In this paper, we address the difficult problems in pipeline leakage diagnosis and investigate the leakage diagnosis algorithm based on vibration signal analysis with pipeline vibration signal as the basis. The SSA-VMD algorithm is used to decompose the pipeline vibration signal and conduct comparative experiments. The results of the comparison experiments show the superiority of the SSA-VMD algorithm. Then the decomposed IMF components are feature extracted to construct single feature vectors and then combined feature vectors. Finally, the support vector machine is optimized using the improved grid search method, and comparative experiments are conducted for the single feature vector and the combined feature vector, respectively. The experimental results show that the support vector machine optimized by the improved grid search method has a higher recognition accuracy.