Real-time Energy Management for PV–battery–wind based microgrid using on-line sequential Kernel Based Robust Random Vector Functional Link Network

Abstract

This paper comprises of an effective Local Energy Management System (LEMS) implementing the Generalized Power Prediction Model (GPPM) for the uncertain performance of various Distributed Generations of a microgrid. The primary DGs considered to compensate intermittency is Wind Power Generation System (WPGS) and Photovoltaic (PV) along with Battery Energy Storage (BES). Solar and wind power are considered here as prediction targets and to cope with the intermittent nature of the Renewable Energy Sources (RESs), Power calculation from the RES data is included in the proposed prediction model. In order to establish arobust reduction in the predicted error, areal-time Online Sequential Kernel based Robust Random Vector Functional Link Network (OS-KRRVFLN) prediction algorithm is developed. Here, the initial randomness of the proposed OS-KRRVFLN is reduced (robust performance) by Hampel’s cost function (eP based) minimization. The PV-BES and WPGS based DGs are designed to be operated under certain uncertainties like PV arc, bus faults etc. as uninterrupted power supply with minimum grid power dependency. An efficient Distributed Adaptive Droop (DAD) control is thus opted as Primary Controller (PC). Addictiveness of DAD’s operational region is estimated while compensating eP”-. The coordination of GPPM with DAD-PC based accurate reference estimation for Independent DG Controllers (IDGCs) operation is considered as LEMS operation.