Multiple Blockage Identification of Drainage Pipeline Based on VMD Feature Fusion and Support Vector Machine

Aiming at the detection problem of Multiple blockage in urban water supply pipelines and drainage pipelines, also the problem of distinguishing commonly used pipe components such as lateral connection from the actual blocking conditions. A multiple-blocking fault identification method based on support vector machine (SVM) combined with a feature extraction approach for component signal are proposed in this paper. Firstly, the variational mode decomposition (VMD) was applied on the acoustic signals collected in the pipeline to obtain a set of finite bandwidth natural mode functions (IMF), multiple time domain indices and center frequencies were extracted as features, then a feature vector set can be constructed and input into the SVM classifier. The experimental results have shown that the method based on VMD feature fusion and support vector machine can effectively identify the multiple congestion faults of drainage pipelines. In addition, the method was compared with back propagation (BP)neural network and the k-nearest neighbor algorithm (KNN). The results suggest that the proposed method has a better performance on the partial blockage recognition with a small number of training samples.