Single-beacon Based Cooperative Path Planning

[bibshow file=MyPublications.bib process_titles=0]

    rangingRecent advancements in the development of autonomous underwater vehicles (AUVs) and underwater communications have made inter-vehicle acoustic ranging a viable option for underwater cooperative positioning and localization. The idea of AUV cooperative positioning is to have a vehicle with good quality positioning information (beacon vehicle), to transmit its position and range information acoustically to supported AUVs (survey AUVs) within its communication range during navigation. Generally, the beacon vehicle is equipped with high accuracy sensors that are able to estimate its position with minimum errors. The range information between the vehicles can then be fused with the data obtained from proprioceptive sensors in the survey AUVs to reduce the pbeaconositioning error during underwater navigation. In [bibcite key=tanoceans2011,Tan:JOE2012], we focused on cooperative path planning algorithms for the beacon vehicle using dynamic programming and Markov decision process formulations. These formulations take intoaccount and minimize the positioning errors being accumulated by the supported AUV. These approaches avoid the use of Long-Base-Line (LBL) acoustic positioning systems as well as allows the supported AUV to remain submerged for a longer period of time with small position error.

    Although managing to achieve some promising results, previous approaches require either high computational load or large number of manually selected representative states for the policy matrix. In [bibcite key=Tan:2012Baltimore], we further extend the work by employing Direct Policy Search (DPS) where the state space is approximated in the form of vlgaVoronoi Tessellation and the states are represented by the Voronoi seeds. The formulation allowed us to deploy the Variable-Length Genetic Algorithm (VLGA) to automatically discover the optimal number of these states while simultaneously learning their corresponding action mappings.  Compared to the previous published approaches, our approach greatly reduces the computational load as well as the size of the policy matrix, yet manages to perform comparatively well in terms of minimizing the survey AUVs’ position errors.