Learning-based Parameter-Adaptive Reference Governors

Reference governors (RGs) provide an effective method for ensuring safety via constraint enforcement in closed-loop control systems. When the parameters of the underlying systems are unknown, but constant or slowly-varying, robust formulations of RGs that consider only the worst-case effect may be overly conservative and exhibit poor performance. This paper proposes a parameter-adaptive reference governor (PARG) architecture that is capable of generating safe trajectories in spite of parameter uncertainties without being as conservative as robust RGs. The proposed approach leverages on-line data to inform algorithms for robust parameter estimation. Subsequently, confidence bounds around parameter estimates are fed to supervised machine learners for approximating robust constraint admissible sets leveraged by the PARG. While initially, due to the absence of on-line data, the PARG may be as conservative as a robust RG, as more data is gathered and the confidence bounds become tighter, such conservativeness reduces, as demonstrated in a simulation example.