This study tackled following two issues:

1) developing robust operational planning methodology to compensate forecast error of energy demand,

2) arranging an environment for proof-of-concept demonstration of residential energy management scheme that this study proposed.

This study adopted the concept of receding horizon for the operational planning problem, in order to use energy demand forecast scenarios with high accuracy by shortening forecast lead time. The numerical case study shows that the receding horizon control with 15-minutes forecast/operational-strategy update interval achieves 9.5% reduction of primary energy consumption compared with the case with 24-hours update interval.

The operational planning problem in previous work minimizes expected value of disutility meaning primary energy consumed in a household when energy supply meets plausible energy demand scenarios forecasted. Here, this study considers two problem settings in order to derive more robust operational strategy: one is to employ expected value and its variance as an objective function; the other is to increase number of scenario forecasted for operational planning problem. As a result, latter case with 9 scenarios shows 3% reduction of primary energy consumption in comparison with the case with only one scenario.