Multiagent system for intelligent Water Demand Management

The pressures exerted by the scarcity of resources and the respect to the environment, among other reasons, have led to the enormous current importance of the Water Demand Management. This means satisfying the demand using the least amount of resources, so that a fundamental aspect is the hourly forecast of this variable. This paper applies modern Artificial Intelligence techniques in solving this problem. More specifically, we have developed a Multiagent System aimed at decision-making in the management, such that it determines the optimal hourly quantity of pumped water that minimizes the management cost. The developed structure relies on complex forecasting methods, such as ARIMA techniques and Neural Networks, coordinated with other intelligent agents. According to the results achieved, this is an interesting alternative to face the problem.