Efficient Learning of Vehicle Controller Parameters via Multi-Fidelity Bayesian Optimization: From Simulation to Experiment

2025-06-10

AI Summary

To address the high cost of physical vehicle testing in parameter tuning for intelligent connected vehicle controllers, this paper proposes a multi-fidelity Bayesian optimization (MFBO) method. The approach innovatively integrates an autoregressive multi-fidelity Gaussian process (AR-MFGP) into the Bayesian optimization framework, enabling cross-fidelity knowledge transfer between high-fidelity simulations and low-fidelity real-world data—without requiring additional low-fidelity experimental data collection. By synergistically modeling simulation and physical experiments and incorporating active learning, the method efficiently identifies optimal controller parameters using only a minimal number of real-vehicle tests. Experimental results demonstrate that the proposed method achieves control performance comparable to full-scale real-vehicle tuning, while substantially reducing development cost and time. Moreover, it natively aligns with the industrial two-stage development workflow.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Intelligent Robots: Learning & Optimization for ROB
• Machine Learning: Auto ML and Hyperparameter Tuning

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

Parameter tuning for vehicle controllers remains a costly and time-intensive challenge in automotive development. Traditional approaches rely on extensive real-world testing, making the process inefficient. We propose a multi-fidelity Bayesian optimization approach that efficiently learns optimal controller parameters by leveraging both low-fidelity simulation data and a very limited number of real-world experiments. Our approach significantly reduces the need for manual tuning and expensive field testing while maintaining the standard two-stage development workflow used in industry. The core contribution is the integration of an auto-regressive multi-fidelity Gaussian process model into Bayesian optimization, enabling knowledge transfer between different fidelity levels without requiring additional low-fidelity evaluations during real-world testing. We validate our approach through both simulation studies and realworld experiments. The results demonstrate that our method achieves high-quality controller performance with only very few real-world experiments, highlighting its potential as a practical and scalable solution for intelligent vehicle control tuning in industrial applications.

Problem

• Reducing costly vehicle controller parameter tuning
• Minimizing reliance on extensive real-world testing
• Integrating multi-fidelity optimization for efficient learning

Innovation

• Multi-fidelity Bayesian optimization for parameter tuning
• Auto-regressive Gaussian process model integration
• Reduced real-world experiments via simulation data