Development and Testing of a Smart Optimization System for EVs Charging in Residential Parking Areas

The widespread adoption of electric vehicles (EVs) is reshaping urban energy consumption, with residential charging emerging as both a necessity and a technical challenge—especially in Italian multi-unit housing, which is the primary focus of this work. This thesis presents the development and laboratory testing of an intelligent Energy Management System (EMS) designed to optimize EV charging in condominium parking areas. Building on the work of previous thesis students, the proposed solution aims to minimize operational costs by aligning charging schedules with real-time electricity prices, while ensuring compliance with contractual power limits, vehicle necessities and user-defined constraints. At its core, the system relies on a Mixed Integer Linear Programming (MILP) optimizer, previously implemented in Python with Pyomo, which calculates optimal power setpoints for each charging point based on vehicle parameters, time constraints, and the Italian day-ahead energy price (PUN). The optimizer is integrated into a functional control architecture capable of real-time communication with AC wallboxes from different manufacturers via Modbus TCP. Users can then communicate their charging preferences through a dedicated interface. In addition,the code includes a communication link with external metering equipment for research validation purposes. An experimental test was performed at Edison's laboratories, simulating a realistic residential charging environment with five stations, three of which were physically absorbing and delivering power. The system was subjected to a range of tests—including virtual simulations, accelerated and long-duration real-time campaigns—to assess its performance under various usage scenarios and input schedules. The results demonstrate that the orchestrator effectively balances power demand and reduces charging costs in typical daily scenarios. Additionally, it can identify unexpected disconnections and provide vehicle-specific state-of-charge (SOC) estimations using a dedicated parameter. These outcomes confirm the technical feasibility and economic relevance of deploying intelligent EMSs in residential EV charging environments, and suggest a promising pathway for broader adoption in condominium infrastructures.