Search | Engineering

订阅投稿

首页学术期刊学术焦点学术视频工程成就工程前沿关于我们 English

资源类型

期刊论文 478

会议视频 14

年份

2023 48

2022 54

2021 54

2020 53

2019 32

2018 28

2017 32

2016 15

2015 21

2014 12

2013 13

2012 12

2011 5

2010 4

2009 7

2008 13

2007 14

2006 9

2005 14

2004 11

展开︾

关键词

神经网络 23

人工神经网络 5

优化 5

人工智能 4

智能制造 4

BP神经网络 3

仿真 3

微波遥感 2

控制 2

机器学习 2

模糊神经网络 2

深度神经网络 2

算法 2

系统 2

网络 2

网络安全 2

能源 2

计算机 2

预测 2

展开︾

检索范围：

排序：展示方式：

基于最小化重构误差的生成对抗网络异常检测 Article

Huan-gang WANG, Xin LI, Tao ZHANG

《信息与电子工程前沿（英文）》 2018年第19卷第1期页码 116-125 doi: 10.1631/FITEE.1700786

摘要：生成对抗网络是机器学习领域近年来最令人瞩目的进展，它通过在二人零和博弈中达到纳什均衡来训练模型。生成对抗网络由一个生成器和一个判别器构成，二者通过对抗学习机制进行训练。本文引入并调查了生成对抗网络在异常检测中的应用。在训练阶段，生成对抗网络从正常数据中学习；然后，基于过去的未知数据，生成器和判别器可以通过学习到的决策边界，区分异常和正常模式。提出的基于生成对抗网络的异常检测方法在MNIST数字数据集和田纳西-伊斯曼标准数据集上的性能表现极具竞争力。

关键词：生成对抗网络；异常检测；田纳西-伊斯曼过程

HTML PDF 收藏

基于机器学习的广彩瓷图案生成系统 Special Feature on Intelligent Design

Steven Szu-Chi CHEN, Hui CUI, Ming-han DU, Tie-ming FU, Xiao-hong SUN, Yi JI, Henry DUH

《信息与电子工程前沿（英文）》 2019年第20卷第12期页码 1632-1643 doi: 10.1631/FITEE.1900399

摘要：由于广彩瓷设计元素和图案种类繁多、复杂多样，准确识别现代及传统瓷器风格是广彩瓷传承工作中的巨大挑战。提出一种基于广彩瓷风格识别和图像合成模块的图案生成系统。在识别模块中，通过主成分分析和所提判别冗余量化策略对特征重要性进行分析和排序，然后分别训练两组神经网络，将最优设计特征与转换后的主成分特征关联，最后利用整体神经网络逻辑回归方法预测未知广彩瓷。基于条件生成对抗网络（cGAN）开发合成模块，要求用户提供自己设计的创意掩码或抽象瓷元素图像，以生成新的广彩瓷风格合成图像。在系统开发过程中，使用603幅广彩瓷图像测试分类模型。测试结果表明，所提模型在精确度、召回率、接受者操作特性曲线（ROC）的曲线下面积（AUC）和混淆矩阵等方面均优于其他方法。对用户设计的各种元素合成图像的案例研究表明，该系统有助于提高学习者对广彩瓷的欣赏和艺术创作能力。

关键词：广彩瓷；分类；生成对抗网络；艺术创作

HTML PDF 收藏

Topology-independent end-to-end learning model for improving the voltage profile in microgrids-integrated power distribution networks

《能源前沿（英文）》 2023年第17卷第2期页码 211-227 doi: 10.1007/s11708-022-0847-3

摘要： With multiple microgrids (MGs) integrated into power distribution networks in a distributed manner, the penetration of renewable energy like photovoltaic (PV) power generation surges. However, the operation of power distribution networks is challenged by the issues of multiple power flow directions and voltage security. Accordingly, an efficient voltage control strategy is needed to ensure voltage security against ever-changing operating conditions, especially when the network topology information is absent or inaccurate. In this paper, we propose a novel data-driven voltage profile improvement model, denoted as system-wide composite adaptive network (SCAN), which depends on operational data instead of network topology details in the context of power distribution networks integrated with multiple MGs. Unlike existing studies that realize topology identification and decision-making optimization in sequence, the proposed end-to-end model determines the optimal voltage control decisions in one shot. More specifically, the proposed model consists of four modules, Pre-training Network and modified interior point methods with adversarial networks (Modified IPMAN) as core modules, and discriminator generative adversarial network (Dis-GAN) and Volt convolutional neural network (Volt-CNN) as ancillary modules. In particular, the generator in SCAN is trained by the core modules in sequence so as to form an end-to-end mode from data to decision. Numerical experiments based on IEEE 33-bus and 123-bus systems have validated the effectiveness and efficiency of the proposed method.

