Pseudo-Linearization Observer-Based Tracking Control of Boiler-Turbine-Generator System Using Linear Matrix Inequalities

This paper deals with the problem of controller and observer design for the nonlinear, highly coupled constrained boiler-turbine-generator system. The input vector (control valve state) has upper and lower limits, which poses a challenge in the controller design process. Moreover, the nonlinear relationship between system output and the inputs and state variables in the observation model, makes the design procedure more challenging. To solve these problems, extended pseudo-linearization technique is utilized, which simplifies the controller design and avoids instability. Since, the resulting pseudo-linear model has state-dependent coefficients, a new state-dependent observer is also designed in this paper. To ensure the stability of the proposed method, a time-varying Lyapunov function is defined to guarantee convergence of both state estimation and tracking errors, throughout Linear Matrix Inequalities (LMIs). Simulations are carried out on a 160-MW boiler-turbine-generator unit to validate effectiveness and stability of the proposed method, confirming control signals saturation limit satisfaction.