Hybrid Data-enabled Predictive Control: Incorporating model knowledge into the DeePC

Abstract: Predictive control can either be data-based (e.g. data-enabled predictive control, or DeePC) or model-based (model predictive control). In this paper we aim to bridge the gap between the two by investigating the case where only a partial model is available, i.e. incorporating model knowledge into DeePC. This has potential advantages over a purely data-based approach in terms of noise and computational expense in some cases, as well as applications to certain linear time-varying and nonlinear systems. We formulate an approach to take advantage of partial model knowledge which we call hybrid data-enabled predictive control (HDeePC) and prove feasible set equivalence and equivalent closed-loop behavior in the noiseless, LTI case. Finally, two examples illustrate the potential of HDeePC.