Development of a Hybrid Bayesian Network Model for Hydraulic Simulation of Agricultural Water Distribution and Delivery

The primary objective of this study is the development of a Hybrid Bayesian network for simultaneously simulation and evaluation of water distribution and delivery in a main irrigation canal. To this end, the HBN model was trained and verified by results of the hydrodynamic model and evaluation process. In this regard, mild, medium and harsh canal inflow fluctuation scenarios were considered for performance investigation of canal in mentioned situations. The results of simulation and evaluation processes showed that in decreasing fluctuations, the operation of the canal in adequacy point of view got in trouble. This issue was even more significant in the harsh scenario and downstream canal reaches. The verification results of HBN model demonstrated good performance of this model. Accordingly, it can be concluded that using the HBN model can be a practical step towards fast (computational time-wise) and more efficient (quality-wise) operational and managerial improvements in main irrigation canals.