Soft Computing Technique-Based Voltage/Frequency Controller for a Self-Excited Induction Generator-Based Microgrid

Abstract

Microgrids (MGs) are small scale energy unit networks that can offer an adequate energy supply to cover local demand by incorporating renewable energy and storage technologies. The system capacity is generally between several kW to several MW. They work in terms of low voltage (LV) level or medium voltage (MV) level. They can also be connected/disconnected from main grid whenever it is necessary. This paper presents a comparison of two soft computing (SC) techniques fuzzy logic (FL)/artificial neural network (ANN) over a conventional proportional integral (PI)-based voltage frequency controllers used for improving the performance of MG under islanding mode. Microgrid is formed by using three 7.5[Formula: see text]kW, four pole, 50[Formula: see text]Hz, self-excited induction generators (SEIGs) driven by small hydro turbine feeding three-phase four-wire consumer load. The proposed topology functions excellently in maintaining phase angle, voltage and frequency (VF) regulation of the micro sources (MSs) in islanded mode as well as in resynchronization when one of the MSs is turned off due to fault or unavailability of resources. The conventional PI controller is replaced by a controller based on SC techniques, as it has disadvantages like explicit description of mathematical model, affected by variations in consumer loads and sources, thus the proposed SC techniques enhance the performance of VF controller. A comparative analysis of PI/FL/ANN controller is also carried out to highlight the superiority of AI controller. The performance of controller with proposed configuration is verified for balanced/unbalanced non-linear load. Microgrid and control schemes are simulated in MATLAB Sim Power Systems environment.