| In the face of limited water resources, better utilization and operation of gravity irrigation systems is essential. The general goal of this research is to improve and upgrade the operation of irrigation systems by using an automatic controller. I developed a downstream canal control algorithm that restores flow conditions by compensating for flow and volume changes in a canal system. The proposed algorithm uses feedback control with multiple inputs and outputs. The proposed algorithm was tuned and tested using the ASCE task committee on canal automation guidelines.;The test case canals and flow scenarios were simulated using CANALMAN a one-dimensional simulation model that employs an implicit, finite difference scheme to solve the unsteady state flow problem as described by Saint Venant. The control parameters of the proposed algorithm were tuned using genetic algorithm an evolutionary optimization technique.;The tuning process yielded better results for the shallow sloped canal scenarios than the scenarios for the steep canal. The algorithm performance resulting from steep canal scenarios indicates that the algorithm is not as reliable for steeper canals as it is for shallow sloped canals.;Some of the more abstract methods perform better than the algorithm presented here. However, this algorithm is adequate for most canal control problems, and it does present advantages that may make it appropriate for many applications. It also introduces the use of evolutionary algorithms for canal control, a class of methods with great potential for this problem.;Using a more rule-based algorithm teamed with a learning classifier system rather that the tuning technique presented here could provide much-improved dynamic performance. |
