A comparison of power plant energy flow prediction using real data applied to linear programming and if-else algorithms

Abstract

In this paper, the energy flow prediction of a grid-connected Hybrid Microgrid (HMG) is studied. The studied system consists of a 19 kW photovoltaic (PV) array, a 6.5 kW wind turbine (WT), and a 59.32 kWh battery Energy Storage System (ESS). Two energy prediction and optimization algorithms, Linear Programming and If-Else (LP-If-Else), are applied to analyze the site's energy flow behavior and accurately predict it one day in advance. Meteorological data collected at the site and real load profiles are used in this study. To highlight the effectiveness of the offline prediction approach, three extreme cases are applied to three energy management strategies and compared. The results of the prediction performed by the two algorithms showed that a single prediction algorithm (LP) is not sufficient to accurately predict the energy flow for the next day.