In this paper, Self-OrganizingMap (SOM) for theMultiple Traveling Salesman Problem (MTSP) with minmax objective is applied\nto the robotic problem of multigoal path planning in the polygonal domain. The main difficulty of such SOM deployment is\ndetermination of collision-free paths among obstacles that is required to evaluate the neuron-city distances in the winner selection\nphase of unsupervised learning.Moreover, a collision-free path is also needed in the adaptation phase, where neurons are adapted\ntowards the presented input signal (city) to the network. Simple approximations of the shortest path are utilized to address this issue\nand solve the roboticMTSP by SOM. Suitability of the proposed approximations is verified in the context of cooperative inspection,\nwhere cities represent sensing locations that guarantee to ââ?¬Å?seeââ?¬Â the whole robotsââ?¬â?¢ workspace. The inspection task formulated\nas the MTSP-Minmax is solved by the proposed SOM approach and compared with the combinatorial heuristic GENIUS. The\nresults indicate that the proposed approach provides competitive results to GENIUS and support applicability of SOM for robotic\nmultigoal path planning with a group of cooperating mobile robots. The proposed combination of approximate shortest paths with\nunsupervised learning opens further applications of SOM in the field of robotic planning.
Loading....