Human-in-the-loop control of an irrigation canal using time instant optimization model predictive control

In the paper we discuss the recently introduced Mobile Model Predictive Control (Mobile MPC) approach for an irrigation canal. Mobile MPC is a configuration of MPC that explicitly incorporates the role of the human operator traveling between the gates as ordered by a remote centralized controller. The operator provides the controller with up-to-date measurements from the locations visited and acts as the actuator as required by the remote controller. Mobile MPC provides a solution in between fully manual and fully automatic canal operation, as the first one may give poor performance and the second one might be impracticable in some situations, where it is not possible to rely on the equipment installed in the field. In the current paper we improve the performance of the original Mobile MPC approach by allowing the controller to decide the exact time instants when the operator should arrive at a specific gate and change the gate's settings as well as we include a penalty in the objective function for the controller to minimize the workload of the human operator. We show that the new approach yields enhanced performance in comparison to the previous method, and we demonstrate the benefits of the new method as opposed to the previous one in a case study.