Search Results

In recent years, the urgent need to fully exploit the fuel economy potential of the Electrified Vehicles (xEVs) through the optimal design of their Energy Management System (EMS) have led to an increasing interest in Machine Learning (ML) techniques. Among them, Reinforcement Learning (RL) seems to be one of the most promising approaches thanks to its peculiar structure, in which an agent is able to learn the optimal control strategy through the feedback received by a direct interaction with the environment. Therefore, in this study, a new Soft Actor-Critic agent (SAC), which exploits a stochastic policy, was implemented on a digital twin of a state-of-the-art diesel Plug-in Hybrid Electric Vehicle (PHEV) available on the European market. The SAC agent was trained to enhance the fuel economy of the PHEV while guaranteeing its battery charge sustainability.

Despite the legislation targets set by several governments of a full electrification of new light-duty vehicle fleets by 2035, the development of innovative, environmental-friendly Internal Combustion Engines (ICEs) is still crucial to be on track toward the complete decarbonization of on road-mobility of the future. In such a framework, the PHOENICE (PHev towards zerO EmissioNs & ultimate ICE efficiency) project aims at developing a C SUV-class plug-in hybrid (P0/P4) vehicle demonstrator capable to achieve a -10% fuel consumption reduction with respect to current EU6 vehicle while complying with upcoming EU7 pollutant emissions limits. Such ambitious targets will require the optimization of the whole engine system, exploiting the possible synergies among the combustion, the aftertreatment and the exhaust waste heat recovery systems.