Electric car charging: leave it to agents

Electric car charging: leave it to agents

Software agents play an important role in reaping the financial benefits from the electric car battery capacity and in charging huge numbers of new electric cars without causing power failures.

Electric cars may well become the buffer of tomorrow’s power grid, argue Micha Kahlen and his co-authors Prof. Ketter and Dr Jan van Dalen. As wind turbines, solar panels and other sustainable forms of energy production are on the rise, energy sources are bound to become more erratic. In this scenario, electric cars can be deployed to balance the grid.

“You can use the batteries of electric cars to store energy when there is an excess supply of wind or solar energy, and use that energy later, when there’s no wind or sun,” Kahlen explains. “Electric cars don’t produce a lot each, but the collective mass of electric cars can be combined into virtual power plants.”


Selling energy back into the grid could become a lucrative enterprise for fleet owners: Car-sharing company Car2Go, which rents out electric vehicles for a short amount of time is an example of trade-off between renting out an electric vehicle and turning it into a virtual power plant. The fleet owner can choose the most profitable option.

Virtual Power Plants

Virtual power plants would also have environmental and societal benefits. Based upon the researchers’ calculations, they would reduce CO2 emissions by 2.4 per cent and the average electricity price by 3.2 per cent.

At around six o’clock, when everyone comes home and turns on their electrical devices, you need additional electric capacity that’s idle the rest of the time. This electricity is generated by turning on carbon-intensive plants, which are both polluting and expensive. You can replace this energy with clean, stored energy, thereby reducing overall CO2 emissions and lowering the average electricity bill for all consumers.

Software agents play a crucial role in this scheme. They will trade on the wholesale energy market on behalf of fleet owners.

We need intelligent agents that act on information from energy markets – they know when it’s profitable to charge a car and when to turn it into a power plant – while also taking the preferences of customers – who want to be mobile – into account.

The agents have been tested extensively in the simulated trading environment Power TAC. An important fact for fleet owners: it is important to be a ‘first mover’. With each additional electric car on the market, the profit-per-car decreases due to price-based competition.

Reduce peaks

The agent learns and tailors itself to individual car owners. The algorithm learns our individual preferences and uses them to optimize charging times. It will allocate charging times differently within the time horizon of the day. That’s what makes the algorithm reduce peaks, and ultimately the demand on the grid becomes less volatile.”

The algorithm is run inside the car. Personalised charging algorithms such as AMEVS can tell your car to charge during the night to benefit from lower tariffs.

Research reports

Involved researchers

Reageer op dit artikel

Your email address will not be published. Required fields are marked *