Model-based dynamic sliding mode control and adaptive Kalman filter design for boiler-turbine energy conversion system

Abstract

The model-based control of a boiler-turbine system (BTS) is a formidable task due to coupling in state variables, nonlinearities and constraints on the control inputs. In this paper, a model-based, multi-variable dynamic sliding mode control (DSMC) is designed for the nonlinear BTS model to maintain the drum pressure, electric power and water level at the desired levels. In DSMC, an implicit sliding manifold is designed for the water level due to its complexity and explicit dependence on the control inputs. For this purpose, an auxiliary function is computed, and the sliding mode is enforced in such a way that the system fluid density tracks the auxiliary function, and subsequently the water level tracks the desired trajectory. Owing to its complexity, the time derivatives of the auxiliary function are computed using the uniform robust exact differentiator (URED). An adaptive Kalman filter (AKF) is designed for the estimation of the unmeasurable state i.e., system fluid density. The design of AKF is based on the quasi-linear model of the BTS. Furthermore, a detailed stability analysis is carried out to ensure the boundedness of the closed-loop system. The simulation results depict that the designed control scheme exhibits the desired tracking performance in the presence of external disturbances, nonlinearities, constraints on the inputs, and measurement and process noises.