Fault samples’scarcity is a key factor that prevents the progress of the intelligent monitoring research.Support vector machine（SVM）

based on the structural risk minimization criterion

can generalize a conclusion from limited samples and keep robust.Using the SVM-based multi-classification algorithm for the stiffened structure health monitoring in aeronautics and astronautics area and comparing with the BP neural network

the satisfying results are obtained.The experimental results demonstrate that SVM has better classification performance in this research case than the BP neural network.