Micro Grid stability enhancement using SVC with fuzzy model reference learning controller algorithm

Maintaining voltage level stability of islanded mode Micro Grids (MG) is a challenging objective due to the limited power flow between sources and loads. The objective of this work is to enhance the dynamic performance of islanded mode Micro grid in the presence of load disturbance using static VAR compensator (SVC). The contribution of this work is the implementation of PI fuzzy model reference learning controller (FMRLC) to SVC control loop. The control algorithm compensates for nonlinearity possessed by MG where fuzzy membership functions and implication imbedded in both controller and inverse model that achieve better presentation of both uncertainty and nonlinearity of the power system dynamics. Hence, MG keeps desired performance as required irrespective of the operating condition. In addition, learning capabilities of the proposed control algorithm compensates for grid parameter variation even with inadequate information about mathematical presentation of load dynamics. The reference model was designed to reject bus voltage disturbance, created by load and wind variation, with achievable desired performance. Accordingly, SVC with fuzzy controller was able to reject bus voltage disturbance by matching closely the reference model performance. Simulations were carried out to study the steady-state and transient performance of MG in islanded mode. The MG is composed of a PV bus supplied by wind turbine and induction generator, a PQ bus connected to nonlinear dynamic load and linear load, a single distribution line connecting the two buses, and SVC. The proposed control algorithm robustness was tested by providing load disturbance in different operating conditions and observing the system dynamic performance. The performance of the proposed controller is compared to a conventional PID controller using overshoot, transient oscillation, Integral-of-Time Multiplied Absolute Error (ITMAE), and integral square error (ISE) as performance parameters. Both ITMAE and ISE values for the proposed controller were much less than conventional PID controller. In addition, for the proposed controller, ITMAE values sustained stable increase while PID controller ITMAE values increased exponentially. These results indicates the progress achieved by proposed controller to enhance disturbance rejection with time due to learning process.