Multibeam bathymetry based on multiple path planning with genetic algorithm

For the survey line route planning, two schemes are proposed, such as single-direction non-closed-loop traveling and multi-direction turning traveling. Although single-direction non-closed-loop driving can realize the shortest path and reasonable overlap, there is a leakage area; while in the discussion of multi-direction turning driving, the feasible driving area of the upper triangle is determined first, and then the area is cut by the microelement method, and the MADDPG algorithm is used to subdivide the parallel lines to determine the approximate direction of driving. Finally, we need to integrate the redundant angles in the cut area, determine the corresponding boundary conditions, adjust the spacing of parallel lines and introduce the TD3 static path planning algorithm to determine the optimal driving path. For the existing sea depth data, normalization and heat map are used to show the general trend of the sea depth following the azimuthal coordinate transformation. Then, the existing data are optimized by genetic algorithm again, and the above path-planning algorithm is applied to solve the driving line paths, and the relevant indexes are calculated to verify its reasonableness.