Power Laws: Optimizing Demand-side Strategies

Storage is critical to flexible and reliable access to renewable energy sources. In this challenge, competitors combined traditional optimization methods and machine learning to build algorithms for controlling a battery charging system as efficiently as possible. #energy

€23,000 in prizes
mar 2018
354 joined

alt-text

Each challenge in the Power Laws series explored a different aspect of energy efficiency and management. The winning algorithms were released under an open source license to spread understanding about how energy modeling works and what approaches are most effective.

Why

Flexibility in energy management is essential to avoid costly reinforcements of the power system and to maintain secure supply while increasing the penetration of renewable sources. Energy storage can increase smart building flexibility, while time of use tariffs can incite use of energy when it is the most available. This is a delicate balance, where algorithms can help battery charging systems to be as efficient as possible (for instance, buy more energy when its price is lowest, and buy less or sell energy when its price is highest).

The Solution

Using data from Schneider Electric, competitors built models to control a battery charging over a simulation period. At the end of the challenge, the simulation was scored across all submitted algorithms, and those that spent the least amount of money over that period rose to the top of the leaderboard.

The Results

The top algorithm drives 20% savings – that is 20% less money spent using the battery than without it. The winning models used traditional optimization tools like Linear Programs and constrained optimization, but also leveraged machine learning and measurements of uncertainty to build the most reliable optimization system.


RESULTS ANNOUNCEMENT + MEET THE WINNERS

WINNING MODELS ON GITHUB

DATASET ON SCHNEIDER ELECTRIC EXCHANGE