METHOD FOR DETERMINING THE SPATIAL PATH OF AVOIDING AN OBSTACLE BY AN AUTONOMOUS UNINHABITED UNDERWATER VEHICLE

Abstract

The problem of safe movement of an autonomous underwater vehicle (AUV) in the presence of
stationary obstacles is considered. Traditionally, information about an obstacle is generated as the
AUV approaches the obstacle, and using it, the AUV control system makes a decision on the parameters
of the AUV’s further movement (course, speed, depth). The goal of the work was to determine the
spatial path to bypass the obstacle based on determining the geometric shape and size of the obstacle
according to digital maps. The paper proposes a method for determining a spatial 3D path to bypass
an obstacle, using complete information about the geometric shape and size of the obstacle, obtained
by supplementing the data from the means of illuminating the situation with data from digital bathymetric
maps of the areas through which the AUV route runs, as well as digital physical maps of the
areas of the earth. surfaces indicating small islands protruding onto the sea surface. The bathymetric
map isobaths are constructed from measurements at grid nodes covering the area under consideration;
the grid spacing exceeds hundreds of meters. To assess the probability of occurrence of bottom
topography anomalies between grid nodes that pose a danger to the movement of AUVs, it is proposed
to use the method of fuzzy probabilistic analysis. Based on the nodal points covering the obstacle,
a two-dimensional autocorrelation function is calculated, and the values of linguistic variables
are formed. Based on these variables, production rules were formed and, using them, the probability
of occurrence of relief anomalies was determined. To determine the shortest distance, the existing
depth grid at the node points of the obstacle is presented in the form of an oriented weighted graph:
the graph nodes are grid nodes with known depths, the edges are assigned weights equal to the spatial
distances between the three-dimensional grid nodes (latitude, longitude, depth). The developed
algorithm for determining the path to bypass an obstacle consists in determining the end point of the
bypass on the route trajectory behind the obstacle and finding the shortest path to bypass the obstacle
by comparing the current path under consideration with those obtained previously. If the length of
the path under consideration exceeds the length of the intermediate node of the previously formed
path, the process of reviewing the current path stops, and the transition to the consideration of the
next path is carried out. The results of the numerical experiments showed that the reduction in the
path around the obstacle compared to the traditional approach in the considered example was 17%.