检索结果-内蒙古大学图书馆

An Empirical Evaluation of Constrained Feature Selection

SN Computer Science 2022年第6期3卷 1-25页

作者： Bach, Jakob Zoller, Kolja Trittenbach, Holger Schulz, Katrin Böhm, Klemens Department of Informatics Karlsruhe Institute of Technology (KIT) Am Fasanengarten 5 Baden-Württemberg Karlsruhe 76131 Germany Department of Mechanical Engineering Karlsruhe Institute of Technology (KIT) Kaiserstraße 12 Baden-Württemberg Karlsruhe 76131 Germany Faculty of Mechanical Engineering and Mechatronics Karlsruhe University of Applied Sciences Moltkestraße 30 Baden-Württemberg Karlsruhe 76133 Germany

While feature selection helps to get smaller and more understandable prediction models, most existing feature-selection techniques do not consider domain knowledge. One way to use domain knowledge is via constraints on sets of selected features. However, the impact of constraints, e.g., on the predictive quality of selected features, is currently unclear. This article is an empirical study that evaluates the impact of propositional and arithmetic constraints on filter feature selection. First, we systematically generate constraints from various types, using datasets from different domains. As expected, constraints tend to decrease the predictive quality of feature sets, but this effect is non-linear. So we observe feature sets both adhering to constraints and with high predictive quality. Second, we study a concrete setting in materials science. This part of our study sheds light on how one can analyze scientific hypotheses with the help of constraints. © 2022, The Author(s).

关键词： Constraints Domain knowledge Feature selection Theory-guided data science

来源：评论

学校读者我要写书评

暂无评论

Data Fusion and State Estimation Using Belief Propagation in Gas Distribution Networks

Data Fusion and State Estimation Using Belief Propagation in...

引用

International Universities Power Engineering Conference (UPEC)

作者： Goekhan Demirel Steven de Jongh Felicitas Mueller Thomas Leibfried Institute of Electrical Energy Systems and High Voltage Technology (IEH) Karlsruhe Institute of Technology Karlsruhe Germany Institute for Automation and Applied Informatics (IAI) Karlsruhe Institute of Technology Eggenstein-Leopoldshafen Germany

ISBN: (数字)9781665455053

ISBN: (纸本)9781665455060

This paper proposes a solution to the state estimation problem in gas networks using the distributed belief propagation (BP) algorithm. Power system identification applications require precise and robust state estimatiors as well as various sensor information. Compared to augmenting the power system with a very large number of sensors, a limited number of sensors and probabilistic graphical models can be used to infer the system state and reduce hardware investments. A novel BP algorithm propagates the pressure quantities at nodes in the gas network based on pressure manometer signals and applies a correction based on the information of neighboring nodes in the fusion step by using additional supporting sensors. Finally, the data fusion algorithm is demonstrated for a 14-node gas distribution network based on real data. This paper presents a novel algorithm aimed at tackling the traditional weighted least squares method to validate the developed novel approach in order to highlight the advantage of the distributed inference algorithm over traditional methods.

关键词： Power engineering Graphical models Data integration Distribution networks Probabilistic logic Inference algorithms Power systems

来源：评论

学校读者我要写书评

暂无评论

Assessing the combination of differential flatness and deterministic automata for controllable hybrid systems

Assessing the combination of differential flatness and deter...

引用

IEEE Conference on Decision and Control

作者： Frederik Zahn Tobias Kleinert Veit Hagenmeyer Institute for Automation and Applied Informatics Karlsruhe Institute of Technology Karlsruhe Germany Chair of Information and Automation Systems for Process and Material Technology RWTH Aachen University Aachen Germany

ISBN: (数字)9781665467612

ISBN: (纸本)9781665467629

Hybrid automata are a powerful model class to analyze and control technical systems with nonlinear continuous dynamics and discrete-event dynamics. However, due to their complexity, controllability is generally hard to achieve in hybrid automata, although it is a desirable property of engineered systems. In the present paper, we assess a novel model class of controllable deterministic hybrid automata, which we call Flat Hybrid Automata. This model class is based on the concept of differential flatness for the nonlinear continuous dynamics and on the property of determinism and controllability for the discrete event dynamics. We review the formal definition and we give the respective conditions and a proof for its controllability. An electrical network example illustrates the methodology and shows its potential regarding modelling, control and system design.

