Frequency-model-based obstruction detection and location in a pipeline by output error minimization

In this paper, the problem of obstruction detection and location in pipelines is considered. On the basis of a frequency modeling of pipeline dynamics together with the obstruction effect, a new algorithm is proposed: it basically amounts to obstruction direct identification, by some output error minimization technique. The proposed method is shown to be pretty efficient via simulation results.