Observer based nonuniform sampling predictive controller for a solar plant

The work presented in this paper exploits the transport characteristic of a solar plant where the transport velocity (a flow) is the manipulated variable, i.e. the control input. The solar field is modelled by a partial differential equation. A non-uniform sampling in time is performed in order to obtain a discrete linear model. Due to the transport dynamics of the plant the resulting transfer function has a finite impulse response and the optimal control derived from a black-box approach of such a systems yield pure feed-forward compensators. The main contribution of this paper is the use of a state-space description of the plant in conjunction with the nonuniform sampling that allows to introduce the feedback mechanism through the state observer. The controller results from the optimization of a multi-step quadratic cost function. Experimental results performed with the solar plant are shown.