Digital Twin for Smart Charging Systems

Digital Twin for Smart Charging Systems: The number of electric vehicles (EVs) has increased every year in the recent past, and we expect it to continue growing in the future. EVs lower emissions and reliance on fossil fuels compared to conventional vehicles, but they also increase the strain on the grid of electricity. The fact that many cars frequently charge at the same hours, which can waste opportunities to use, raises costs, and overloads the electricity grid with peaks of energy demand. Nowadays, the most common strategy used to manage the charging of EVs is the so-called First-In First-Served rule (FIFS), which only prioritizes EVs charging based on their arrival time. Smarter algorithms can be instead implemented to make the charging process more efficient, cheaper and faster, depending on our goals. At the same time, we may require techniques that are not just inexpensive but also more environmentally friendly moving towards a greener future. The main contribution of this thesis is the development of a digital twin platform in Python. Beyond serving as a simulator, our control panel acts as a practical tool for safely testing charging strategies before they are applied in real systems. Thanks to its interactive and easy-to-use design, the platform supports researchers, operators, and policymakers in exploring different approaches and in shaping charging systems that are efficient, sustainable, and responsive to user needs. Technically speaking, we used several libraries such as CVXPY, matplotlib, numpy, pandas, and customtkinter for making our panel complete and professional in order to test any smart charging models desired using different values and scenarios. Three distinct smart charging models are also presented in this thesis and tested using our simulation platform. The first is the Power Allocation model, which aims to reduce the electricity costs of EVs charging by shifting the power allocation to low-peak periods, without failing the user demand satisfaction. The second is the CO2 Emission model, which aims to minimize the carbon emission due to the electricity usage, together with a usual cost minimization. The third is the Cohort model, which enables the management of big fleets of EVs by aggregating them into different classes according to their similarities and energy demand. These models collectively address the three primary objectives of smart charging: cost reduction, emission reduction, and efficiency.