VeMo: A Lightweight Data-Driven Approach to Model Vehicle Dynamics

2025-10-08

AI Summary

To address the challenge of constructing high-fidelity vehicle dynamics models for high-performance vehicles under structural information deficiency, this paper proposes a lightweight, purely data-driven encoder–decoder framework that predicts future vehicle states solely from onboard sensor measurements and driver command sequences. The method employs a gated recurrent unit (GRU)-based end-to-end sequence-to-sequence architecture, eliminating the need for physics-based modeling priors while ensuring computational efficiency and implicit physical consistency. Under extreme dynamic maneuvers, the model achieves a maximum mean relative error of less than 2.6% for critical states—including longitudinal velocity, acceleration, and yaw rate—and demonstrates robustness to input noise and high prediction accuracy within the target frequency band. By obviating reliance on detailed vehicle design parameters, the approach significantly reduces deployment barriers for autonomous driving systems and enables rapid integration onto existing vehicle platforms.

Technology Category

• Intelligent Robots: State Estimation
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Developing a dynamic model for a high-performance vehicle is a complex problem that requires extensive structural information about the system under analysis. This information is often unavailable to those who did not design the vehicle and represents a typical issue in autonomous driving applications, which are frequently developed on top of existing vehicles; therefore, vehicle models are developed under conditions of information scarcity. This paper proposes a lightweight encoder-decoder model based on Gate Recurrent Unit layers to correlate the vehicle's future state with its past states, measured onboard, and control actions the driver performs. The results demonstrate that the model achieves a maximum mean relative error below 2.6% in extreme dynamic conditions. It also shows good robustness when subject to noisy input data across the interested frequency components. Furthermore, being entirely data-driven and free from physical constraints, the model exhibits physical consistency in the output signals, such as longitudinal and lateral accelerations, yaw rate, and the vehicle's longitudinal velocity.

Problem

• Modeling vehicle dynamics without full structural information
• Developing data-driven approach using onboard sensor measurements
• Achieving accurate predictions under extreme driving conditions

Innovation

• Lightweight encoder-decoder model using GRU layers
• Correlates future vehicle state with past data
• Data-driven approach without physical constraints