Novel genetic algorithm for scheduling of appliances

The introduction of smart metering has brought more detailed information on the actual load profile of a house. With the ability to measure, comes the desire to control the load profile. Furthermore, advances in renewable energy have made the consumer to become supplier, known as Prosumer, who therefore also becomes interested in the detail of his load, and also his energy production. With the lowering cost of smart plugs and other automation units, it has become possible to schedule electrical appliances. This makes it possible to adjust the load profiles of houses. However, without a market in the demand side, the use of load profile modification techniques are unlikely to be adapted by consumers on the long term. In this research, we will be presenting work on scheduling of energy appliances to modify the load profiles within a market environment. The paper will review the literature on algorithms used in scheduling of appliances in residential areas. Whilst many algorithms presented in the literature show that scheduling of appliances is feasible, many issues arise with respect to user interaction, and hence adaptation. Furthermore, the criteria used to evaluate the algorithms is often related only to reducing energy consumption, and hence CO2. Whilst this a key factor, it may not necessarily meet the demands of the consumer. In this paper we will be presenting work on a novel genetic algorithm that will optimize the load profile while taking into account user participation indices. A novel measure of the comfort of the customer, derived from the standard deviation of the load profile, is proposed in order to encourage the customer to participate more actively in demand response programs. Different scenarios will also be tested.