检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Kouw, Wouter M. Podusenko, Albert Koudahl, Magnus T. Schoukens, Maarten The Bayesian Intelligent Autonomous Systems Lab The Department of Electrical Engineering TU Eindhoven Eindhoven Netherlands The Control Systems Group The Department of Electrical Engineering TU Eindhoven Eindhoven Netherlands

We propose a variational Bayesian inference procedure for online nonlinear system identification. For each output observation, a set of parameter posterior distributions is updated, which is then used to form a posterior predictive distribution for future outputs. We focus on the class of polynomial NARMAX models, which we cast into probabilistic form and represent in terms of a Forney-style factor graph. Inference in this graph is efficiently performed by a variational message passing algorithm. We show empirically that our variational Bayesian estimator outperforms an online recursive least-squares estimator, most notably in small sample size settings and low noise regimes, and performs on par with an iterative least-squares estimator trained offline. © 2022, CC BY.

关键词： Message passing

来源：评论

学校读者我要写书评

暂无评论

Decision-Focused Learning for Complex System Identification: HVAC Management System Application

arXiv

引用

arXiv 2025年

作者： Favaro, Pietro Toubeau, Jean-François Vallée, François Dvorkin, Yury Power Systems and Markets Research Group University of Mons Mons Belgium Ralph O’Connor Sustainable Energy Institute Department of Electrical and Computer Engineering Department of Civil and System Engineering Johns Hopkins University BaltimoreMD United States

As opposed to conventional training methods tailored to minimize a given statistical metric or task-agnostic loss (e.g., mean squared error), Decision-Focused Learning (DFL) trains machine learning models for optimal performance in downstream decision-making tools. We argue that DFL can be leveraged to learn the parameters of system dynamics, expressed as constraint of the convex optimization control policy, while the system control signal is being optimized, thus creating an end-to-end learning framework. This is particularly relevant for systems in which behavior changes once the control policy is applied, hence rendering historical data less applicable. The proposed approach can perform system identification — i.e., determine appropriate parameters for the system analytical model — and control simultaneously to ensure that the model’s accuracy is focused on areas most relevant to control. Furthermore, because black-box systems are non-differentiable, we design a loss function that requires solely to measure the system response. We propose pre-training on historical data and constraint relaxation to stabilize the DFL and deal with potential infeasibilities in learning. We demonstrate the usefulness of the method on a building Heating, Ventilation, and Air Conditioning day-ahead management system for a realistic 15-zone building located in Denver, US. The results show that the conventional RC building model, with the parameters obtained from historical data using supervised learning, underestimates HVAC electrical power consumption. For our case study, the ex-post cost is on average six times higher than the expected one. Meanwhile, the same RC model with parameters obtained via DFL underestimates the ex-post cost only by 3%. © 2025, CC BY-NC-ND.

关键词： Convex optimization

来源：评论

学校读者我要写书评

暂无评论

Physics–Guided Neural Networks for Feedforward control: From Consistent Identification to Feedforward controller Design*

Physics–Guided Neural Networks for Feedforward Control: Fro...

引用

IEEE Conference on Decision and control

作者： Max Bolderman Mircea Lazar Hans Butler Department of Electrical Engineering Control Systems Group Eindhoven University of Technology The Netherlands ASML Veldhoven The Netherlands

ISBN: (数字)9781665467612

ISBN: (纸本)9781665467629

Model–based feedforward control improves tracking performance of motion systems if the model describing the inverse dynamics is of sufficient accuracy. Model sets, such as neural networks (NNs) and physics–guided neural networks (PGNNs) are typically used as flexible parametrizations that enable accurate identification of the inverse system dynamics. Currently, these (PG)NNs are used to identify the inverse dynamics directly. However, direct identification of the inverse dynamics is sensitive to noise that is present in the training data, and thereby results in biased parameter estimates which limit the achievable tracking performance. In order to push performance further, it is therefore crucial to account for noise when performing the identification. To address this problem, this paper proposes a forward system identification using (PG)NNs from noisy data. Afterwards, two methods are proposed for inverting PGNNs to design a feedforward controller. The developed methodology is validated on a real–life industrial linear motor, where it showed significant improvements in tracking performance with respect to the direct inverse identification.

关键词： Training System dynamics Tracking Dynamics Training data Feedforward neural networks System identification

来源：评论

学校读者我要写书评

暂无评论

On feedforward control using physics-guided neural networks: Training cost regularization and optimized initialization

arXiv

引用

arXiv 2022年

作者： Bolderman, Max Lazar, Mircea Butler, Hans Control Systems Group department of Electrical Engineering Eindhoven University of Technology Netherlands ASML Veldhoven Netherlands

Performance of model–based feedforward controllers is typically limited by the accuracy of the inverse system dynamics model. Physics–guided neural networks (PGNN), where a known physical model cooperates in parallel with a neural network, were recently proposed as a method to achieve high accuracy of the identified inverse dynamics. However, the flexible nature of neural networks can create overparameterization when employed in parallel with a physical model, which results in a parameter drift during training. This drift may result in parameters of the physical model not corresponding to their physical values, which increases vulnerability of the PGNN to operating conditions not present in the training data. To address this problem, this paper proposes a regularization method via identified physical parameters, in combination with an optimized training initialization that improves training convergence. The regularized PGNN framework is validated on a real–life industrial linear motor, where it delivers better tracking accuracy and extrapolation. Copyright © 2022, The Authors. All rights reserved.

