Multi-objective Optimization Technique for Droop Controlled Distributed Generators in AC Islanded Microgrid

Abstract

• This work proposes a novel droop-controlled optimal controller to regulate multi-load-bus voltage and accurate reactive power sharing in islanded ac microgrid. • Real power from each distributed generation unit is considered as equality constraints of quadratic programming. • One important aspect of proposed methodology is that, only voltage magnitude information is required to send at ith inverters local controller. • Stability analysis also investigated through small signal model. Accompanied experiments validates the MATLAB/Simulink results. The subject of an optimal control of distributed generation (DG) units in AC islanded microgrid (IMG) has gain an important research attention in a recent time. Such control method promises voltage regulation and resolve inaccuracy of power sharing caused by high loads and line differences in a droop-controlled IMG. This research work proposes a multi-objective optimization droop control strategy, where three objective functions corresponding to: (i) load-bus voltage regulation; (ii) accurate real power sharing; and (iii) reactive power sharing, are constructed to improve the system performance and stability in IMG. The numerical approximation method through Lagrange multiplier is employed, subject to equality constraint function. An important aspect of proposed solving algorithm related to objective function is the voltage magnitude information only required to be transferred to each inverter's local voltage controller. Further, a supervisory secondary control layer is used to regulate the system frequency deviations. The results obtained from MATLAB/Simulink and experiment show the effectiveness for optimizing the reactive power injected by each DG unit and load-bus-voltage, under series of parameters uncertainties. Modeling and analysis are investigated through improved mathematical small signal models.