Aim at the limitation of common non-destructive technology, an active real-time monitoring system based on the fiber Bragg grating sensors is set up in this paper. Support Vector Machine (SVM) is applied to detect the impact damage for the carbon fiber composite laminated plates, and compared with the traditional BP neural network. The results show that SVM possesses the better ability of regression and higher accuracy than the traditional BP neural network. Furthermore, the monitoring system can identify the locations and degrees of impact damage real-time effectively.