摘要
Water resources are highly prone to mismanagement due to uncertainties introduced by climate change and political interferences in the operations of water distribution systems. In countries like Pakistan, India and Egypt the irrigation networks are extremely vast and complex; the water sources are few and drying, while the infrastructure is in a constant decay, dating back by decades and centuries. Mismanagement and inefficiencies in such networks add up to floods and droughts like situations. Managing such complex and large networks poses challenges on multiple levels. The basin manager has to respect the international & provincial accords while designing policies. The water managing body has to ensure equity and work within the bounds of local water distribution schemes such as irrigation rosters. All of these factors demand more robust and reliable policies from water manager. In this paper we propose model- and data-driven frameworks to analyze the operations of water networks under varying conditions. Canal operations are posed as hierarchical optimization problems in which the low level captures physical regulation and the highest level corresponds to legal seasonal interpretations of an international or provincial accord. These frameworks then represent the decision support systems connecting to the next higher and lower layers. Tools such as risk sensitive control, model predictive control and fuzzy set theory are used to analyze the operations of canals as single and multi-objective optimization problem. In single objective optimization problem decision maker (DM) is modeled two different types of agents namely risk neutral and risk averse agents based on different economic instruments. A multi criteria decision making technique based on fuzzy set theory is also used to manage risks to accommodate non-quantifiable losses. These frameworks allow the basin manager to play what-if scenarios and create future projections in order to plan for long term future.