Study on Defect Detection Based on D-S Evidential Reasoning in Natural Gas Pipeline

Based reasoning data fusion method, Dempster-Shafer evidence reasoning applies to natural gas pipeline detecting experiments. Through the use of multiple sensors to sample the defect information in natural gas pipeline, extracting characteristic information, using neural network to carry on the fusion recognition, and then the neural networks for normalized output value as evidence, Dempster combination rule to fuse the data further and improve reliability, the final identification of the effective decision-making a type of pipeline defect. This experiment shows the D-S evidence reasoning has a strong ability to deal with uncertain information, the result proved the correctness and effectiveness.