Integrating swarm intelligence in a Multi-Agent System for wastewater discharges management
Marta Verdaguer, Montse Aulinas, Narcís Clara, Manel Poch · 2010
The integrated management of the Urban Wastewater System (UWS) involves to deal with a high level of complexity. This complexity is a result of the need to reach the UWS global goal (i.e. to discharge wastewater with an appropriate quality into the receiving media) and, at the same time, the requirements from the particular dynamics and interactions between their components (i.e. sewer system, urban wastewater treatment plant and the receiving media). This work considers a Multi-Agent System approach of the UWS using GAIA methodology that permits to abstract all the components and their relations. Through its development we can identify a NP-Hard problem with multiple choice and multiple constraints. We suggest a new approach to solve the wastewater discharges management task based on a combinatorial optimization process. We propose the use of swarm intelligence with an ant system algorithm. We define a function objective involving costs, constraints related to total volume and total pollutants' load in order to not overload the Wastewater Treatment Plant. Finally, we propose a heuristic equation that guides the construction of solutions towards bigger volumes and concentration of pollutants.