Model Reduction Techniques Applied to a Physical Vehicle Model for HiL Testing

This document explores the development and application of model reduction techniques for complex physical vehicle models within the MORSE (Model-based Real-time Systems Engineering) project using Dymola and Modelica.

The research focuses on simplifying detailed multi-domain vehicle subsystems — including transmissions, drivelines, suspensions, tyres, and chassis systems — to achieve real-time capable Hardware-in-the-Loop (HiL) simulation while preserving the dynamic behaviour and accuracy required for drivability analysis and virtual calibration. By generating reduced-order equivalents from high-fidelity physical models through automated parameterisation, lumped loss mapping, and subsystem simplification strategies, the study demonstrates significant improvements in simulation speed, event reduction, and real-time performance with minimal loss of accuracy during tip-in and tip-out manoeuvres. The work highlights how scalable model reduction approaches can support efficient virtual vehicle development, control validation, and real-time simulation workflows for advanced automotive engineering applications.