AICSS: Automatic simulator-controller/optimizer model of open channels

The coupling between simulators and controllers in control systems of irrigation canals has always been a challenging issue. Consequently, simplified or approximated models of Saint-Venant equations (SVEs) have unavoidably been used thus far. In this research, a coupled simulator-controller/optimizer model named Automatic Irrigation Canal Simulation System (AICSS) was developed to control irrigation conveyance system simulation (ICSS) using MATLAB, which allows setting input data, retrieving input or output data and opening, running and closing the ICSS model, enabling any real-time modelling of canals with mathematical and statistical experiments as well as calibration of the discharge coefficient and Manning's roughness. To this end, several submodels were written in FORTRAN to automatically guide orders in ICSS received from MATLAB. The developed model is started by pressing the F5 key in MATLAB. To introduce the AICSS application and evaluate it, the proportional-integral-derivative (PID) controller was used and tested in the E1R1 Dez and Aghili East canals located in Khuzestan province (Iran), showing successful results when simulating and controlling the canals and leading to maximum and average water depth errors of 2.3% and 0.07% in the E1R1 canal and 5.3% and 0.36% in the Aghili East canal, respectively.

Automated summary

This research developed a coupled simulator-controller/optimizer model called Automatic Irrigation Canal Simulation System (AICSS) to control irrigation conveyance system simulation (ICSS). The model allows real-time modeling of canals, mathematical and statistical experiments, as well as calibration. Using FORTRAN submodels written in MATLAB, orders received from MATLAB can be automatically guided in ICSS. The application of AICSS was evaluated using proportional-integral derivative (PID) controller in Dez and Aghili East canals, resulting in successful simulation and control of the canals with low water depth errors.