关键词： Feedforward control

来源：评论

学校读者我要写书评

暂无评论

control of Nonlinear systems Under Multiple Time-Varying Output Constraints: A Single Funnel Approach

Control of Nonlinear Systems Under Multiple Time-Varying Out...

引用

IEEE Conference on Decision and control

作者： Farhad Mehdifar Lars Lindemann Charalampos P. Bechlioulis Dimos V. Dimarogonas Division of Decision and Control Systems KTH Royal Institute of Technology Stockholm Sweden Thomas Lord Department of Computer Science University of Southern California Los Angeles CA USA Department of Electrical and Computer Engineering Division of Systems and Control University of Patras Patra Greece

This paper proposes a novel control framework for handling (potentially coupled) multiple time-varying output constraints for uncertain nonlinear systems. First, it is shown that the satisfaction of multiple output constraints boils down to ensuring the positiveness of a scalar variable (the signed distance from the time-varying output-constrained set's boundary). Next, a single funnel constraint is designed properly, whose satisfaction ensures convergence to and invariance of the time-varying output-constrained set. Then a robust and low-complexity funnel-based feedback controller is designed em-ploying the prescribed performance control method. Finally, a simulation example clarifies and verifies the proposed approach.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Physics-Guided Neural Networks for Feedforward control: From Consistent Identification to Feedforward controller Design

arXiv

引用

arXiv 2022年

作者： Bolderman, Max Lazar, Mircea Butler, Hans Control Systems Group Department of Electrical Engineering Eindhoven University of Technology Netherlands Asml Veldhoven Netherlands

Model-based feedforward control improves tracking performance of motion systems, provided that the model describing the inverse dynamics is of sufficient accuracy. Model sets, such as neural networks (NNs) and physics-guided neural networks (PGNNs) are typically used as flexible parametrizations that enable accurate identification of the inverse system dynamics. Currently, these (PG)NNs are used to identify the inverse dynamics directly. However, direct identification of the inverse dynamics is sensitive to noise that is present in the training data, and thereby results in biased parameter estimates which limit the achievable tracking performance. In order to push performance further, it is therefore crucial to account for noise when performing the identification. To address this problem, this paper proposes the use of a forward system identification using (PG)NNs from noisy data. Afterwards, two methods are proposed for inverting PGNNs to design a feedforward controller for high-precision motion control. The developed methodology is validated on a real-life industrial linear motor, where it showed significant improvements in tracking performance with respect to the direct inverse identification. Copyright © 2022, The Authors. All rights reserved.

关键词： Feedforward control

来源：评论

学校读者我要写书评

暂无评论

Adaptive PI-Fuzzy Like control of a Stack Pneumatic Actuators Testbed for Multi-Configuration Small Scale Soft Robotics 6

Adaptive PI-Fuzzy Like Control of a Stack Pneumatic Actuator...

引用

6th International Conference on Manipulation, Automation, and Robotics at Small Scales, MARSS 2023

作者： Al Mhdawi, Ammar K. Wright, Nick Humaidi, Amjad J. Azar, Ahmad Taher Newcastle University Department of Electrical and Electronic Engineering Newcastle United Kingdom Edge Hill University United Kingdom Prince Sultan University Asscl Lab Saudi Arabia University of Technology Control and Systems Engineering Department Baghdad Iraq College of Computer & Information Sciences Prince Sultan University Riyadh Saudi Arabia Benha University Faculty of Computers and Artificial Intelligence Egypt

ISBN: (纸本)9798350330397

It has become evident that pneumatic muscle-based robots are the most compliant of all soft and semi-soft robotic testbeds under investigation, despite the fact that the artificial muscle-based robot is the most challenging due to its hyperelastic behavior resulting from the constitutive elastic material. The purpose of this paper is to examine the development of a physical pneumatic actuation testbed and the application of control strategies for precision small scale soft robotics control. Moreover, an adaptive PI-Fuzzy like control is proposed for reference signal tracking. It is imperative to note that in this application, the controller's performance is determined by its ability to track a reference signal as well as its ability to minimize the absolute tracking error. The Fuzzy Logic controller (FLC) controller, including the nature of the membership functions and the rule base, is described in detail. Simulations and hardware testing have revealed that the adaptive PI-Fuzzy like controller outperforms the traditional FLC controller in terms of experimental results. Additionally, the actuation speed is controlled under different loading conditions. It is considered that the proposed testbed is capable of controlling and managing multi-configurations of medium to small scale soft-based robots efficiently and effectively. © 2023 IEEE.

关键词： controllers

来源：评论

学校读者我要写书评

暂无评论

A Privacy-Preserving Protocol Level Approach to Prevent Machine Learning Modelling Attacks on PUFs in the Presence of Semi-Honest Verifiers

A Privacy-Preserving Protocol Level Approach to Prevent Mach...

