Intelligent Control of the F-100 Turbofan Engine for Full Flight Envelope Operation

摘要

A nonlinear intelligent controller is proposed in this paper for the main fuel flow of the F-100 turbofan engine. Three different kinds of method, fuzzy control, neural network, and genetic algorithm are combined to design the controller. These three methods are known to be effective methods to deal with nonlinear systems. The fuzzy logic controller is designed in a way that the turbo engine will respond to the input signal at a minimum time without violating any safety limits. The genetic algorithm is applied for off-line learning the most suitable design parameters of the fuzzy logic controller. For full flight-envelope operation, a feedforward error-back-propagation neural network is used to learn the mapping of engine's altitude and Mach number into the parameters of the fuzzy logic controller. Finally, to reduce the tracking error of the rotor speed, an additional fuzzy controller is used to adjust the scaling factor of the main fuzzy controller.