A Data-Driven methodology applied to X-In-the-Loop environments for Electric Vehicle development

The use of data-driven identification can provide high confidence models for mechatronic systems with real-time capabilities. These models are highly valuable for X-in-the-Loop technologies, giving a powerful way of monitoring the system evolution and automatically updating their structure and parameters in real-time. In this work, extended Dynamic Mode Decomposition (eDMD) is used for the development of these reduced order models based on simulation and experimental data. This methodology is applied to the modeling of an active suspension, which will be used in a shared X-in-the-Loop environment for electric vehicles involving several subsystems tested simultaneously in different locations.