The Outlier Elimination Methodology of Ultra-Short BaseLine Positioning Data

Ultra-short baseline positioning is an important acoustic positioning method for underwater vehicles. The main purpose of this paper is to reduce the positioning error caused by outlier of acoustic measurements. The paper first introduces positioning theory and error source of ultra-short baseline, then analyzes the influence of motion trajectory of underwater vehicles on the variation characteristics of time difference of arrival and range. On the basis of Kalman filter, an outlier elimination methodology based on dynamic innovation threshold has been proposed. Experimental data from submarine “Exploration No.1” testing in South China Sea in 2017 is used to verify the effectiveness of proposed method. The paper combines multiple channel information of acoustic array to eliminate positioning outlier and analyzes the depth positioning results before and after the elimination of outlier. The results show that outlier eliminating rate can reach up to 88.2 %, and the effective positioning data rate increases by 4.9%.