Recursive Algorithm for AUV Positioning and Determining its Orientation

The work is devoted to the problem of determining the coordinates and orientation angles of an autonomous underwater vehicle (AUV) in the interests of bringing the vehicle to a stationary underwater station using a high-frequency short-range sonar system. This process involves AUV maneuvering and its approach to the underwater station, which requires formation of three zones with different acoustic visibility of the station emitters by the AUV receivers. The paper presents a mathematical formulation of the navigation problem, the solution of which is reduced to the use of a recursive algorithm with the formation of a system of algebraic equations, where the coordinates and orientation angles of the AUV are assumed to be unknown, and the time of the signal front detection by the AUV receivers is used as initial parameters. The solution of the system of algebraic equations becomes possible when the domain of the desired variables is limited, i.e. when the area of the AUV current location and the orientation angles are specified. The formation of a neighborhood is possible using the machine learning method - training with a teacher, where a training sample is formed using a model of signal reception by the AUV receivers from the station beacons on a given coordinate grid – orientation angles. The paper presents the results of the problem solution using the method based on the k-nearest neighbors (k-NN) algorithm, a decision tree, neural network, and random forest.