Automatic dimensionality reduction of Twin-in-the-Loop Observers

2024-01-18

arXiv.org

AI Summary

To address the challenges of difficult filter parameter tuning and poor real-time performance in Twin-in-the-Loop (TiL) observers—stemming from tightly coupled, high-dimensional digital twin models—this paper proposes a supervised and unsupervised learning–co-driven automatic dimensionality reduction framework. For the first time, it integrates Principal Component Analysis (PCA) with neural network regression, augmented by linear time-invariant output-error correction and Bayesian optimization, enabling interpretable and scalable parameter-space compression in black-box digital twin environments. The method drastically reduces the effective tuning dimensionality while preserving high-accuracy estimation of vehicle speed and yaw rate on real-world driving data. It achieves a 3.2× improvement in convergence speed, establishing a novel paradigm for real-time deployment of closed-loop observers based on complex digital twin models.

Technology Category

• Intelligent Robots: State Estimation
• Machine Learning: Dimensionality Reduction/Feature Selection
• Computer Vision: Learning & Optimization for CV

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
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

State-of-the-art vehicle dynamics estimation techniques usually share one common drawback: each variable to estimate is computed with an independent, simplified filtering module. These modules run in parallel and need to be calibrated separately. To solve this issue, a unified Twin-in-the-Loop (TiL) Observer architecture has recently been proposed: the classical simplified control-oriented vehicle model in the estimators is replaced by a full-fledged vehicle simulator, or digital twin (DT). The states of the DT are corrected in real time with a linear time invariant output error law. Since the simulator is a black-box, no explicit analytical formulation is available, hence classical filter tuning techniques cannot be used. Due to this reason, Bayesian Optimization will be used to solve a data-driven optimization problem to tune the filter. Due to the complexity of the DT, the optimization problem is high-dimensional. This paper aims to find a procedure to tune the high-complexity observer by lowering its dimensionality. In particular, in this work we will analyze both a supervised and an unsupervised learning approach. The strategies have been validated for speed and yaw-rate estimation on real-world data.

Problem

• Reducing complexity in Twin-in-the-Loop Observers for vehicle dynamics
• Addressing high-dimensional optimization in black-box simulator tuning
• Validating supervised and unsupervised learning for dynamics estimation

Innovation

• Unified Twin-in-the-Loop Observer architecture
• Real-time correction with linear error law
• Complexity reduction via supervised and unsupervised learning