Challenges in Model Predictive Control Application for Transient Stability Improvement Using TCSC

摘要

Performance of a system is always dominated by constraints rather than dynamics. Conventional controllers determine off-line, a feedback policy that provides optimal control action based on minimization of one or more cost functions with or without constraints using linear or nonlinear model of the system. Increased system complexity and demanding performance requirements have rendered classical control laws inadequate in spite of their simplicity as in case of PID loop or robustness in case of H_2 or H_∞ control designs. The only generic control that can meet these challenges is Model Predictive Control (MPC). The most challenging MPC application would be maintaining stability after large disturbances in highly nonlinear, complex and hybrid system such as power system. The practical difficulties are due to large system having strong interactions in various parameters with hard constraints. The paper discusses various issues in tuning and maintaining constraints on states and control variables for a Single Machine Infinite Bus (SMIB) system using Thyristor Controlled Series Compensator (TCSC) as a controller.