The application and analysis of H∞ filter based on SLAM in detection and location of underwater oil pipeline leakage

摘要

Aiming at improving the positioning accuracy of the submarine oil pipeline leak location, a new positioning algorithm SINS combined SLAM is proposed. When AUV is navigation positioning, it can realize the tracking of Submarine oil pipeline's simultaneous localization and mapping, and use the location information of map features in the submarine oil pipeline to correct method of SINS which has more advantage and makes the system error not accumulate with time. By setting specific pipeline mathematical function, collecting its two-dimensional position information with sensors,the actual graphics was built, and processing the marked accurate leakage point from image to achieve accurate positioning purposes; Using the extended H filter (EHF) to do simulation experiment on position information acquired by SINS/SLAM integrated navigation. Which show data association of noise and observation are robust to model uncertainties, and filtering accuracy reaches 0.1% of total distance. The oil pipeline leak location precision is also guaranteed, and passive navigation for underwater pipeline tracking detection is feasible.