Verification and evaluation methods for pipeline leak detection algorithms

Abstract

As an important equipment in energy transportation and utilization, state detection is a key link to ensure the safety and environmental protection of the equipment. At present, there are many types of pipeline leakage detection algorithms, but there is less verification and evaluation of the actual performance and characteristics of the algorithm. Therefore, based on the evaluation index system, the entropy weight method (EWM), and the TOPSIS, we build a pipeline leakage inspection evaluation model of the measurement algorithm. In addition, the actual liquid pipeline transportation experiment platform is established. The negative pressure wave method is selected and combined with the CNN-LSTM algorithms to perform experiments. The results show that the algorithm used in this paper is better than the traditional negative pressure wave algorithm (NPW). The performance and calculation results can provide a reference for scientific research, engineering, teaching, etc.