Optimal Survey Path for MCM Operations using Variable Length Genetic Algorithms

We are interested in the optimal survey path that an autonomous underwater vehicle (AUV) should follow during the detection phase of mine countermeasures (MCM) operations. In our study, we suppose that the sea bottom survey is made by collecting high resolution images generated by a side looking sonar. We suppose the seabed map is available a priori. Our purpose is to minimize the travelled length under coverage constraint. The optimal coverage path is computed by means of a variable length genetic algorithm. Classical lawnmower pattern with horizontal tracks as well as grid pattern with both horizontal and vertical tracks are considered. The latter pattern is performed in series (in parallel) with one drone (two drones) respectively. It is shown that the use of a second drone for perpendicular tracks leads to significant reductions in mission execution times which are mandatory for swift clearing of mine like objects.