Supervised Gain-Scheduling Multimodel versus Linear Parameter Varying Internal Model Control of Open-Channel Systems for Large Operating Conditions

Abstract

In this paper, two internal model control (IMC) controllers using gain-scheduling techniques are proposed and compared for open-channel systems that allow to deal with large operating conditions. In particular, in one side, a linear parameter varying (LPV) model for an open-flow channel system based on a second-order delay Hayami model is proposed. This model will allow one to design a classic gain-scheduling strategy for the IMC controller. On the other side, the LPV model is discretized in a set of linear time invariant (LTI) models corresponding to different operating points. For each LTI model a LTI IMC controller is designed off-line. Then, a supervised gain-scheduler detects on-line which is the LTI model that represents better the open-flow channel system at the current operating point and decides which is the LTI controller that should be used. Finally, both approaches will be applied to a simulated open canal: the Lunax gallery located at Gascogne, France.