Abstract
This paper presents a robust and optimal operation tracking Energy Management System (EMS) for Mobile Base Transceiver Station (BTS) Microgrid equipped with Battery, PV panels and Diesel Engine Generator (DEG) for unreliable grid. The contribution is particularly focused on minimizing the DEG fuel (with DEG ON/OFF frequency) at the time of power-grid outage (i.e. Blackout). The EMS is designed to work with less sophisticated local BTS controller. The prediction (load predictor, PV power generator predictor and blackout duration predictor) modules and the optimization modules are envisioned to be hosted in a server. The optimization module provides optimal battery control policies that are robust against the temporal uncertainty imposed by unknown blackout duration while minimizing the expected DEG fuel cost. The blackout duration predictor is facilitated by combining Linear Logistic Regression classifier and blackout duration frequency distribution. The load predictor utilizes Support Vector Regression for creating predictive models. The optimizer performs a local search over candidate solution space to seek the optimal control policy. Numerical simulation corroborates that the designed EMS, applied on 14 BTS sites in India, is capable of reducing 26% (on average) of DEG fuel compared to that of current operation over 7-Days.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call