Inverse Analysis of Road Contact Force and Contact Location Using Machine Learning with Measured Strain Data

Abstract

<div class="section abstract"><div class="htmlview paragraph">To adapt to Battery Electric Vehicle (BEV) integration, the significance of protective designs for battery packs against ground impact caused by road debris is very high, and there is also a keen interest in the feasibility assessment technique using Computer-Aided Engineering (CAE) tools for prototype-free evaluations. However, the challenge lies in obtaining real-world empirical data to verify the accuracy of the predictive CAE model. Collecting real-world data using actual battery pack can be time-consuming, costly, and accurately ascertaining the precise direction, magnitude, and location of the force applied from the road to the battery pack poses a challenging task. Therefore, in this study, we developed a methodology using machine learning, specifically Gaussian process regression (GPR), to perform inverse analysis of the direction, magnitude, and location of vehicle-road contact forces during rough road conditions. This was achieved by measuring the strain distribution of the plate-like device attached to the vehicle's underbody, as explanatory variables for the regression. To create a regression model, strain distribution patterns of the measurement device were gathered as a training dataset by altering the direction, magnitude, and application location of the forces in a laboratory environment. Finally, we conducted a test on a bumpy test road at different speeds (20km/h and 30km/h) to assess the vehicle-road interaction. We qualitatively confirmed the reliability of the analysis results regarding the time series data of the direction, magnitude, and location of vehicle-road contact forces obtained through regression by comparing them with contact marks on the device, vehicle behavior captured by a high-speed camera, and deformation of the device measured using laser displacement sensors, verifying their consistency. The findings were employed to identify parameters to be set in predictive CAE further expediting Model-Based Development (MBD).</div></div>