Intelligence Optimization in Parameter Identification of the Border Irrigation Model

With the aim of estimating infiltration properties of surface irrigation and further saving water efficiently, a zero-inertia model was adopted for simulating the surface flow of border irrigation. The parameters identification of the model has been derived from hybrid volume balance model coupling artificial neural networks and numerical inversion approaches including differential evolution. With some special treatments to the advance and/or recession fronts of surface flow as its kinematical boundary, the discretization and/or the further linearization of zero-inertia model have been solved through the Newton-Raphson method and the pursuit algorithm. The validations of the identification of parameters and/or the model were verified by comparing the simulated data with measured and/or recorded data for advance or recession phase of border irrigation. The result shows that the optimization algorithm and/or model are appropriate and accurate.