Design of Observer-Based Robust Power System Stabilizers

Abstract

Power systems are subject to undesirable small oscillations that might grow to cause system shutdown and consequently great loss of national economy. The present manuscript  proposes two  designs for observer-based robust power system stabilizer (PSS) using Linear Matrix Inequality (LMI) approach to damp such oscillations. A model to describe power system dynamics for different loads is derived in the norm-bounded form. The first controller design is based on the derived model to achieve  robust stability against load variation. The design is based on a new Bilinear matrix inequality (BMI) condition. The BMI optimization  is solved interatively in terms of Linear Matrix Inequality (LMI) framework. The condition contains a symmetric positive definite full matrix to be obtained, rather than the commonly used block diagonal form. The difficulty in finding a feasible solution is thus alleviated. The resulting LMI is of small size, easy to solve. The second PSS design shifts the closed loop poles in a desired region so as to achieve a favorite  settling time and damping ratio via a non-iterative solution to a set of LMIs.  The approach provides a systematic way to design a robust output feedback PSS which  guarantees good dynamic performance for different loads. Simulation results based on single-machine and multi-machine power system models verify the ability of the proposed PSS to satisfy control objectives for a wide range of load conditions.