Optimal control of irrigation canals based on control vector parametrization

Abstract

This paper presents the control vector parametrization (CVP) method to solve the optimal control problem of irrigation canals with boundary constraints of the gate openings. The mathematical model of a multi-pool cascaded canal is built based on the Saint-Venant equations. The formulation of the optimal control problem with the gate opening boundary constraints of the cascaded irrigation canals is then established. The control vector parametrization (CVP) method was then introduced to solve this optimal control problem. Through the CVP method, the original infinite dimensional problem to obtain the optimal control trajectory is reduced to a finite dimensional nonlinear programming (NLP) problem. The computational burden of solving the optimal control problem is reduced greatly, the boundary constraints on the gate openings are treated effectively as well. Thus, the optimal control algorithm for the irrigation canal systems is designed. The proposed algorithm is solved by MATLAB simulations for both a single canal system and a 2-pool cascaded canal. The control performance of the algorithm is implemented and evaluated using a fully nonlinear model SICC.