Vibration Control of a Smart Composite Engine Fan Blade by Neural Network Controller

This research analyzes the adoption of an artificial neural network (ANN) method for a smart engine fan blade bonded with a piezoelectric (PZT) patch made of macro fiber composite (MFC) for vibration control and suppression. The MFC material that functions as a sensor and an actuator is embedded on the surface of the fan blade to provide structural active damping on the structure. The finite element model (FEM) also the system state-space model of the engine fan blade is formulated to establish the structural numerical model as a basis to initialize the impulse input disturbance to measure the performance of the controllers. To predict the future reaction of the smart fan blade, the recommended controller is trained using the nonlinear autoregressive moving average with an exogenous input model of order 2 (NARMA-L2). The PID-supplemented compensation method improves the controller's performance, and the NARMA-L2 and fuzzy generate the necessary control signal. The performance will then be assessed by contrasting the reactions of the NARMA-L2-PID, fuzzy-PID, and classical PID controllers under a specified dynamic excitation. The outcomes show that the proposed controller model is effective and reliable regardless of the input disturbance exerted on the structure's vibration response.