Scheduling Smart Home Appliances in the Stockholm Royal Seaport

This thesis investigates the optimal scheduling of smart home appliances with respect to economic benefits (electricity bill) and reducing environmental impacts (CO2 emissions) for the Stockholm Royal Seaport project which is a new urban district developing in the eastern Stockholm which will house 10,000 new apartments and 30,000 new office spaces where modern living is combined with environmental thinking to create sustainable living. Previous work used a time slot based approach to schedule the smart home appliances with respect to the electricity price which concluded that the solve time rapidly increased as the number of appliances increased. The smart home control devices that will be used in the Stockholm Royal Sea project apartments will have a CPU and memory similar to those of a smart phone. Therefore there exists a need for faster implementation. The previous work will be extended by including CO2 emission intensities. An alternative method using a piecewise linear function (PLF) based approached is introduced and evaluated for faster implementation of the scheduling of the smart home appliances. Technical and operation specifications of the smart home appliances such as sequential processing are enforced into the optimization framework as well as consumer preferences. The results of the tradeoff between electricity bill and CO2 emissions are studied through a Pareto frontier. A well balanced result between minimizing the electricity cost and reducing the CO2 emissions for an unusual cold day in Sweden (2010-01-05) with three typical home appliances showed that for the time slot based approach one could save up to 35.8 % of electricity costs as well as reducing the CO2 emissions with up to 16.6 %, while for the piecewise linear function approach one could save up to

34.6 % of electricity cost as well as reducing the CO2 emissions with up to 15.9 %. These savings are with respect to the worst case scheduling for respective approach. After evaluating both methods one could show that the solve time of the piecewise linear based approach is about fivefold faster than the time slot based approach. Finally a variant of the Ramer-Douglas-Peucker algorithm is applied to further reduce the solve time of the piecewise linear function based approach and finally tradeoffs between the approaches are discussed as well as extensions of the current framework.