Scholarlink

<p>This study presents a data-driven approach to derive physically interpretable equations for tire vertical load and effective rolling radius (ERR) using data measured by an intelligent tire. Conventional estimation methods suffer from the model uncertainty of vehicle dynamics models or the lack of physical interpretability in artificial intelligence-based black-box models. To overcome these issues, this study introduces a hierarchical modeling strategy utilizing the SINDy (Sparse Identification of Nonlinear Dynamics). First, the difficult-to-measure vertical load is modeled as a hidden variable, which is then integrated into ERR. Experimental evaluation on a Flat-trac test rig demonstrates that the proposed method achieves high accuracy (R² = 0.983, RMSE = 0.10&#xa0;mm). The main contribution of this work is in overcoming the limitations of black-box models by deriving physically interpretable equations directly from data, thereby enhancing the reliability and explainability of tire dynamics estimation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Physically Interpretable Modeling of Tire Effective Rolling Radius and Vertical Load via Hierarchical SINDy

  • 期刊论文

  • 发表日期：

    2026年3月31日

查看全文
  • Dongyoel Won,
  • Kyoungtae Ji,
  • Youngsam Yoon,
  • Sehwan Ku,
  • Jiyeon Hong,
  • Yonghyun Ryu,
  • Hyungjoo Kim,
  • Jaekil Lee,
  • Dae-Un Sung,
  • Hojong Lee,
  • Kyoungseok Han

摘要

This study presents a data-driven approach to derive physically interpretable equations for tire vertical load and effective rolling radius (ERR) using data measured by an intelligent tire. Conventional estimation methods suffer from the model uncertainty of vehicle dynamics models or the lack of physical interpretability in artificial intelligence-based black-box models. To overcome these issues, this study introduces a hierarchical modeling strategy utilizing the SINDy (Sparse Identification of Nonlinear Dynamics). First, the difficult-to-measure vertical load is modeled as a hidden variable, which is then integrated into ERR. Experimental evaluation on a Flat-trac test rig demonstrates that the proposed method achieves high accuracy (R² = 0.983, RMSE = 0.10 mm). The main contribution of this work is in overcoming the limitations of black-box models by deriving physically interpretable equations directly from data, thereby enhancing the reliability and explainability of tire dynamics estimation.