Canal Controllability Identification Based on Automation Theory to Improve Water Delivery Efficiency in Irrigation Canal Systems

Abstract

Canal controllability is an open and significant topic that addresses the unreliability/uncertainty in water delivery, efficiency, and modernization in the irrigation system. The study developed an analysis tool for canal controllability to estimate controllable performance and reduce the uncertainty of water delivery under steady and unsteady flow. We designed a linear quadratic control algorithm and applied it to the Nongchang test canal to verify the tool reliability and improve efficiency of water utilization. Finally, the effects of hydraulic variables on canal controllability were quantified. The numerical model of unsteady flow showed satisfactory predictions of water level (correlation coefficient, root mean square error, and mean absolute percentage error were 0.929%, 0.293%, and 12.86%, respectively). The Nongchang test canal was rather controllable (controllability indicator 0.265 to 0.279), implying that linear quadratic control algorithm was appropriate. The algorithm performed well under all conditions tested; the maximum absolute error was 3.66%–8.65%. The water level remained stable (deviation usually less than 0.05 m) and water delivery met user demands in terms of flow rate (the measure of performance relative to adequacy was 92.51%–98.94%) with relatively small gate movements. The bottom slope and roughness were the principal contributors to controllable performance, explaining approximately 46% of the variance. Specifically, controllability was highest when the bottom slope was low, and the roughness and side slope were high. We offer a reliable and flexible tool for assessment of canal controllability. Study results demonstrated that stakeholders benefit when conventional canal operation is modernized. The work will guide automation and upgrading of canal hardware as irrigation becomes optimized.