Design of robust-fuzzy controller for SMIB based on power-load cluster model with time series analysis

Abstract

Dynamic stability analysis is one of important issue in electrical power system study. This paper aims to analyze and design the control of power generation operation system in small disturbance events. This condition is affected by changes in the prime mover of mechanical power input in generator system due to power fluctuation in load, so that the system becomes unstable. In the analysis of electrical power distribution, generation unit provides power output based on regulations of fluctuating power. This distribution system that provides continuous data periodically is actually performing a pattern of dynamic time series model. Within the statistical methods analysis, the presentation of load data will be analyzed through clustering method based on the average distribution and peak loads. This kind of pattern description is purposed to enable the control system for anticipating the changes in the load model, where each of load cluster represents one dynamic system model in appropriate operation condition. The solution of dynamic models control systems, performed by Takagi-Sugeno Fuzzy Inference System (TS-FIS) as multiple soft switching controllers and optimal control gain for each dynamic model. Those can be distributed into TS-FIS outputs, to achieve robustness of power generation system that affected by changes in huge variation of load power. In this study, the cluster analysis technique has produced seven data's groups with interval of 18 MVA. By performing Robust-Fuzzy control through TS-FIS as multiple soft switching, can be proved that the power generating performance is better than using Linear Quadratic Regulator (LQR) optimal control, since Robust-Fuzzy control Integral Absolute Error (IAE) is better than LQR optimal control IAE.