Support Vector Machine for Function Degradation Prediction of Macro Fiber Composite

Macro fiber composite (MFC) will experience function degradation under the influence of the stress field, which is one of the key concerns in its application. In order to prevent the deterioration of the vibration control effect of the MFC actuator on the structure, it is necessary to propose an effective method to predict the function degradation of MFC. In this study support vector machine (SVM) is adopted in the function degradation prediction modeling of MFC. Aiming at the typical problem of stress-induced function degradation of MFC, a nonlinear prediction model based on SVM is established. The experimental data is used as samples to train the SVM model, and an effective method based on the established model for the function degradation prediction of MFC is proposed. The fitting error of the model to the experimental data is less than 1%, and the maximum prediction error for the MFC function degradation curve under 30MPa stress is 6.2%.