neural networks can successfully be used for this application. The spin avoidance and recovery system functioned properly. In addition a weighting system was developed to help test pilots and flight test engineers predict possible spin characteristics on aircraft where the moment of inertia about x axis is greater than the moment of inertia about the y axis, the moment of inertia about x axis is equal to moment of inertia about the y axis, and where the moment of inertia about x axis is less than the moment of inertia about the y axis. |
