Combining multiple perspectives on clustering. Node-pipe case in hydraulic sectorization

Abstract

This work addresses the problem of water network division into district metered areas by the use of clustering respect to multiple views. In our application will use two data perspectives: one of these will be associated with consumption nodes of the water supply network, meanwhile the other one will be related to pipes. It also will be needed to combine the information from these multiple points of view. Thus, we communicate both solutions negotiating the respective cluster memberships by the interaction of their con gurations based on schemes of intelligent agents that can merge the respective information layers of each structure. The result of this proposal is a very useful approach because it combines pipe properties, nodes, and graph information associated with the supply network as strong criterion to achieve system division. A real water supply system will serves as a case-study to check the performance of the division proposed.