DAWN: A platform for evaluating water-pricing policies using a software agent society
Lately there is a transition in water management: policy makers leave aside traditional methods focused on additional-supply policies and focus on water conservation using demand control methods. Water Agencies use water-pricing policies as an instrument for controlling residential water demand. However, design and evaluation of a water-pricing policy is a complex task, as economic, social and political constraints have to be incorporated. In order to support policy makers in their tasks, we developed DAWN, a software tool for evaluating water-pricing policies, implemented as a multi-agent system. DAWN simulates the residential water demand-supply chain and enables the design, creation, modification and execution of different scenarios. Software agents behave as water consumers, while econometric and social models are incorporated into them for estimating future consumptions. Scenarios and models can be parameterized through a friendly graphical user interface and software agents are instantiated at runtime. DAWN’s main advantage is that it supports social interaction between consumers, which is activated using agent communication. Thus, variables affecting water consumption and associated with consumer’s social behavior can be included into DAWN scenarios. In this paper, DAWN’s agent architecture is detailed and agent communication using ontologies is discussed. Focus is given on the econometric and social simulation models used for agent reasoning. Finally, the platform developed is presented along with real-world results of its application at the region of Thessaloniki, Greece.
I. N. Athanasiadis, P. Vartalas, P. A. Mitkas, DAWN: A platform for evaluating water-pricing policies using a software agent society, 2nd Intl Congress on Environmental Modelling and Software (iEMSs 2004), vol. 2, pg. 643-648, 2004, International Environmental Modelling and Software Society (iEMSs).
You might also enjoy (View all publications)
- A weakly supervised framework for high-resolution crop yield forecasts
- Learning latent representations for operational nitrogen response rate prediction
- Simulation-assisted machine learning for operational digital twins