Design of a NARX-ANN-Based SP Controller for Control of an Irrigation Main Canal Pool

The management of irrigation main canals are studied in this research. One way improving is designing an efficient automatic control system the water that flows through canal pools, which usually carried out by PI controllers. However, since pools systems with large time delays and nonlinear hydrodynamics, these PIs tuned a very conservative so closed-loop instability may appear depending on chosen operation regime avoided. These controllers inefficient because they have slow responses. In order to obtain faster responses remain stable independently regime, combines Smith predictor, appropriate linear delays, NARX artificial neural network (ANN), models dynamics proposed. By applying identification procedures, two NARX-ANN-based mathematical model real pool were obtained. applied implement modified SP controller conventional controller. Experimental results our showed overcomes both setpoint tracking load disturbance rejection.