Parameter and State Estimation of Managed Pressure Drilling System Using the Optimization-Based Supervisory Framework

This brief proposes a method for the simultaneous estimation of unknown parameters and unmeasured states of nonlinear continuous-time systems. The proposed methodology has been inspired by the supervisory estimation approaches that use a bank of observers and are designed based on a set of fixed nominal parameter values. Consequently, the methods have a high computational load, and the estimated parameters marginally converge to the true parameters. For the reduction of computational load, a new objective function with a lower computational load that is computed during a sliding time window is proposed. Then, an observer with an optimization-based framework is introduced for nonlinear systems with affine unknown parameters, that replaces the bank of observers with a single observer. The unknown parameters update is based on the defined objective function. The performance of the method is illustrated in a simulation study of a managed pressure drilling (MPD) system. In the MPD process, downhole measurements encounter problems such as transmission delay and slow sampling rate. Also, the drilling process has parametric uncertainties due to the unknown friction coefficient. The simulation results indicate that the simultaneous estimations of the friction coefficient and drill bit flow rate in the MPD system converge to their true values in finite time.