부산대학교 도서관

In this paper, a neural network controller for F-100 twin-spool turbofan engine is proposed. Neural network with its learning ability is used to learning the inverse dynamic property between the main-fuel rate and the output of the high pressure turbine at the operation point of see level (zero altitude and zero Mach Number). Then, this well-trained neural network plays the role as a controller to guide the high pressure turbine to follow a desired speed trajectory. Computer simulation showed that the proposed control scheme can drive the turbo engine to its desired output with faster rising time and without any overshooting.