The current usage of alternative sources is sub-optimal. With this optimization we encourage consumers to use alternative power systems thus further reducing the dependency on heavy carbon based energy

What it does

Inverters decide when to use energy from the grid and when to use from alternative sources. If the decision is a predix-enabled smart decision the cost of a clean energy solution can be reduced, and the grid can benefit from a prediction based on historical data.

How we built it

We based our algorithm on a realistic consumer profile , hourly price and generation profile of a commercial of-the-shelf solar panel system After this we emulated 2 possible algorithms for deciding when to use and when to store. One algorithm is the industry standard not-so-smart inverter strategy, the second algorithm is an optimized smart decision - takes into consideration average hourly price, battery level and overflow ( and many other internal factors ). To make a clear difference between the 2 we graphically represented 2 charts with the consumption profile of each version.

Challenges we ran into

Alternative sources are unpredictable and require storage. Even more so an optimal 'decision' of when to use from the grid and when to store in a battery requires complex analytics which takes into hourly price, consumption forecast, energy generation and capacity.

Accomplishments that we're proud of

  • Our solution helps in minimizing consumer costs by creating an optimized load profile deciding when to switch from renewable sources to Grid usage.
  • Clearly visualization of the difference between the smart and not-so-smart versions.
  • User friendly and responsive interface with grid map, consumers and visualization charts.

What we learned

Costs for implementing a clean energy solution can be reduced by 50% if a SMART solution is correctly implemented

What's next for Smart Inverters Solution

Algorithm can be further improved by using linear optimization analytic ( predix service ) and taking into account weather and solar radiance forecast.

You can see our application using the link bellow. usr/psw: TEAM_31/TEAM_31

Built With

  • analytic
  • asset
  • linear-optimization
  • polymer
  • predix-ui
  • timeseries
  • uaa
Share this project: