Near-infrared (NIR) technology has gained wide acceptance in practical processes and is now the measurement of choice in many sectors. However, with increasing spectral dimensionality, it is challenging to establish a...
详细信息
Near-infrared (NIR) technology has gained wide acceptance in practical processes and is now the measurement of choice in many sectors. However, with increasing spectral dimensionality, it is challenging to establish a prediction model with satisfactory stability and generalization. Stochastic configuration networks (SCNs) based on supervisory learning mechanism have demonstrated significant advantages in developing nonlinear learners. However, existing incremental learning strategies make it difficult to achieve fast convergence while obtaining a suitable-scale network in high-dimensional spectra modeling. In addition, the linear or regularization weight estimation methods are vulnerable to outliers and noise in NIR analysis. To accelerate model construction and improve model performance in high-dimensional spectra analysis, the adaptive robust SCN (AR-SCN) algorithm is proposed in this work, which can perform adaptive incremental learning according to the prediction residual and robustly estimate the output weights by the global-local shrinkage strategy. Comparison results on three benchmark NIR datasets and real-world gasoline blending process verify the effectiveness of the proposed method. Compared with the state-of-the-art SCNs, the AR-SCN method can simultaneously improve the construction efficiency and robustness of SCNs.
Increasing energy demand in today's world emphasizes the importance of optimal scheduling for distributed energy resources to minimize energy costs and greenhouse gas (GHG) emissions. The efficiency of this decisi...
详细信息
Increasing energy demand in today's world emphasizes the importance of optimal scheduling for distributed energy resources to minimize energy costs and greenhouse gas (GHG) emissions. The efficiency of this decision -making process relies on accurate modeling. In this paper, reinforcement learning (RL), an artificial intelligence -based approach, is proposed to optimize the energy management system (EMS) of an energy hub (EH). This EH contains renewable energy resources (RER), a combined heat and power (CHP), and a gas furnace. In order to meet electrical and thermal energy demand, available options such as day -ahead and real-time purchases from the main grid, RERs, and natural gas consumption are managed, with the preference of RERs to minimize GHG emissions and energy costs. With the adaptable RL method, a non-linear model of the CHP operation is constructed, considering the operational costs of the CHP. Furthermore, the natural gas tariff is varied according to the consumption level of the microgrid. Finally, this paper presents an RL-based method for EMS optimization of an EH with day -ahead and real-time scheduling, applied to a 24 -hour case study with linear and nonlinear modeling of the problem and sensitivity analysis of the parameters. Corresponding simulation results show the efficiency of the presented approach.
The development process of complex products is typically characterized by high integration across multiple domains and disciplines. Design changes are an integral part of this process, and effective management of freq...
详细信息
Generative Adversarial Networks (GANs) are widely used for modeling complex data. However, the dynamics of the gradient descent-ascent (GDA) algorithms, often used for training GANs, have been notoriously difficult to...
详细信息
Generative Adversarial Networks (GANs) are widely used for modeling complex data. However, the dynamics of the gradient descent-ascent (GDA) algorithms, often used for training GANs, have been notoriously difficult to analyze. We study these dynamics in the case where the discriminator is kernel-based and the true distribution consists of discrete points in Euclidean space. Prior works have analyzed the GAN dynamics in such scenarios via simple linearization close to the equilibrium. In this work, we show that linearized analysis can be grossly inaccurate, even at moderate distances from the equilibrium. We then propose an alternative non-linear yet tractable second moment model. The proposed model predicts the convergence behavior well and reveals new insights about the role of the kernel width on convergence rate, not apparent in the linearized analysis. These insights suggest certain shapes of the kernel offer both fast local convergence and improved global convergence. We corroborate our theoretical results through simulations.
The use of tilting photogrammetry technology to build a refined urban 3D model has become the industry research hotspot with the rapid development of China's "digital city."To investigate and analyze the...
详细信息
作者:
Huang, WenjunCui, YunduanLi, HuiyunWu, XinyuUniv Chinese Acad Sci
Sch Artificial Intelligence Beijing 101408 Peoples R China Chinese Acad Sci
Shenzhen Inst Adv Technol Shenzhen 518055 Guangdong Peoples R China Chinese Acad Sci
Shenzhen Inst Adv Technol CAS Key Lab Human Machine Intelligence Synergy Sys Shenzhen 518055 Guangdong Peoples R China Chinese Acad Sci
Shenzhen Inst Adv Technol Guangdong Hong Kong Macao Joint Lab Human Machine Shenzhen 518055 Guangdong Peoples R China
Gaussian process (GP) offers a robust solution for modeling the dynamics of unmanned surface vehicles (USV) in model-based reinforcement learning (MBRL). However, the rapidly increasing computational complexity with a...
详细信息
Gaussian process (GP) offers a robust solution for modeling the dynamics of unmanned surface vehicles (USV) in model-based reinforcement learning (MBRL). However, the rapidly increasing computational complexity with a large sample capacity of GP limits its application in complex scenarios that require substantial samples to cover the state space. In this article, a novel probabilistic MBRL approach, probabilistic neural networks model predictive control (PNMPC) is proposed to tackle this issue. With an iterative learning framework, PNMPC properly models the USV dynamics using neural networks from a probabilistic perspective to avoid the computational complexity associated with sample capacity. Employing this model to effectively propagate system uncertainties, a model predictive control (MPC) policy is developed to robustly control the USV against external disturbances. Evaluated by position-keeping and multiple targets-tracking scenarios on a real USV data-driven simulation, the proposed method consistently demonstrates its significant superiority in both model accuracy and control performance compared to not only GP model-based approaches but also the probabilistic neural networks-based MBRL baselines, across various scales of external disturbances.