关键词： end-to-end learning microgrids voltage profile improvement generative adversarial network

HTML PDF 收藏

深度学习中的对抗性攻击和防御 Feature Article

任奎, Tianhang Zheng, 秦湛, Xue Liu

《工程（英文）》 2020年第6卷第3期页码 346-360 doi: 10.1016/j.eng.2019.12.012

摘要：

在深度学习（deep learning, DL）算法驱动的数据计算时代，确保算法的安全性和鲁棒性至关重要。最近，研究者发现深度学习算法无法有效地处理对抗样本。这些伪造的样本对人类的判断没有太大影响，但会使深度学习模型输出意想不到的结果。最近，在物理世界中成功实施的一系列对抗性攻击证明了此问题是所有基于深度学习系统的安全隐患。因此有关对抗性攻击和防御技术的研究引起了机器学习和安全领域研究者越来越多的关注。本文将介绍深度学习对抗攻击技术的理论基础、算法和应用。然后，讨论了防御方法中的一些代表性研究成果。这些攻击和防御机制可以为该领域的前沿研究提供参考。此外，文章进一步提出了一些开放性的技术挑战，并希望读者能够从所提出的评述和讨论中受益。

关键词：机器学习深度神经网络对抗实例对抗攻击对抗防御

HTML PDF 收藏

基于Wasserstein GAN的新一代人工智能小样本数据增强方法——以生物领域癌症分期数据为例 Article

刘宇飞, 周源, 刘欣, 董放, 王畅, 王子鸿

《工程（英文）》 2019年第5卷第1期页码 156-163 doi: 10.1016/j.eng.2018.11.018

摘要：本研究提出了一种基于生成对抗网络（generative adversarial network，GAN）和深度神经网络（deep neural network，DNN）分类器的方法。

关键词：人工智能生成式对抗网络深度神经网络小样本癌症

HTML PDF 收藏

深度学习的几何学解释 Article

雷娜, 安东生, 郭洋, 苏科华, 刘世霞, 罗钟铉, 丘成桐, 顾险峰

《工程（英文）》 2020年第6卷第3期页码 361-374 doi: 10.1016/j.eng.2019.09.010

摘要：

本文从几何角度来理解深度学习，特别是提出了生成对抗网络（GAN）的最优传输（OT）观点。自然数据集具有内在的模式，该模式可被概括为流形分布原理，即同一类高维数据分布于低维流形附近。 GAN主要完成流形学习和概率分布变换两项任务。其中，后者可以用经典的OT方法来实现。从OT的角度来看，生成器用于计算OT映射，而判别器用于计算生成数据分布与真实数据分布之间的Wasserstein距离；两者都可以归结为一个凸优化过程。此外， OT理论揭示了生成器与判别器之间的内在关系是协作的而不是竞争的，并且解释了模式崩溃的根本原因。在此基础上，我们提出了一种新的生成模型，该模型利用自编码器（AE）进行流形学习，并利用OT映射进行概率分布变换。这个AE-OT模型提升了深度学习理论的严谨性和透明性、提高了计算的稳定性和效率，尤其是避免了模式崩溃问题。实验结果验证了我们的假设，并充分展示了我们提出的AE-OT模型的优点。

关键词：生成，对抗，深度学习，最优传输，模式崩溃

HTML PDF 收藏

SmartPaint：一种基于生成式对抗神经网络的人机协同绘画系统 Special Feature on Intelligent Design

Lingyun SUN, Pei CHEN, Wei XIANG, Peng CHEN, Wei-yue GAO, Ke-jun ZHANG

《信息与电子工程前沿（英文）》 2019年第20卷第12期页码 1644-1656 doi: 10.1631/FITEE.1900386

摘要：当前人工智能在模仿和大批量生产设计作品中扮演重要角色（如电商广告），而在与用户合作创作时表现欠佳。人们有能力使用草图表达创意想法，但缺乏专业绘画技巧完成精美画作。已有人工智能方法无法基于用户输入草图的语义输出具有艺术美感的画作。本文开发了一种基于生成式对抗神经网络的人机协作绘画系统——SmartPaint，支持人机合作创作动漫风景画作。该系统使用动漫图像数据及其相应语义标注图、边缘检测图训练生成式对抗神经网络。通过此种方式，该系统能够同时理解动漫风格以及风景图像中物体的语义和空间关系。在使用中，用户输入草图作为语义标注图，系统自动为其合成边缘图；根据合成的边缘图生成具有恰当风格纹理的画作，从而稳定地处理多样化草图。实验证明该系统可有效满足用户创作需求，生成高质量动漫风格画作。

