Dissertation Title: Nonlinear Modeling, Sparse Estimation, and Input Design for Vehicle Dynamics

Date: 2025/12/04 – 2025/12/04

Dissertation Title: Nonlinear Modeling, Sparse Estimation, and Input Design for Vehicle Dynamics

Speaker: Le Wang

Time: 22:00, Thursday, Dec 4, 2025(Beijing Time)

Location: Room 202, Global Institute of Future Technology

Abstract

This dissertation investigates system identification challenges in vehicle dynamics under complex and nonlinear operating conditions. It focuses on estimating hard-to-measure quantities such as wheel center torque and vehicle mass, which are essential for safe and efficient autonomous driving. The research makes three main contributions. First, a soft sensor is developed for torque estimation using a MISO-FIR model combined with a classifier to handle nonlinear dead-zone effects. Second, a regularized-based sparse modeling and parameter merging method is proposed to reduce model redundancy and improve multi-condition estimation. Third, an application-oriented input design framework is established for vehicle mass estimation, providing optimal acceleration and velocity trajectories under different objectives and physical constraints. Real-world vehicle experiments validate the effectiveness and feasibility of all proposed methods. Overall, the dissertation presents an integrated framework for nonlinear modeling, sparse estimation, and input design that enhances the reliability of system identification for intelligent vehicle systems.

Biography

Le Wang is currently pursuing a joint Ph.D. degree with KTH Royal Institute of Technology and Shanghai Jiao Tong University, majoring in electrical engineering and control science and engineering, respectively. She received her Bachelor’s degree in Automation from Shandong University in 2019. Her research interests include system identification, input design, optimization algorithms, signal processing, and their related applications in vehicle systems.