The continuously increasing amount of noisy data demands the development of accurate and efficient models for analysis, modeling, and control. In this article, we propose a novel data-driven moment matching method whi...
详细信息
ISBN:
(纸本)9798331540920;9783907144107
The continuously increasing amount of noisy data demands the development of accurate and efficient models for analysis, modeling, and control. In this article, we propose a novel data-driven moment matching method which employs Tikhonov regularization in the Reproducing Kernel Hilbert Spaces (RKHSs). Specifically, considering a realistic scenario in which the system's plant is unknown and only noisy measured data are available, we provide an estimation of the moment of the unknown plant by solving a regularized optimization problem on RKHS. For, we first demonstrate that the estimation of the moment can be improved via tuning the regularization term, and further, we show under which condition the effect of the transient improves the performance of the estimation. Then, we construct a parameterized model characterized by a kernel-based output mapping. Finally, the proposed data-driven approach is validated and discussed by means of a DC-to-DC C ' uk converter driven by a Van der Pol oscillator.
In manufacturing process monitoring, the obtained data is always affected by multi-noise resources with different statistics features, including non-homologous noises, which adversely affect dataanalysis, especially ...
详细信息
In manufacturing process monitoring, the obtained data is always affected by multi-noise resources with different statistics features, including non-homologous noises, which adversely affect dataanalysis, especially the performance of data driven models which are increasingly developed in manufacturing applications. Due to the lack of prior knowledge of the noises, traditional denoising methods based on modeling noise distribution with statistical features have major limitations in denoising non-homologous noises. To address this issue, this paper presents a denoising method for monitoring data with non-homologous noises based on causal inference. The causal relationship between raw data from multi sources, and the noises and monitoring data from sensors is modeled. To remove the influence of noises on monitoring data, the instruments variable is introduced into causal model, which forming new back-door paths between non-homologous noises sources. Then the noise could be denoised according to the prior causal knowledge of the causal model. The method is verified in both simulation and actual machining environment, which lays a data foundation for establishing accurate and stable prediction and control models during manufacturing processes.
作者:
Oliveira, DanielaTeixeira, LeonorAlvelos, HelenaUniv Aveiro
Dept Econ Management Ind Engn & Tourism DEGEIT P-3810193 Aveiro Portugal Univ Aveiro
Inst Elect & Informat Engn Aveiro IEETA Lab Associado Sistemas Inteligentes LASI Dept Econ Management Ind Engn & Tourism P-3810193 Aveiro Portugal Univ Aveiro
Ctr Invest & Desenvolvimento Matemat & Aplicacoes Dept Econ Management Ind Engn & Tourism P-3810193 Aveiro Portugal
The energetic and environmental crisis currently being experienced around the world has brought an exponential growth on the demand for wind energy, and, consequently, real challenges for companies in this sector. Sin...
详细信息
The energetic and environmental crisis currently being experienced around the world has brought an exponential growth on the demand for wind energy, and, consequently, real challenges for companies in this sector. Since this industry is relatively new, its production processes require a significant amount of human intervention and have a high variability in the process and high lead times. Therefore, there is a need to focus on the quality of the processes and their improvement. This paper presents the use of the Six Sigma DMAIC approach integrated with Business process Management, to reduce the number of defects and the repair time in a leading wind blade manufacturing company in Portugal. During the measurement and analysis phases, data was collected to understand the performance of the process and to identify the main problems and their causes. The data was analysed through several graphical and statistical techniques. In the analysis phase, processmodeling was also utilized, using Business process Model and Notation, to identify the critical steps that are responsible for producing the majority of defects. One month after beginning to use this approach, it was possible to notice a reduction in the number of defects by more than 30%, in the repair time by 14%, and an increase in the process sigma level by more than 100%. Thus, this hybrid approach has allowed the company to become more competitive and responsive to the constant growth in demand. (c) 2023 The Authors. Published by ELSEVIER B.V.
This paper introduces our development of software designed for OCD (Optical Critical Dimension) modeling, utilizing 3D graphics design functionality. In the OCD metrology, the role of analysis software is crucial for ...
详细信息
ISBN:
(纸本)9781510672178;9781510672161
This paper introduces our development of software designed for OCD (Optical Critical Dimension) modeling, utilizing 3D graphics design functionality. In the OCD metrology, the role of analysis software is crucial for accurately and precisely extracting CD parameters from intricate device structures. Our software incorporates calculation engines grounded in Physics and Machine Learning - RCWA (Rigorous Coupled Wave analysis) and DL (Deep Learning). The software's advanced 3D modeling engine supports complex structure manipulation and precise adjustments of a broad range of parameters, including optical properties. This facilitates detailed device geometry exploration through a cohesive interface. The DL algorithm has been developed ensuring consistency between RCWA and DL predictions, essential for accurate and rapid OCD metrology. We have conducted a comprehensive evaluation process to assess the consistency between RCWA calculations and 3D representations, encompassing both 2D and 3D structures. Further evaluation is planned, specifically focusing on real patterned wafers.
暂无评论