Prototype Measurement System: To better understand how energy is
generated and consumed in homes powered by renewable sources, we have
deployed the first version
of a measurement system in an off-grid home in Arkansas. We collect
the instantaneous residual battery voltage (in Volts) and the energy
consumed by the house (in KWh). We have also deployed networked,
off-the-shelf WattsUp meters to measure power consumption of the
television and refrigerator in the home. Analysis of the data suggests the following:
- Finding 1 - Traditional energy management techniques are insufficient: We first explore whether
established norms of energy management, for example the 7PM-7AM policy
that recommends users run important household appliances between 7PM and
7AM when the demand for power is low, are appropriate for green homes.
We find that, though energy consumption varies considerably over 24
hours, a large portion of energy is consumed between 10AM and 8PM.
Conversations with the home residents indicate that this consumption
pattern is largely driven by the residual energy in the batteries; the
residents try to use energy during the day when the most energy is being
generated. These findings demonstrate a need to develop new energy
management techniques that consider the unique energy harvesting
conditions in homes powered by renewable technologies.
- Finding 2 - Energy generation and consumption is both variable and predictable:
We next observe that both energy harvested from the panels and energy
consumed by the house is highly variable, yet predictable. There is
variance in generation and consumption across a single day, across
several days, and across seasons. This suggests that fixed energy
management strategies are insufficient and adapting to variability is a
key element for green homes. Moreover, we observe that there is
considerable predictability in the data, pointing to the feasibility of
automated and proactive energy management schemes.
|
 |
|
|
Recommendation System and Smartphone Application: We are
developing a recommendation-based system that uses information collected
by the measurement system to provides users with advice regarding
energy consumption, including warnings in advance of critical battery
situations, recommendations for the best times to execute high-power
tasks such as running a clothes dryer, and opportunities to adjust the
power states of devices to reduce energy consumption. We have chosen a
recommendation-based model to minimize user irritation and ensure that
control of household appliances ultimately resides with the user. A
smartphone application notifies the user when the system suggests
changes to the power states of devices, for example suggests that the
user turn an appliance on. The user is responsible for implementing the
suggestion, and control of some devices is supported via the smartphone
application. |
|
|
