Automated demand side management For pv enabled residential home Aiding in demand response

Abstract

The integration of renewable energy sources, primarily newlinephotovoltaic (PV), for a residential home is quite suitable as its power newlinegeneration profile suits the energy requirement profile very well. This newlineintegration balances both PV energy generated and the residential home newlinedemand with minimum grid power drawn, leading to a reduction in the newlineelectricity bill. With active participation in demand response programs under newlinedemand either at PV available hours or low tariff hours. newlineThis thesis proposes an energy-demand management scheme newlineconsisting of schedule pattern generation algorithms enabled with resident newlinecomfort window identifier and solar irradiance forecaster for automating newlinedemand-side management for a PV enabled residential home aiding in newlinedemand response. newlineThe resident comfort window identifier employs an artificial neural newlinenetwork technique to identify the end-user preferable operational time slots newlinefor an appliance operation for the next day. The solar irradiance forecaster newlineuses the random decision forest-based machine learning algorithm to forecast newlinethe next-day irradiance for generating the PV power profile through the newlineMATLAB Simulink model. The schedule-pattern generation algorithms newlinefacilitate the residential user in planni - newlineeffectively at suitable time slots of day-ahead PV power profile and dayahead newlineelectricity tariffs within identified resident comfort windows. newlineThe results exhibit that the proposed scheme within a PV enabled newlineresidential home maximizes the benefits of renewable energy sources, newlineminimizes the total energy consumption cost, and minimizes the peak newlinedemand while preserving user comforts. newline