INTERFERENCE MODELIZATION IN MULTI-ROBOT AUCTION METHODS

Task allocation is one of the main problems in multi-robot systems, very specially when the robots form coalitions. To get a good task allocation, we have to take into account, among other factors, the interference effect or the interaction between robots. Interference is not easy to modelize because it depends on a lot of dynamic factors. This paper models the interference impact using auction methods and support vector regression. We will show how the performance of the auction utility function can be improved if the interference impact is included in it. This method has been tested using transport like tasks taking into account an specific interference called physical interference.