A Simulink-Based Scheme for Simulation of Irrigation Canal Control Systems

Abstract

Computer simulation is the main tool in the design and analysis of control systems for irrigation canal automation, which involves several disciplines and backgrounds. While hydraulic engineers are more familiar with equation-oriented, highly nonlinear, distributed parameter models, control engineers are usually comfortable with linear models and block-oriented descriptions of systems. This article presents a Simulink-based scheme for the simulation of irrigation control systems, which combines library blocks with user-custom blocks, fitting different user preferences and reducing implementation complexity.

Fault detection using interval LPV models in an open-flow canal

“…The real canal behaviour is accurately modelled by SaintVenant's equations (Chow, 1959) and reproduced using a highfidelity simulator developed by the group of ''Modelling and Control of Hydraulic Systems'' at the UPC (Bolea & Blesa, 2000;Mantecó n, Gó mez, & Rodellar, 2002). This simulator solves numerically the Saint-Venant's equations as usually done in existent commercial open-flow canal simulators (Malaterre, 2006).…”

Distributed model predictive control of irrigation canals

“…Control objectives: The control objectives of the MPC controller are typically to regulate (downstream) water levels [2][3][4]13,26], to minimize costs on water supply, treatment, and elevation [5], to minimize costs on pressure regulation, flow regulation and water quality [5], to make sure that the right amount of water is at the right place [8,40], to ensure that operational spills are avoided [40], and to make sure that reservoirs are emptied as fast as possible [3]. Such control objectives are usually rephrased as objectives with respect to water levels staying close to a given reference set-point [2,4,24,26,30,39,40], input values being minimized [24], or changes in inputs being minimized [2,4,30,39,40]. Control measures: To achieve the control objectives, the available control measures have to be manipulated.…”

“…Control measures: To achieve the control objectives, the available control measures have to be manipulated. Usually the control measures consist of the (upstream) discharges through gates [2,5,13,26,40], although in some cases the gate openings [3,4], the water flow towards reservoirs [8], or pump flows [5] are controlled. Prediction models: For making the predictions of the behavior of particular water systems a wide variety of prediction models has been considered in the literature, including an analytically obtained linear time-invariant statespace model derived from Saint Venant's equations discretized through the Preismann implicit scheme [24], a linear time-varying state-space model [3], polynomial models based Diophantine equations [2,30], a nonlinear discretetime state-space model [5], a black-box identified linear model transferred into linear state-space form [4], a linear model known as the Muskingum model in combination with a discretized version of a differential equation for representing storage [13,26], the integrator delay model [39,40], and artificial neural networks using radial basis functions [8].…”

“…Water system: There are various types of water systems for which MPC has been employed, such as irrigation canals [2,4,24,26,30,39,40], river systems [3], water supply networks [8], and water distribution systems [5]. Test canals: The performance of the MPC approaches has been illustrated on various test canals, including the ASCE test canal 1 (first 2 canal reaches [24], the first 3 canal reaches [30], and the complete canal [39]), ASCE test canal 2 [2,13,26], a model consisting of large portions of the Arizona Canal, the Grand Canal, and the South Canal in Arizona [40], a model of the river basin Demer in Belgium [3], and a practical prototype lab setup [4] Simulation packages: Models of the test canals have been implemented using different simulation packages, such as SOBEK [3,39,40], SIC [24,30], and Simulink [13,26]. Performance criteria: In order to evaluate the performance of the controllers, different performance criteria have been considered, such as the performance indices defined by the ASCE committee [2,24] (including the maximum absolute error, the integrated absolute error, the steady-state error, and the change in integrated absolute discharge).…”

Model predictive control of an experimental water canal

“…The Saint-Venant hydraulic equations are known to hydraulic engineers as being mostly accurate [16]. Since these equations are hyperbolic partial derivatives, whose solution in real geometry is generally not known, a finite differences scheme is normally used, typically the Preissmann scheme [17]. The integration of this multivariable prediction model into the MPC scheme will enable the design of a realistic controller that is able to consider known in advance water offtakes, while guaranteeing the overall physical constraints satisfaction.…”