关键词：协同绘画；深度学习；图像生成

HTML PDF 收藏

生成式人工智能的惊人进展引发惊讶和担忧

Dana Mackenzie

《工程（英文）》 2023年第25卷第6期页码 9-11 doi: 10.1016/j.eng.2023.04.004

HTML PDF 收藏

基于双向深度生成模型和功能磁共振成像数据的大脑编码和解码 Review

杜长德, 李劲鹏, 黄利皆, 何晖光

《工程（英文）》 2019年第5卷第5期页码 948-953 doi: 10.1016/j.eng.2019.03.010

摘要：

通过功能磁共振成像（fMRI）进行大脑编码和解码是视觉神经科学的两个重要方面。尽管以前的研究人员在大脑编码和解码模型方面取得了显著进步，但是现有方法仍需要使用先进的机器学习技术进行改进。例如，传统方法通常会分别构建编码和解码模型，并且容易对小型数据集过度拟合。实际上，有效地统一编码和解码过程可以进行更准确的预测。在本文中，我们首先回顾了现有的编码和解码方法，并讨论了“双向”建模策略的潜在优势。接下来，在体系结构和计算规则方面，我们证明了深度神经网络和人类视觉通路之间存在的对应关系。此外，深度生成模型[如变分自编码器（VAE）和生成对抗网络（GAN）]在大脑编码和解码研究中产生了可喜的成果。最后，我们提出了最初为机器翻译任务设计的对偶学习方法，该方法通过利用大规模未配对数据提高了编码和解码模型的效果。

关键词：大脑编码和解码功能磁共振成像深度神经网络深度生成模型双重学习

HTML PDF 收藏

Toward Trustworthy Decision-Making for Autonomous Vehicles: A Robust Reinforcement Learning Approach with Safety Guarantees

Xiangkun He,Wenhui Huang,Chen Lv,

《工程（英文）》 doi: 10.1016/j.eng.2023.10.005

摘要： While autonomous vehicles are vital components of intelligent transportation systems, ensuring the trustworthiness of decision-making remains a substantial challenge in realizing autonomous driving. Therefore, we present a novel robust reinforcement learning approach with safety guarantees to attain trustworthy decision-making for autonomous vehicles. The proposed technique ensures decision trustworthiness in terms of policy robustness and collision safety. Specifically, an adversary model is learned online to simulate the worst-case uncertainty by approximating the optimal adversarial perturbations on the observed states and environmental dynamics. In addition, an adversarial robust actor-critic algorithm is developed to enable the agent to learn robust policies against perturbations in observations and dynamics. Moreover, we devise a safety mask to guarantee the collision safety of the autonomous driving agent during both the training and testing processes using an interpretable knowledge model known as the Responsibility-Sensitive Safety model. Finally, the proposed approach is evaluated through both simulations and experiments. These results indicate that the autonomous driving agent can make trustworthy decisions and drastically reduce the number of collisions through robust safety policies.

关键词： Autonomous vehicle Decision-making Reinforcement learning Adversarial attack Safety guarantee

HTML PDF 收藏

针对工业故障分类系统的单变量攻击及其防御 Article

卓越, Yuri A.W. Shardt, 葛志强

《工程（英文）》 2022年第19卷第12期页码 240-251 doi: 10.1016/j.eng.2021.07.033

摘要：

近年来，工业过程故障分类系统主要是由数据驱动的，得益于大量的数据模式，基于深度神经网络的模型显著地提高了故障分类的准确性。但是，这些数据驱动模型容易受到对抗攻击，因此，在样本上的微小扰动会导致模型提供错误的故障预测。最近的研究已经证明了机器学习模型的脆弱性以及对抗样本的广泛存在。本文针对安全、关键的工业故障分类系统提出了一种具有极端约束的黑盒攻击方法：只扰动一个变量来制作对抗样本。此外，为了将对抗样本隐藏在可视化空间中，本文使用了雅可比矩阵来引导扰动变量的选择，使降维空间中的对抗样本对人眼不可见。利用单变量攻击（OVA）方法，文本探究了不同工业变量和故障类别的脆弱性，有助于理解故障分类系统的几何特征。基于攻击方法，文本还提出了相应的对抗训练防御方法，该方法能够有效地防御单变量攻击，并提高分类器的预测精度。在实验中，将本文所提出的方法在田纳西-伊士曼过程（TEP）和钢板（SP）故障数据集上进行了测试。本文探索了变量和故障类别的脆弱相关性，并验证了各种分类器和数据集的单变量攻击和防御方法的有效性。对于工业故障分类系统，单变量攻击方法的攻击成功率接近（在TEP上）甚至高于（在SP 上）目前最有效的一阶白盒攻击方法（该方法需要对所有变量进行扰动）。

