This article designs a data-driven unsupervised defense scheme for nonlinear systems by proposing a machine learning approach called gate recurrent unit-based modified denoising and stable image representation-aided a...
详细信息
This article designs a data-driven unsupervised defense scheme for nonlinear systems by proposing a machine learning approach called gate recurrent unit-based modified denoising and stable image representation-aided autoencoders. The proposed scheme decomposes original data into two subspaces through orthogonal projection. For secure transmission, information related to the system's dynamics, which is in the image space of the controlled system, is hidden through filtering, whereas only the dynamic-independent information is plaintext for transmission, which supplements the cryptographic encryption methods from a control perspective. Moreover, attack detection for nonstealthy and stealthy attacks is achieved simultaneously under the same framework. A case study is conducted for validation on the a hardware-in-the-loop platform with a mecanum-wheeled vehicle. The comparative experiments with well-known unsupervised data-driven methods show the high detection accuracy of the proposed defense scheme for nonstealthy and stealthy attacks and the excellent encryption capability.
AC microgrids typically use droop-operated gridforming inverters for voltage and frequency regulation as this allows stabilisation of the microgrid without communication. However, in cases with very low X/R line ratio...
详细信息
ISBN:
(纸本)9798350372793;9798350372786
AC microgrids typically use droop-operated gridforming inverters for voltage and frequency regulation as this allows stabilisation of the microgrid without communication. However, in cases with very low X/R line ratios, there is greater coupling between real and reactive power and stability issues can occur. This paper proposes the use of an adaptive data-driven compensator which operates in conjunction with droop-control. The proposed data-drivencontrol uses an LQR formulation for improved performance, and is updated in real-time according to developed criteria in order to adapt to network changes. We demonstrate improved stability regions via a large-scale MATLAB/Simulink study of a two-inverter test system where the line and load parameters are varied randomly.
The rapid evolution of Intelligent Transportation systems (ITS) in the Big data era, propelled by the Internet of Things (IoT), has led to advanced data-driven vehicle traffic forecasting. Graph Neural Networks (GNNs)...
详细信息
ISBN:
(纸本)9781728190549
The rapid evolution of Intelligent Transportation systems (ITS) in the Big data era, propelled by the Internet of Things (IoT), has led to advanced data-driven vehicle traffic forecasting. Graph Neural Networks (GNNs), particularly the Attention-Based Spatial-Temporal Graph Neural Networks (ASTGNN), are promising in traffic forecasting but face limitations in integrating Big data with privacy-preserving Federated learning (FL) due to unique data topology processing. This paper introduces the Availability Aware Federated Attention-based Spatial-Temporal Graph Neural Network (FastFlow), an innovative framework that enhances ASTGNN by integrating Federated learning across entities and employing Big data methodologies. FastFlow's distinctiveness lies in its availability-aware approach, aggregating adjacency matrices for global topology and utilizing a novel communication protocol that prioritizes data availability and correlation among organizations. Our evaluation of Caltrans Performance Measurement System (PEMS) data in a simulated setup demonstrates FastFlow's ability to balance predictive accuracy and data security in a multi-organizational context.
The growing interest in robust designs and data-driven technologies for safe control problems underscores the critical need to understand uncertainty for ensuring reliable safety guarantees. This review offers a conci...
详细信息
The growing interest in robust designs and data-driven technologies for safe control problems underscores the critical need to understand uncertainty for ensuring reliable safety guarantees. This review offers a concise survey of recent advancements in the control barrier function (CBF) method, widely recognized as a principled and effective approach to safe control, particularly in the context of uncertainty. From a unified perspective, we classify uncertainty into three types based on their learnability and transferability. Then we explore the techniques associated with each type of uncertainty found in the existing literature. Additionally, we highlight a knowledge-based safe control framework that utilizes meta-learning techniques to address dynamic uncertainty, shedding light on the potential for future investigations into practical learning algorithms and control problems. Furthermore, we employ topic modeling technologies to identify and generalize topics from the literature, thus revealing research trends and ongoing real-world applications withing the scope of safe control.
