Adaptive Learning Agents for Electric Vehicle Customer Decision Support

Electric Vehicles (EVs) have already been one of main the discussion topics between energy policy makers on the road to sustainability. However, large scale EV integration will put the energy grid under critical strain since the peaks in the power demand will increase radically. We tackle this challenge by proposing a decentralized charging algorithm implemented through an IS artifact which focuses on the individual EV owners preferences. We show that it alleviates the energy grid accounting for the individual customers' valuation of EV charging. Specifically, we see significant peak demand and energy price reduction, as result of our artifact's adoption. We evaluate the proposed algorithm in various customer population scenarios, creating energy policy recommendations for each scenario.

Poster Number 22