Localization of an Underwater Control Network Based on Quasi-Stable Adjustment

Abstract

There exists a common problem in the localization of underwater control networks that the precision of the absolute coordinates of known points obtained by marine absolute measurement is poor, and it seriously affects the precision of the whole network in traditional constraint adjustment. Therefore, considering that the precision of underwater baselines is good, we use it to carry out quasi-stable adjustment to amend known points before constraint adjustment so that the points fit the network shape better. In addition, we add unconstrained adjustment for quality control of underwater baselines, the observations of quasi-stable adjustment and constrained adjustment, to eliminate the unqualified baselines and improve the results’ accuracy of the two adjustments. Finally, the modified method is applied to a practical LBL (Long Baseline) experiment and obtains a mean point location precision of 0.08 m, which improves by 38% compared with the traditional method.

Positioning Accuracy Model of Sailing‐Circle GPS‐Acoustic Method

“…During the iteration, stop criteria and the external parameters are of various choices. For instance, Zhao, Chen, Zhang, and, Fend (2018b) utilize dichotomy algorithm in which the prior depth of the underwater transponder measured by pressure sensor is regarded as the external parameter to perform the inverse ray tracing method to get the accurate incident angle. Chadwell and Sweeney (2010) employ downhill simplex method in which the prior plane distance of the transducer and the transponder is taken as the external parameter to adjust the incident angle under the condition that the plane coordinates of the transducer and the transponder are known.…”

Weighted quasi-stable combined adjustment of two-order horizontal control network using posteriori variance to determine the weight

“…However, this method did not weaken the influence of the initial data error [18,19] proposed an adjustment method for appending datum taking initial data errors into account, which was suitable for deformation monitoring data processing [20] considered the influence of initial data error on coefficient matrix of error equation, and studied the total least squares method considering initial data error. However, the precision of adjustment result is reduced [21] compared the precision of the adjustment with initial data error and integrated adjustment and reported that the precision of high-order control points can improved after integrated adjustment [22] used the initial data as quasi-stable points and performed weighted quasi-stable adjustments to weaken the influence of the initial data error. In this method, the prior variance matrix of the high-order control points was used to determine the weight matrix of datum [23] assigned different weights to different IGS stations when processing GPS network data, which not only controlled the error and determined the reference frame but also avoided large distortion of the GPS network [24] adjusted the spatial network coordinates with the ground network observations, and determined the weights of different kinds of observations by variance component estimation.…”

“…A reasonable datum can not only offer coordinate system and control the accumulated survey error, but also weaken the distortion of the observations of the low-order control network caused by the high-order control points. In order to obtain more reasonable datum, inspired by the work conducted in [21][22][23][24][25], this study proposes a new method to adjust the observations of a two-order control network. In this method, the coordinates of high-order control points are unchanged and are used as the quasi-stable datum.…”