engleski

Multi-agent system for managing energy storage using electric vehicles

The thesis addresses the challenges related to management of energy storage with electric vehicles (EVs). Charging of EV batteries is a time-consuming process with the current state of technology and therefore, the process of charging is often coupled with the process of parking. This thesis builds on that assumption and proposes a solution how to design a parking lot with EV charging functionality so that the parking lot operator’s profit is maximized while, simultaneously, the need of EV owners for parking and energy is fulfilled. Therefore, this thesis focuses on EV energy storage management through answering the questions when, where, at what price, and how long an EV should park in an EV- enabled parking spot. To answer such questions, a novel framework for studying EV-enabled parking lots (EVPLs) based on simulations aligned with real-world data was designed and implemented. Such tool allows quantifying the impact of different attributes, such as the number of chargers, on different aspects of the parking lot, such as profitability and payback period. The use of the proposed framework is illustrated for a public parking lot in Melbourne, Australia. Besides uncovering complex patterns of how the number of chargers and the charging tariff affect different aspects of the underlying park lot, the thesis also provides some suggestions given different EV adoption rate and the number of chargers within the parking lot. For example, an evidence suggested that in futuristic scenario where 30% of the cars in Melbourne are EVs, the best strategy for a cautious investor is to invest in charging infrastructure for 18% of the total number of City Square parking spots. Moreover, the developed framework suggests that such parking spots are best monetized when the premium hourly fee is set at 0.45 Australian dollars.

electric vehicles ; charging infrastructure ; multi-agent systems ; simulation ; energy storage

nije evidentirano