Inspiration
As Dean Kumar stated, the factors enabling clean energy are dropping in cost with every passing year. Clean energy is the future, but using the technology already available today we are able to save costs for consumers.
What it does
The energy industry provides electrical power to consumers from a variety of sources, including gas-based, hydroelectric plants, nuclear and coal-based power plants.
The price of electricity in any given area is governed by complex dynamics, driven by many factors including day-to-day and seasonal variation in demand, seasonal variation in temperature, availability of electricity from surrounding regions, and cascade effects when plants are shut down.
How we built it
In our model, which we refer to as a dynamic supply-demand model, we simultaneously capture electricity price and usage time series.
This model is built on the basic economic principle that on each day, the price and quantity in a competitive market can be determined as the intersection of supply and demand curves.
The model incorporates temperature, seasonality effects and gas-availability as factors by expressing the supply and demand curves as explicit functions of these factors.
Since the model is nonlinear and non-Gaussian, and the supply and demand curves are not directly observable, traditional methods for parameter estimation and forecasting are not applicable.
Hence, we are using artificial neural networks in this context. We have chosen MATLAB as our platform of choice. The final model is very accurate, with the artificial neural network handling power system analysis and modelling.
Challenges we ran into
MATLAB is trickier to integrate with than Python, so communicating through Node proved difficult.
Accomplishments that we're proud of
The model requires quite a significant amount of data to regularly train on, and as such cannot be loaded onto user phones. Instead, the MATLAB model lives on a remote Windows server and responds to requests from the application handled via a Node.js server.
We are proud to have figured out this solution.
What we learned
We learned quite a bit about app development and inter-service communication by integrating MATLAB with Node.js.
What's next for Electrika Energy Saving App
The predictive capability of the model, enabled by its unique ability to offload compute-heavy tasks to the remote server, can be leveraged in the future for other applications such as an industry-tailored device sitting directly between a machine and its electricity source.
Using predictive capabilities would enable functionality similar to the Amazon AWS model of “spot instance” bidding, for electricity. For periods where usage is predicted to be low, electricity would be permitted to be drawn.
Log in or sign up for Devpost to join the conversation.