关键词：对抗样本黑盒攻击工业数据安全故障分类系统

HTML PDF 收藏

A multi-sensor relation model for recognizing and localizing faults of machines based on network analysis

《机械工程前沿（英文）》 2023年第18卷第2期 doi: 10.1007/s11465-022-0736-9

摘要： Recently, advanced sensing techniques ensure a large number of multivariate sensing data for intelligent fault diagnosis of machines. Given the advantage of obtaining accurate diagnosis results, multi-sensor fusion has long been studied in the fault diagnosis field. However, existing studies suffer from two weaknesses. First, the relations of multiple sensors are either neglected or calculated only to improve the diagnostic accuracy of fault types. Second, the localization for multi-source faults is seldom investigated, although locating the anomaly variable over multivariate sensing data for certain types of faults is desirable. This article attempts to overcome the above weaknesses by proposing a global method to recognize fault types and localize fault sources with the help of multi-sensor relations (MSRs). First, an MSR model is developed to learn MSRs automatically and further obtain fault recognition results. Second, centrality measures are employed to analyze the MSR graphs learned by the MSR model, and fault sources are therefore determined. The proposed method is demonstrated by experiments on an induction motor and a centrifugal pump. Results show the proposed method’s validity in diagnosing fault types and sources.

关键词： fault recognition fault localization multi-sensor relations network analysis graph neural network

HTML PDF 收藏

Novel interpretable mechanism of neural networks based on network decoupling method

《工程管理前沿（英文）》 2021年第8卷第4期页码 572-581 doi: 10.1007/s42524-021-0169-x

摘要： The lack of interpretability of the neural network algorithm has become the bottleneck of its wide application. We propose a general mathematical framework, which couples the complex structure of the system with the nonlinear activation function to explore the decoupled dimension reduction method of high-dimensional system and reveal the calculation mechanism of the neural network. We apply our framework to some network models and a real system of the whole neuron map of Caenorhabditis elegans. Result shows that a simple linear mapping relationship exists between network structure and network behavior in the neural network with high-dimensional and nonlinear characteristics. Our simulation and theoretical results fully demonstrate this interesting phenomenon. Our new interpretation mechanism provides not only the potential mathematical calculation principle of neural network but also an effective way to accurately match and predict human brain or animal activities, which can further expand and enrich the interpretable mechanism of artificial neural network in the future.

关键词： neural networks interpretability dynamical behavior network decouple

HTML PDF 收藏

Multiscale computation on feedforward neural network and recurrent neural network

Bin LI, Xiaoying ZHUANG

《结构与土木工程前沿（英文）》 2020年第14卷第6期页码 1285-1298 doi: 10.1007/s11709-020-0691-7

摘要： Homogenization methods can be used to predict the effective macroscopic properties of materials that are heterogenous at micro- or fine-scale. Among existing methods for homogenization, computational homogenization is widely used in multiscale analyses of structures and materials. Conventional computational homogenization suffers from long computing times, which substantially limits its application in analyzing engineering problems. The neural networks can be used to construct fully decoupled approaches in nonlinear multiscale methods by mapping macroscopic loading and microscopic response. Computational homogenization methods for nonlinear material and implementation of offline multiscale computation are studied to generate data set. This article intends to model the multiscale constitution using feedforward neural network (FNN) and recurrent neural network (RNN), and appropriate set of loading paths are selected to effectively predict the materials behavior along unknown paths. Applications to two-dimensional multiscale analysis are tested and discussed in detail.

关键词： multiscale method constitutive model feedforward neural network recurrent neural network

HTML PDF 收藏

Heat, mass, and work exchange networks

Zhiyou CHEN, Jingtao WANG

《化学科学与工程前沿（英文）》 2012年第6卷第4期页码 484-502 doi: 10.1007/s11705-012-1221-5

摘要： Heat (energy), water (mass), and work (pressure) are the most fundamental utilities for operation units in chemical plants. To reduce energy consumption and diminish environment hazards, various integration methods have been developed. The application of heat exchange networks (HENs), mass exchange networks (MENs), water allocation heat exchange networks (WAHENs) and work exchange networks (WENs) have resulted in the significant saving of energy and water. This review presents the main works related to each network. The similarities and differences of these networks are also discussed. Through comparing and discussing these different networks, this review inspires researchers to propose more efficient and convenient methods for the design of existing exchange networks and even new types of networks including multi-objective networks for the system integration in order to enhance the optimization and controllability of processes.

关键词： process system engineering integration methods heat exchange network mass exchange network work exchange network