Robust Controllers for Precision Power Sharing among DG Units of Islanded Microgrid Under Load Variations

Abstract

Islanded microgrids are prone to frequent fluctuations in power sharing among distributed generation (DG) units. This is primarily due to the microgrids inherent characteristics, such as load uncertainty, the lack of accurate control methods, and the variable generation patterns of renewable energy sources. The droop control method combined with a Proportional-Integral (PI) anti-windup technique alone is insufficient for ensuring proper power sharing among DG units. It also struggles to maintain standard voltage and frequency levels. This research proposes a droop control approach based on real power and frequency, combined with higher-order sliding mode control and a single hidden layer feed-forward neural network. The higher-order sliding mode controller is responsible for regulating the voltage, while the single hidden layer feed-forward neural network is employed to control the current. The controllers are designed to ensure equal power sharing among DG units even as the load varies. Experiments are conducted under three different loading conditions to evaluate the stability of power sharing among the DG units. To verify the effectiveness of the proposed method, the real and reactive power sharing between DG1 and DG2 is compared using both the existing and proposed controllers across each case. Additionally, to assess the robustness of the controller, voltage and frequency comparisons are carried out under time-varying load conditions against other controllers. The proposed microgrid is modeled and simulated in the MATLAB/Simulink environment. The simulation results demonstrate that the proposed controllers significantly enhance the performance of the droop controller and achieve equal power sharing among DG units despite load variations.