Artificial Neural Network-Based Model for Galloping-Based Piezoelectric Energy Harvester

Accurate prediction of power output for galloping-based piezoelectric energy harvesting (GPEH) system depends on the representation of the aerodynamic force. Traditionally, higher-order polynomials are used to approximate the galloping force based on quasi-steady theory. It is very difficult to fit the coefficient of lift and angle of attack curve by using polynomials. In this paper, an artificial neural network (ANN) technique is proposed to approximate aerodynamic force. The results from both ANN and polynomial approximation-based models are compared and conclusions are drawn. It is found that for low wind speeds, ANN-based model gives good results compared to polynomial approximation.