On the feasibility of applying pattern recognition concepts for an aircraft approach progress monitor

Pattern recognition concepts are investigated as a possible synthesis technique for an aircraft approach progress monitor. Some aircraft state variables are chosen as components of the pattern vector x which represents the state of the aircraft relative to a prescribed set of tolerances. A curve fitting algorithm is used as a method of solution for the required non-linear discriminant function, F(??(x)) = 0. The resultant decision surface is considered to be an adjunct to the pilot´s decision process. Simulation results are discussed which indicate a potential application for future system´s design.