Methods Research In Pipelines Leak Detection And Location Based On Kalman Filter

Pipeline as one of the most important transportations is playing more and more important role in the national economy. However, aging pipeline, corrosion and vandalism, cause the pipeline leaking frequently and lead to huge economic losses and environmental pollution, therefore, the pipeline leak detection and location have become an important research topic. At present, based on the pipeline mechanism model Kalman filtering algorithms have achieved well results in leak detection, but because of the algorithm complexity and large amount of calculation, their applications are limited; Based on the pipeline signal model, Kalman filter as a relatively simple model can't accurately alarm, and the model uncertainty and the signal with strong disturbance cause low positioning precision and poor system robustness. As a result, researching Kalman filter algorithm, in the premise of, reducing model uncertainty and signal disturbance on pipeline leak detection and location accuracy, and decreasing the leakage false alarm have significant theoretical significance and application value. In this paper, anglicizing the filter-based pipeline leakage detection and location, using pipeline historical operating data to identify the pipeline model, and proposing Kalman filter with unknown inputs algorithm to detect pipeline leak that can decrease the false alarms caused by frequent operation; establishing the random walk model with pressure data,using the strong tracking filter to locating the pipeline leak, and extracting leaking point from the fading factor sequence to locate the leakage,that can improve the filtering performance and accurately locate the leak.Experimental results showed that the proposed Kalman filter-based pipeline leak detection and location algorithm is able to improve the filter performance, to increase robustness of model uncertainty, to decrease the fault alarms efficiently, raise precision of location.