Leakage Detection in Pipelines Using Decision Tree and Multi-Support Vector Machine

Abstract

Abstract-In order to solve the problem of leakage detection in the case of complex conditions and limited training samples, a multivariate classification recognition model was built by using Decision Tree and Support Vector Machine, which has advantages of rapid speed and high efficiency in classification and outstanding characteristics in small samples binary classification. The model was trained with a fault feature vector which is a dimensionless value extracted from the pipeline pressure signal characteristic parameters, and then using the model to test the samples. The results show that this method not only can complete the model learning training in the case of small samples, but also has been greatly improved over the neural network method in terms of the recognition performance, and can be effectively applied to leakage detection in pipelines.

A Novel PPA Method for Fluid Pipeline Leak Detection Based on OPELM and Bidirectional LSTM

“…In this section, several conventional ML based leak detection methods such as the BPNN in [9], SVM in [10], KNN in [13], ELM in [14], NB in [16] and DT in [17] are compared with the proposed method, and BiLSTM is also tested in the first-stage detection. Furthermore, to verify the effectiveness of BiLSTM in eliminating false alarms, the aforementioned ML methods with BiLSTM added as the second stage are also implemented.…”

“…It is shown that the time spent on model learning is greatly reduced. Applications of other ML methods in leak detection are also reported, including naive Bayesian (NB) based [16] and decision tree (DT) based classifier [17].…”