On physical and data driven modelling of irrigation channels

In this paper we test the St. Venant equations against real data in order to examine their accuracy and capability to describe the relevant dynamics of an irrigation channel. The St. Venant equations are simulated using the Preissmann scheme, and the simulated and real measured water levels are compared. Some of the parameters in the St. Venant equations are not exactly known and these parameters are also estimated from the observed data in order to improve the accuracy. Finally, a comparison with system identification models is performed in order to examine which model type is more suitable for control design and prediction purposes. The results show that the St. Venant equations can adequately capture the dynamics of real channels, and naturally, the model with estimated parameters is more accurate than the one with physical parameters. However, system identification models are as accurate as the St. Venant equations with estimated parameters, and they are preferred over the St. Venant equations for control and prediction purposes since they are much easier to use.