引用

IEEE International Workshop on Hardware-Oriented Security and Trust (HOST)

作者： Owen Millwood Fei Hongming Prosanta Gope Oğuz Narlı Meltem Kurt Pehlivanoğlu Elif Bilge Kavun Biplab Sikdar Department of Computer Science The University of Sheffield Sheffield UK Department of Electrical and Computer Engineering The National University of Singapore Singapore Department of Computer Engineering Kocaeli University Kocaeli Turkey Secure Intelligent Systems Research Group University of Passau Passau Germany

With the ubiquitous and distributed nature of the Internet-of-Things (IoT), various qualities of traditional communication methods for end devices and their verifiers prove insufficient in solving the challenges this new paradigm faces. Many new hardware and software technologies are proposed in an attempt to provide IoT systems with desired security properties while meeting performance requirements. Physically Unclonable Functions (PUFs) are one such technology receiving particular interest from the wider research community by promising to provide low-cost and highly secure key data to enable lightweight authentication protocols for devices operating over publicly accessible networks. PUFs have been the target of Machine Learning Modelling Attacks (ML-MA), which aim to clone the intrinsic behaviour of the PUF to undermine their security. While many PUF-based schemes have been proposed to defend against adversaries who are guaranteed to be dishonest, an area which has not seen significant consideration is one where a normal communication participant cannot always be assumed to act honestly. To the best of our knowledge, this work is the first to consider the concept of ‘semi-honest verifier’ for PUFbased authentication, taking initial steps to shed light on this prominent issue in IoT by proposing a privacy-preserving mutual authentication protocol which considers security against MLMA in the presence of such verifiers. Furthermore, this work describes hardware-level considerations for PUF obfuscation by utilising a combination of strong PUF, configurable One-Way Function (OWF) and secure DRAM-PUF and is, therefore, one of the first to integrate PUF obfuscation comprehensively at the protocol level.

关键词： Performance evaluation Protocols Authentication Cloning Machine learning Physical unclonable function Hardware

来源：评论

学校读者我要写书评

暂无评论

Multiplier analysis of Lurye systems with power signals

Multiplier analysis of Lurye systems with power signals

引用

IEEE Conference on Decision and control

作者： William P. Heath Joaquin Carrasco School of Computer Science and Engineering Bangor University Dean St LL57 1UT UK Department of Electrical and Electronic Engineering Control Systems and Robotics Group Nancy Rothwell Building University of Manchester M13 9PL UK

ISBN: (数字)9798350316339

ISBN: (纸本)9798350316346

Multipliers can be used to guarantee both the Lyapunov stability and input-output stability of Lurye systems with time-invariant memoryless slope-restricted nonlinearities. If a dynamic multiplier is used there is no guarantee the closedloop system has finite incremental gain. It has been suggested in the literature that without this guarantee such a system may be critically sensitive to time-varying exogenous signals including noise. We show that multipliers guarantee the power gain of the system to be bounded and quantifiable. Furthermore power may be measured about an appropriate steady state bias term, provided the multiplier does not require the nonlinearity to be odd. Hence dynamic multipliers can be used to guarantee Lurye systems have low sensitivity to noise, provided other exogenous systems have constant steady state. We illustrate the analysis with an example where the exogenous signal is a power signal with non-zero mean.

关键词： Sensitivity Power measurement Noise Stability analysis Steady-state Time-varying systems Lyapunov methods

来源：评论

学校读者我要写书评

暂无评论

Infrastructure-dependent ramp-metering control for METANET-s

Infrastructure-dependent ramp-metering control for METANET-s

引用

IEEE Conference on Decision and control

作者： Ayda Kamalifar Carlo Cenedese Michele Cucuzzella Antonella Ferrara Department of Electrical Computer Science and Biomedical Engineering University of Pavia Pavia Italy Department of Information Technology and Electrical Engineering ETH Zürich Zürich Switzerland Center for Systems and Control ENTEG Faculty of Science and Engineering (FSE) University of Groningen (UG) Nijenborgh 4 Groningen the Netherlands

ISBN: (数字)9798350316339

ISBN: (纸本)9798350316346

In this paper, we propose a novel infrastructure-dependent ramp-metering control for the recently proposed METANET with service station (METANET-s) model, i.e., a second-order macroscopic traffic model that, compared to the classical METANET, incorporates the dynamics of service stations on highways. We study the effect of a ramp-metering control scheme on a highway stretch with a service station and show that it is capable of actively regulate internal traffic demand attempting to exit the service station via its on-ramp, on top of contributing to decrease the traffic congestion on the mainstream. In fact, the proposed control scheme effectively prevents the backlog of vehicles attempting to merge back onto the mainstream. This dynamic control mechanism is further endowed by a route guidance control strategy increasing the share of vehicles stopping at the service station during main-stream congestion periods, e.g. via incentives. The combined effect of our control schemes allows to take full advantage of the presence of service stations, reducing the overall traffic congestion. Simulation results demonstrate the effectiveness of the proposed control strategies.

关键词： Simulation Merging Numerical models Vehicle dynamics Traffic congestion

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：