Model Predictive Control of an Irrigation Canal Using Dynamic Target Trajectory

Abstract

Model predictive control (MPC) has been extensively studied for controlling irrigation canals. Most of the studies use lumped models and fixed target for the controller. This setup works fine if the water delivery is on demand. However, in the case of on-supply delivery, balancing mismatches between demand and supply and spreading mismatches among all users are crucial. A fixed control target will not meet the request due to the mismatches. In this paper, a dynamic target trajectory approach is proposed to calculate changes of control targets that are used by MPC. The dynamic target trajectory calculates the percentage change of each setpoint based on the total volume mismatch spreading over the available capacity in each canal pool. The approach is applied to the Central Main Canal in Arizona. The results demonstrate that the dynamic trajectory can help the optimization find the optimum more quickly and facilitate tuning of the controller. Moreover, the mismatches can be better distributed over canal pools.