Periodic Reporting for period 1 - ChargFlex (Smart Synergy Mechanism between Electric Vehicle Charging and Flexibility Markets)
Periodo di rendicontazione: 2023-09-01 al 2025-08-31
The transition toward climate neutrality needs rapid decarbonisation of both the electricity and transportation sectors, which together account for the majority of greenhouse gas emissions in Europe. Electric vehicles (EVs) play a central role in this transition; however, their large-scale adoption is still slowed down by limited charging infrastructure availability, operational inefficiencies, and uncertainty in energy systems with high shares of renewable energy sources.
ChargFlex was designed to address these challenges by developing intelligent, uncertainty-aware solutions for EV charging infrastructures. The project aimed to improve the economic viability and operational reliability of EV charging systems while enhancing the quality of charging experience for EV drivers. By combining advanced energy management methods with artificial intelligence, ChargFlex sought to support better integration of renewable energy, reduce range anxiety, and contribute to more resilient and sustainable charging ecosystems aligned with European climate and energy objectives.
ChargFlex was designed to address these challenges by developing intelligent, uncertainty-aware solutions for EV charging infrastructures. The project aimed to improve the economic viability and operational reliability of EV charging systems while enhancing the quality of charging experience for EV drivers. By combining advanced energy management methods with artificial intelligence, ChargFlex sought to support better integration of renewable energy, reduce range anxiety, and contribute to more resilient and sustainable charging ecosystems aligned with European climate and energy objectives.
ChargFlex developed advanced methodologies for the planning, operation, and decision support of EV charging infrastructures operating under uncertainty. A central achievement of the project is the development of a high-fidelity stochastic energy management framework for EV charging stations, explicitly accounting for uncertain electricity prices, renewable generation, and charging demand. Unlike many existing approaches in relevant literature, the proposed framework preserves modeling accuracy while remaining computationally tractable and suitable for real-time operation.
In parallel, the project delivered an AI-driven charging station recommendation system that combines fuzzy logic, to cover the subjectivity of the charging experience of the drivers and deep learning algorithms to support EV drivers’ decision-making. This system accounts for user preferences, charging prices, waiting times, and sustainability considerations, thereby improving charging convenience and reducing range anxiety.
The project also implemented early-stage real-time-oriented validation using simulation platforms like OPAL-RT and software prototyping, demonstrating the practical feasibility of the developed methods. Together, these achievements provide a comprehensive set of tools linking charging infrastructure operation, market participation, and user-centric decision support.
In parallel, the project delivered an AI-driven charging station recommendation system that combines fuzzy logic, to cover the subjectivity of the charging experience of the drivers and deep learning algorithms to support EV drivers’ decision-making. This system accounts for user preferences, charging prices, waiting times, and sustainability considerations, thereby improving charging convenience and reducing range anxiety.
The project also implemented early-stage real-time-oriented validation using simulation platforms like OPAL-RT and software prototyping, demonstrating the practical feasibility of the developed methods. Together, these achievements provide a comprehensive set of tools linking charging infrastructure operation, market participation, and user-centric decision support.
ChargFlex produced results that go beyond the state of the art in several key aspects. The project addressed previously unresolved modeling challenges associated with comprehensive EV charging infrastructures that include grid constraints, renewable energy sources, realistic non-linear energy storage systems, and EV chargers, which lead to a high-dimension higly non-linear, non-convex and NP hard programming that there is no solver to be able to solve such a realistic program. By developing exact reformulation and convexification techniques for highly non-linear and non-convex optimization problems, the project established a level of modeling rigor not commonly found in existing literature.
Furthermore, ChargFlex demonstrated that economically viable operation of EV charging systems in uncertain environments requires both high modeling accuracy and real-time solvability. To this end, the project integrated artificial intelligence techniques to accelerate complex optimization processes, enabling practical deployment where small approximation errors could otherwise eliminate profitability.
Beyond technical novelty, the project laid a strong foundation for future large-scale integration of charging and flexibility-oriented energy systems. Continued validation, industrial collaboration, and market-oriented deployment are identified as key steps to further exploit and scale the project outcomes.
Furthermore, ChargFlex demonstrated that economically viable operation of EV charging systems in uncertain environments requires both high modeling accuracy and real-time solvability. To this end, the project integrated artificial intelligence techniques to accelerate complex optimization processes, enabling practical deployment where small approximation errors could otherwise eliminate profitability.
Beyond technical novelty, the project laid a strong foundation for future large-scale integration of charging and flexibility-oriented energy systems. Continued validation, industrial collaboration, and market-oriented deployment are identified as key steps to further exploit and scale the project outcomes.