Improved Proportional Integral (PI) controller for water level control in open channel systems: A case study of the Middle Route Project for South-to-North Water Transfer

The three pools near the terminal of the main canal of the Middle Route Project for South-to-North Water Transfer in China. The conventional Proportional Integral (PI) controller neglects the potential for improved control performance by utilizing a PI controller with adaptable control parameters based on different water level conditions, particularly in large-scale open channels with significant water level variations. To address this limitation, this study proposes a novel hierarchical Proportional Integral (H-PI) control approach. The H-PI approach incorporates a weight coefficient into the control logic of the PI controller, allowing flexible adjustment of control parameters. These parameters are tuned using a hydrodynamic model integrated with a genetic algorithm for optimization. The simulation results of the study area manifest the efficacy of the proposed method in substantively diminishing gate control frequencies in contrast to other methods, specifically for gate deadbands of 3 cm and 6 cm, respectively. Moreover, the proposed method exhibits higher sensitivity to significant water level deviation changes. The robustness of the proposed method is evaluated under different disturbance scenarios, confirming its resilience and practical feasibility for implementation in the study area. In summary, the proposed H-PI control approach offers distinct advantages in minimizing water level deviation and gate control frequency. Furthermore, its robustness under different disturbance scenarios further supports its practical utility. [Display omitted] •A novel hierarchical PI controller (H-PI) is introduced for water level control.•The controller incorporates a optimization model to fine-tune its parameters.•The controller effectively control water levels even with significant gate deadband.•The controller demonstrates excellent performance and robustness.