Parallel operation of inverters by using Model Predictive Control in Islanded Microgrid

Abstract

With the energy demand of the world increasing day-by-day, aging the present power system infrastructure and the increasing integration of renewable energy resources in quenching the load demand, the development of microgrids (MGs) is imminent. The performance analysis of AC MG and DC MG is made based on the concentration of the type of distributed energy resources (DER) in a development, availability of energy storage systems (ESS), use of converters and the economic/physical feasibility of connection and synchronization with the utility grid. The importance and demand of renewable resources has been rising in wake of the rising fuel costs and limited availability of conventional fuel sources demand the increased penetration of solar PV and wind energy into the grid. This can be implemented with the help of parallel inverter-based MG technology, which can not only help integrate the renewable energy into the main grid, reducing the load on fossil fuels, but also increase the reliability of electric power supply as MGs can operate in stand-alone mode as well, independent of the grid. This requires robust, fast and precise control system algorithms which can control the AC bus voltage and frequency even with variable loads associated with the MGs. Two different control algorithms, conventional droop control and Model Predictive Control (MPC) based droop control can be implemented for independent control of the MG. In MPC based operation, the voltage and frequencies are predicted and comparing it through a cost function with the reference value generated, the appropriate switching pattern of VSI is generated to produce the output voltage. In this paper, these two control methods have been implemented and their performance with active and reactive power sharing are compared