Title: Optimizing Indicator Choosing for Canal Control System and Simulation Study Authors
|Guanghua, Guan -|
|Changde, Wang -|
Submitted to: Meeting Proceedings
Publication Type: Proceedings
Publication Acceptance Date: April 1, 2011
Publication Date: June 1, 2011
Citation: Guan, G., Changde, W., Clemmens, A.J. 2011. Optimizing Indicator Choosing for Canal Control System and Simulation Study. Proceedings 2011 IEEE International Conference on Networking, Sensing and Control (ICNSC 2011), 11-13 April, Delft, The Netherlands, pp. 192-197. Interpretive Summary: Automatic control of irrigation canals is one method for conserving water supplies. Downstream-water-level-control techniques can maintain constant deliveries to users by adjusting gates so that canal inflow equals outflow. Poor design of downstream controllers can make downstream control ineffective, or cause canal overtopping or lining failure. In this paper we examine the criteria used for controller design and suggest a new approach based on a combination of indicators. This should improve the overall performance of downstream canal control. These results will be of use to irrigation and large water districts, the Bureau of Reclamation, and consultants.
Technical Abstract: One Key problem for canal system control is how to select appropriate performance indicators and how to tune the controller with these indicators. A canal system is a multi-input and multi-output (MIMO) system. The judging of control performance can be extremely complicated. In this paper, frequently used canal system performance indicators are evaluated. Then these indicators are used individually to tune the control parameters for a simple proportional-integral (PI) controller on a test canal. Then we evaluated these controllers on this test canal through simulation. Based on these results, we develop a single performance indicator that is a combination of the individual indicators. For this test case, it is shown that this indicator represents a good compromise. It is hoped that this approach can be used to develop a more robust method for determining criteria for controller tuning.