Identification of Vehicle Dynamics Parameters Using Simulation-based\n Inference

2021/08/27 by Ali Boyali, Boyali, Ali, Simon G. Thompson +3 · 1 citation
Chemistry · Computer Science · Mathematics · #Mass Spectrometry Techniques and Applications #Machine Learning and Algorithms #Markov Chains and Monte Carlo Methods

paper · pdf · doi:10.48550/arxiv.2108.12114

Abstract

Identifying tire and vehicle parameters is an essential step in designing\ncontrol and planning algorithms for autonomous vehicles. This paper proposes a\nnew method: Simulation-Based Inference (SBI), a modern interpretation of\nApproximate Bayesian Computation methods (ABC) for parameter identification.\nThe simulation-based inference is an emerging method in the machine learning\nliterature and has proven to yield accurate results for many parameter sets in\ncomplex problems. We demonstrate in this paper that it can handle the\nidentification of highly nonlinear vehicle dynamics parameters and gives\naccurate estimates of the parameters for the governing equations.\n

Cited by

• Bayesian Optimization-based Tire Parameter and Uncertainty Estimation for Real-World Data

Related

• Bayesian Optimization-based Tire Parameter and Uncertainty Estimation for Real-World Data
• Adaptive Planning and Control with Time-Varying Tire Models for Autonomous Racing Using Extreme Learning Machine
• System Identification for Dynamic Modeling of Large Steering Angle Vehicles
• Bridging the Sim-to-Real Gap with Bayesian Inference
• Strengthening the Case for a Bayesian Approach to Car-following Model\n Calibration and Validation using Probabilistic Programming