Compensation of INS/LBL Navigation Errors in a Polynomial Sound-Speed-Profile

Abstract

This paper presents an autonomous underwater vehicle (AUV) navigation scheme that pairs an inertial navigation system (INS) and a long baseline (LBL) acoustic positioning system. The INS is assigned to be the main navigation aid because of its faster rate. Meanwhile, the LBL provides position reference for compensation of the INS’ main inherent drawback, i.e., accumulating errors. However, the LBL has to deal with time-of-flight (ToF) measurements that may not be carried out under line-of-sight (LoS) circumstances. This is because the propagation speed of underwater acoustic waves is subject to the sound-speed-profile (SSP) of the area in question. This paper’s contribution is to consider the SSP in ToFs while addressing the above scheme. Specifically, the discrete approach to raytracing was implemented. For a given ToF, the Snell’s parameter of the wave is estimated and subsequently used to compute the horizontal range. The ToF results are then used to estimate the position of the AUV, while the  position is obtained from a depth sensor. It was shown by simulation that the estimators can provide navigation with accuracy <0.5 m2, as it manages to compensate for errors. Since the estimation of Snell’s parameter is prone to exhibit imaginary numbers, future work should consider a more robust method to tackle this problem.