We use cookies essential for this site to function well. Please click “Accept” to help us improve its usefulness with additional cookies. Learn about our use of cookies, and collaboration with select social media and trusted analytics partners here

EnergyTech
Maximizing energy output using renewables equipment
ShipTech
Optimizing vessels' engines for greener maritime
HealthTech
Enabling remote diagnostics using data from wearables
Life Sciences
Utilizing AI across the medical treatment value chain

Smart grid control

Speak with an AI/ML expert in your industry
Schedule a call
View our cases to discover what we have reaserched
Smart grid control - 1

About customer

Our customer is an energy services provider and electricity producer focusing on sustainable and climate-friendly energy sources and operations.

Challenge

The share of renewable energy resources like solar photovoltaic (PV), geothermal, tidal waves, wind power, and biomass has been growing rapidly in the energy market.

But currently most popular types of renewable resources like solar and wind are highly variable and the resulting fluctuations in the generation capacity cause instability in power grids. It happens because in the energy generation process the stability of the power grid relies on the equilibrium between demand and supply.

Thus, it is necessary to generate power when resources are available and store it for later use for achieving self-consumption based only on green energy.

Solution

Wind and solar PV energy are expensive to store; thus, careful energy generation management is needed.

In this case Machine learning help to resolve two main challenges:

  1. Forecast of renewable energy production (PV, Wind)
  2. Prediction of energy demand (or consumption).

Forecasting power output from a renewable energy power plant is crucial as this depends on many non-human-controllable factors such as environmental parameters. Depending on the energy source it uses, the power plant exhibits certain characteristics that enable machine learning techniques for prediction purposes.

Result

Before

  • The heuristic pipeline for estimation of the equilibrium between demand and supply wasn’t accurate enough
  • More energy consumed from the traditional grid due to the lack of an energy management system and forecasting for renewable sources. As a result, increases CO2 emissions.

After

  • Machine learning models to predict PV output and the consumption for 3 hours ahead were developed with 10% and 5% forecasting errors respectively
  • Eventually, combining these two models we developed a control strategy for battery charging and reduced grid consumption by 40%