This article is concerned with a fault-tolerant control (FTC) scheme for feedback controlsystems with multiplicative faults by optimizing system performance with the aid of a reinforcement learning (RL) approach. To ...
详细信息
This article is concerned with a fault-tolerant control (FTC) scheme for feedback controlsystems with multiplicative faults by optimizing system performance with the aid of a reinforcement learning (RL) approach. To be specific, initially, based on the Youla-Kucera (YK) and dual YK parameterizations, a new performance-driven FTC method is proposed and its capability in dealing with multiplicative faults is proven. Then, data-driven implementation of this method using RL is elaborated. This implementation shows that RL can be applied efficiently by utilizing both plant model and data to recover the fault-induced system performance degradation. Finally, a benchmark study on an inverted pendulum system demonstrates the application of the proposed performance-driven FTC method.
Derived from the Hindmarsh-Rose (HR) neuron model, the extended Hindmarsh-Rose (e-HR) neuron model with time delay is constructed considering the effect of intracellular ion transfer and the inherent slow current dela...
详细信息
The industrial sector is currently undergoing a transformative era of intelligent automation driven by Artificial Intelligence (AI) capabilities. This synergy greatly enhances efficiency and seamlessly enables data-dr...
详细信息
The industrial sector is currently undergoing a transformative era of intelligent automation driven by Artificial Intelligence (AI) capabilities. This synergy greatly enhances efficiency and seamlessly enables data-driven decision-making processes. These advantages enable more efficient resource allocation and enhance production planning precision. This paper aims to provide state-of-the-art and ongoing developments in the AI landscape within the manufacturing industry. In addition, the review explores the key areas where AI is being applied in manufacturing, such as predictive maintenance, quality control, process optimization, supply chain management, robotics and automation, and intelligent decision support systems. The review also encompasses an exploration of the challenges encountered by the manufacturing sector, alongside an investigation into the potential of AI to mitigate these challenges. Furthermore, this work thoroughly reviews recent AI advancements, including explainable AI, human-robot collaboration, edge computing, and the Internet of Things (IoT) integration. The review concludes by providing recommendations, highlighting best practices, and providing insights into potential collaborative opportunities.
Environments for autonomous driving can vary from place to place, leading to challenges in designing a learning model for a new scene. Transfer learning can leverage knowledge from a learned domain to a new domain wit...
详细信息
ISBN:
(纸本)9798350384581;9798350384574
Environments for autonomous driving can vary from place to place, leading to challenges in designing a learning model for a new scene. Transfer learning can leverage knowledge from a learned domain to a new domain with limited data. In this work, we focus on end-to-end autonomous driving as the target task, consisting of both perception and control. We first utilize information bottleneck analysis to build a causal graph that defines our framework and the loss function;then we propose a novel domain-agnostic learning method for autonomous steering based on our analysis of training data, network architecture, and training paradigm. Experiments show that our method outperforms other SOTA methods.
In industrial positioning systems where rapid response and high-precision are crucial, minor model inaccuracies due to unknown dynamics and identification errors in controller design significantly impede achieving des...
详细信息
ISBN:
(纸本)9798350355376;9798350355369
In industrial positioning systems where rapid response and high-precision are crucial, minor model inaccuracies due to unknown dynamics and identification errors in controller design significantly impede achieving desired positioning accuracy. This paper introduces and evaluates a novel, direct data-drivencontrol-based additive feedforward (FF) compensation method, aimed at enhancing precision in positioning while streamlining the design process. The purpose of this additive FF compensation is to attenuate undesirable error responses resulting from modeling errors in the existing model-based FF design. The proposed method enhances control performance by utilizing data-driven prediction of positioning response and optimizing the predicted response. Moreover, this work presents a newly developed design theory for the additive FF controller and highlights its design efficiency. The effectiveness of the proposed approach is substantiated through comprehensive experiments with a galvano scanner in printed circuit board laser drilling applications, demonstrating significant improvements in positioning accuracy and response time.
Bio-inspired swarm robotics is an emerging field at the intersection of biology, robotics, and artificial intelligence, offering novel capabilities by integrating living organisms with robotic systems. This paper pres...
详细信息
暂无评论