Underwater robot multi-source acoustic information fusion accurate navigation method

Abstract

The invention provides an underwater robot multi-source acoustic information fusion accurate navigation method, which comprises the following steps that an AUV (Autonomous Underwater Vehicle) acquires submarine topography surveying and mapping information and distance observation information between the AUV and a beacon through multi-beam sonar and beacon communication, and distinguishes and eliminates distance observation outliers caused by a weak communication environment through observability test and a probability graph division algorithm, and finally, takes AUV submarine topography surveying and mapping information, distance observation information between the AUV and the beacon and AUV navigation system information as input to construct a multi-source acoustic information fusion SLAM pose graph model, and through a pose graph optimization algorithm, estimate the current position of the AUV, the historical track of the AUV and beacon positioning information, so that the problem of long-term, hidden, all-weather, accurate and real-time navigation of the AUV in deep sea is solved. According to the method, a multi-source acoustic information fusion image is constructed to optimize an SLAM algorithm framework, and real-time and accurate navigation information is provided for AUV large-range deep sea operation.

Description

technical field [0001] The invention belongs to the field of underwater navigation of an autonomous underwater vehicle (AUV), and in particular relates to a multi-source acoustic information fusion precise navigation method for an underwater robot. Background technique [0002] Autonomous Underwater Vehicle (AUV) has become an important tool for high-resolution deep-sea topographic mapping, but the accuracy of the bathymetric map drawn by AUV depends on the error of the AUV navigation system. Inertial navigation systems, long baseline acoustic positioning systems, ultra-short baseline acoustic positioning systems, and geophysical navigation systems all provide navigation information for AUV deep-sea operations. However, even with the addition of expensive equipment such as Doppler velocity logs and attitude and heading reference systems, the navigation error of inertial navigation systems will grow infinitely over time. The positioning results produced by the long baseline ...

AI Technical Summary

Problems solved by technology

Accurate data associations can only be generated when the AUV retraces to previously visited locations, but in deep-sea missions, performing terrain retracing by AUVs is often time-consuming and labor-intensive

Therefore, terrain backtracking usually occurs after a long period of time when the mission is performed, which means that the AUV will use the inertial navigation system to determine its own pose for a long time after the start of the mapping mission. During this period, the real-time position estimation of the AUV is extremely poor. precise

For beacon SLAM, the error constraint results are relatively accurate in a short period of time when the beacon maintains communication, but its operating range is severely limited by the beacon, which cannot provide accurate navigation information for AUV long-term and large-scale operations

Images