arXiv subjects

Herman Van der Auweraer

Publications and source records attributed to Herman Van der Auweraer.

7 recordsLinked to original sources

arXiv2024

Learning Based NMPC Adaptation for Autonomous Driving using Parallelized Digital Twin

In this work, we focus on the challenge of transferring an autonomous driving controller from simulation to the real world (i.e. Sim2Real). We propose a data-efficient method for online and on-the-fly adaptation of parametrizable control architectures such that the target closed-loop performance is optimized while accounting for uncertainties as model mismatches, changes in the environment, and task variations. The novelty of the approach resides in leveraging black-box optimization enabled by Executable Digital Twins (xDTs) for data-driven parameter calibration through derivative-free methods to directly adapt the controller in real-time. The xDTs are augmented with Domain Randomization for robustness and allow for safe parameter exploration. The proposed method requires a minimal amount of interaction with the real-world as it pushes the exploration towards the xDTs. We validate our approach through real-world experiments, demonstrating its effectiveness in transferring and fine-tuning a NMPC with 9 parameters, in under 10 minutes. This eliminates the need for hours-long manual tuning and lengthy machine learning training and data collection phases. Our results show that the online adapted NMPC directly compensates for the Sim2Real gap and avoids overtuning in simulation. Importantly, a 75% improvement in tracking performance is achieved and the Sim2Real gap over the target performance is reduced from a factor of 876 to 1.033.

cs.RO↗

arXiv2023

Safety Envelope for Orthogonal Collocation Methods in Embedded Optimal Control

Orthogonal collocation methods are direct approaches for solving optimal control problems (OCP). A high solution accuracy is achieved with few optimization variables, making it more favorable for embedded and real-time NMPC applications. However, collocation approaches lack a guarantee about the safety of the resulting trajectory as inequality constraints are only set on a finite number of collocation points. In this paper we propose a method to efficiently create a convex safety envelope containing the trajectory such that the solution fully satisfies the OCP constraints. We make use of Bernstein approximations of a polynomial's extrema and span the solution over an orthogonal basis using Legendre polynomials. The tightness of the safety envelope estimation, high accuracy in solving the underlying differential equations, fast rate of convergence and little conservatism are properties of the presented approach making it a suitable method for safe real-time NMPC deployment. We show that our method has comparable computational performance to pseudospectral approaches and can accurately approximate the original OCP up to 9 times more quickly than standard multiple-shooting method in autonomous driving applications, without adding complexity to the formulation.

math.OC↗

arXiv2023

The Executable Digital Twin: merging the digital and the physics worlds

While the digital twin has become an intrinsic part of the product creation process, its true power lies in the connectivity of the digital representation with its physical counterpart. Data acquired on the physical asset can validate, update and enrich the digital twin. The knowledge contained in the digital representation brings value to the physical asset itself. When a dedicated encapsulation is extracted from the digital twin to model a specific set of behaviors in a specific context, delivering a stand-alone executable representation, such instantiated and self-contained model is referred to as an Executable Digital Twin. In this contribution, key building blocks such as model order reduction, real-time models, state estimation and co-simulation are reviewed, and a number of characteristic use cases are presented. These include virtual sensing, hybrid testing and hardware-in-the loop, model-based control and model-based diagnostics.

cs.OH↗

arXiv2022

Real-time Nonlinear MPC Strategy with Full Vehicle Validation for Autonomous Driving

In this paper, we present the development and deployment of an embedded optimal control strategy for autonomous driving applications on a Ford Focus road vehicle. Non-linear model predictive control (NMPC) is designed and deployed on a system with hard real-time constraints. We show the properties of sequential quadratic programming (SQP) optimization solvers that are suitable for driving tasks. Importantly, the designed algorithms are validated based on a standard automotive XiL development cycle: model-in-the-loop (MiL) with high fidelity vehicle dynamics, hardware-in-the-loop (HiL) with vehicle actuation and embedded platform, and full vehicle-hardware-in-the-loop (VeHiL). The autonomous driving environment contains both virtual simulation and physical proving ground tracks. NMPC algorithms and optimal control problem formulation are fine-tuned using a deployable C code via code generation compatible with the target embedded toolchains. Finally, the developed systems are applied to autonomous collision avoidance, trajectory tracking, and lane change at high speed on city/highway and low speed at a parking environment.

eess.SY↗

arXiv2022

Virtual Reality Assisted Human Perception in ADAS Development: a Munich 3D Model Study

As the development of autonomous driving (AD) and advanced driver assistance systems (ADAS) progresses, the relevance of the comfort of users is gaining increasing interest. It becomes significant to test and validate perceived comfort performance from the early phase of system development before driving on roads. Most of the present ADAS test procedures are not efficient in performing such comfort evaluation. One of the main challenges is to integrate high-quality, realistic and predictable virtual traffic scenarios into an ADAS testing framework that has physics-based sensors capable of sensing the virtual environment. In this paper, we present our development of a virtual reality based ADAS testing framework that enhances human perception evaluation. The main contribution relies on three aspects. First, we introduce our development of a large and high-quality (realism, structure, texture) 3D traffic model of the Munich city in Germany. Second, we optimize the 3D model for virtual reality purpose, and real-time capable for human-in-the-loop ADAS testing. Finally, the model is then integrated into an ADAS framework for testing and validating ADAS functionalities and perceived comfort performance. The developed framework components are presented with illustrative examples.

cs.HC↗

arXiv2022

Sim2real for Autonomous Vehicle Control using Executable Digital Twin

In this work, we propose a sim2real method to transfer and adapt a nonlinear model predictive controller (NMPC) from simulation to the real target system based on executable digital twin (xDT). The xDT model is a high fidelity vehicle dynamics simulator, executable online in the control parameter randomization and learning process. The parameters are adapted to gradually improve control performance and deal with changing real-world environment. In particular, the performance metric is not required to be differentiable nor analytical with respect to the control parameters and system dynamics are not necessary linearized. Eventually, the proposed sim2real framework leverages altogether online high fidelity simulator, data-driven estimations, and simulation based optimization to transfer and adapt efficiently a controller developed in simulation environment to the real platform. Our experiment demonstrates that a high control performance is achieved without tedious time and labor consuming tuning.

eess.SY↗

arXiv2021

Safe Imitation Learning on Real-Life Highway Data for Human-like Autonomous Driving

This paper presents a safe imitation learning approach for autonomous vehicle driving, with attention on real-life human driving data and experimental validation. In order to increase occupant's acceptance and gain drivers' trust, the autonomous driving function needs to provide a both safe and comfortable behavior such as risk-free and naturalistic driving. Our goal is to obtain such behavior via imitation learning of a planning policy from human driving data. In particular, we propose to incorporate barrier functions and smooth spline-based motion parametrization in the training loss function. The advantage is twofold: improving safety of the learning algorithm, while reducing the amount of needed training data. Moreover, the behavior is learned from highway driving data, which is collected consistently by a human driver and then processed towards a specific driving scenario. For development validation, a digital twin of the real test vehicle, sensors, and traffic scenarios are reconstructed toward high-fidelity and physics-based modeling technologies. These models are imported to simulation tools and co-simulated with the proposed algorithm for validation and further testing. Finally, we present experimental results and analyses, and compare with the conventional imitation learning technique (behavioral cloning) to justify the proposed development.

eess.SY↗