Resumen
The effect of stress depolarization will cause actuation function degradation of a piezoelectric actuator, which can eventually trigger function failure of the piezoelectric smart structure system. In the present study, we experimentally demonstrate the degradation process of the actuation function of the Macro Fiber Composite (MFC) piezoelectric actuator. Actuation function degradation data of MFC actuators undergoing cyclic loads with four different stress amplitudes have been measured. Based upon the experimental results, the radial basis function (RBF) neural network learning algorithm was adopted to establish a neural network model, in order to predict the actuation function degenerative degree of the MFC actuator, undergoing arbitrary cyclic load within the concerned stress amplitude range. The maximum relative error between the predicted result and our experimental result is 4%.