Active Disturbance Rejection Control for Time Varying Disturbances: Comparative Study on a DC-DC Boost Converter

Abstract

DC-DC boost converter (DBC) finds essential role in battery-powered applications such as HVDC systems, electric vehicles, micro-grids, robotics but its non-linear and non-minimum phase dynamics presents a challenging control problem. Model based control strategies yield unsatisfactory closed loop performance in the presence of time-varying internal and external disturbances, hence model free control techniques are preferred for disturbance rejection such as active disturbance rejection control (ADRC), uncertainty and disturbance estimator based control. For the above said class of control problems, conventional extended state observer (ESO) fails to reject time-varying disturbances. Compared to a conventional ESO, two of its alternatives namely, cascade extended state observer (CESO) and generalized proportional integral observer (GPIO) disturbance-estimator strategies yield improved regulatory performance. We investigate the efficacy of these two strategies based on their structures to reject the total disturbance which is evaluated on a DBC subjected to input voltage disturbance and load disturbance. This study suggests that GPIO outperforms CESO in terms of disturbance attenuation, reference tracking and regulatory performance.