Robust control design considering time delay for wide area power systems

This thesis concerns robust control technology as applied to wide area power systems. The main application is to enhance system stability with a Supervisory Power System Stabilizer (SPSS). With the power industry deregulation, many tie lines between control areas are being operated near their maximum capacity. Stressed operating conditions can increase the possibility of inter-area oscillations and may cause breakup of the system. The centralized SPSS using system-wide data with the enhanced Phasor Measurement Units (PMUs) is suggested to improve the dynamic performance of large interconnected power systems.; The delay of a feedback signal in a wide area power system is usually considered to be in the order of 100ms. Long time delays may be detrimental to system stability and may degrade system performance. The long time delay and its uncertainty are modeled using Pade approximation and Linear Fractional Transformation (LFT). Time delay tolerance of the SPSS and the associated performance trade-off are analyzed using a new criterion based on the small gain theorem. Special attention is paid to the choice of weighting functions in the SPSS design. It is found that time delay tolerance decreases when the system bandwidth increases, while the nominal system time domain performance is concomitantly improved. Several approaches that can maintain good system performance while increasing the time delay tolerance are suggested and compared. A modern controller design technique, Gain Scheduling (GS) via Linear Matrix Inequalities (LMIs), is evaluated for the design of the SPSS accounting for various time delays.; Because of the limitations of available commercial software, a program in MAT LAB Simulink is developed for nonlinear simulations of the power system with the proposed robust controller. The nonlinear model simulation results are consistent with those from commercial software and the small signal model. Small disturbance and transient cases can both be simulated. A high order centralized Multiple Input and Multiple Output (MIMO) controller can be fully implemented in the nonlinear model. The nonlinear simulation results validate the effectiveness of the SPSS in steady and transient states. This program also offers users a lot of flexibility to accommodate new components and can be applied to test other control schemes and power electronic devices.