关键词： Adaptation models Analytical models Power system dynamics Automata Switches Controllability Trajectory

来源：评论

学校读者我要写书评

暂无评论

Hierarchical-Stochastic Model Predictive Heating Control for a Grid-Interactive Multi-Zone Residential Building with Distributed Energy Resources

SSRN

引用

SSRN 2023年

作者： Langner, Felix Frahm, Moritz Wang, Weimin Matthes, Jörg Hagenmeyer, Veit Institute for Automation and Applied Informatics Karlsruhe Institute of Technology Eggenstein-Leopoldshafen76344 Germany Department of Engineering Technology University of North Carolina at Charlotte CharlotteNC28223 United States

Using Model Predictive Control (MPC) is a promising method for enabling grid-interactive efficient buildings. Since MPC relies on a building model and the forecasts of external disturbances to derive optimal inputs, the uncertainties due to forecast errors and model inaccuracies can deteriorate the control performance. Most existing MPC studies for built environment are deterministic MPC (DMPC) without consideration of uncertainties even though several different methods (e.g., stochastic MPC) are available in literature to deal with them. Hence, in the present paper, a hierarchical-stochastic MPC (HSMPC) is proposed and its performance is compared with deterministic, stochastic, and hierarchical MPC for power management in a residential building with distributed energy resources. The control objective is the cost-optimal scheduling of a heat pump, a battery for energy storage, and a rooftop photovoltaic system. Measurement data is used to identify the building model. The results for one-week simulation in winter show that 1) the deterministic MPC results in an unacceptable level of temperature constraint violations in two out of five rooms;2) the hierarchical MPC can reduce the temperature constraint violations to an acceptable level at the expense of increased cost;3) the stochastic MPC achieves the same reduction in temperature constraint violations as the hierarchical MPC but at slightly lower costs;and 4) the new proposed HSMPC results in both lower temperature constraint violations and lower financial expenses than the use of stochastic or hierarchical MPC individually. © 2023, The Authors. All rights reserved.

关键词： Uncertainty analysis

来源：评论

学校读者我要写书评

暂无评论

StateFlow as a Tool of the Hybrid X-in-the-Loop Technique in the Model-Based Design Approach to the Control Systems Design in Industrial automation

StateFlow as a Tool of the Hybrid X-in-the-Loop Technique in...

引用

MECHATRONIKA International Symposium

作者： Martin Juhás Bohuslava Juhásová Igor Halenár Milan Daňo Lenka Halenárová Fedor Burčiar Faculty of Materials Science and Technology in Trnava Institute of Applied Informatics Automation and Mechatronics Slovak University of Technology in Bratislava Trnava Slovakia

ISBN: (数字)9798350394900

ISBN: (纸本)9798350394917

The rapid dynamics of changes in the definition of requirements when deploying new technologies means an increase in demands on the volume and scope of skills and knowledge of experts in the design of solutions and their implementation. PLC-based control systems designed for industrial automation can also be included in this problem area. The article deals with the possibility of rapid design and testing of PLC control algorithms using X-in-the-Loop techniques of Model-based Design approach. The control algorithm is implemented using visual programming techniques in the form of state diagrams utilizing the Matlab StateFlow tool. The given proposal is analyzed from the point of view of different approaches to the implementation of communication between the development PC and the real control PLC SIMATIC S7-300 connected to the real part of the model production line. Data exchange is carried out in two ways. The first way of communication is the exchange of uncompressed and compressed images of inputs and outputs via IP communication. The second way is to use the OPC UA architecture. The result of the article is a case study of the use of the XiL approach in the design phase of the control of the AFB Factory production line module with a summary of the possibilities, advantages and disadvantages of alternative partial solutions.

关键词： Visualization automation Mechatronics Programming Control systems Mathematical models Production facilities Proposals MATLAB Testing

来源：评论

学校读者我要写书评

暂无评论

Automated Style-Aware Selection of Annotated Pre-Training Databases in Biomedical Imaging

Automated Style-Aware Selection of Annotated Pre-Training Da...

引用

IEEE International Symposium on Biomedical Imaging

作者： Miguel Molina-Moreno Marcel P. Schilling Markus Reischl Ralf Mikut Department of Signal Theory and Communications Universidad Carlos III de Madrid Madrid Spain Institute for Automation and Applied Informatics Karlsruhe Institute of Technology Eggenstein-Leopoldshafen Germany

Due to the broad use of deep learning and its need for big data, annotated and available databases for different tasks are constantly appearing. Nevertheless, they often remain unexploited due to the difficulty of effectively performing transfer learning between different databases. In medical imaging, the task of transfer learning is challenging due to: the variety of image modalities, organ/cell shapes, etc., and the lack of available and annotated data. In this paper, we propose an automated pipeline for predicting the similarity values of new database compared to known annotated databases. The system consists of an autoencoder trained on a comprehensive loss function that considers image reconstruction, style features, and dataset membership. A similarity measure is defined based on the resulting 2D latent space, which is demonstrated to have a correlation with the pre-training results on not annotated databases. Hence, our similarity measure could be used to select the most suitable known database for transfer learning or domain adaptation.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Automated Production Data Integration and Processing via Simulation Based Digital Twin

Automated Production Data Integration and Processing via Sim...

引用

MECHATRONIKA International Symposium

作者： Fedor Burčiar Martin Juhás Bohuslava Juhásová Pavel VaŽan Igor Halenár Lenka Halenárová Institute of Applied informatics Automation and Mechatronics Faculty of Materials Science and Technology in Trnava Slovak University of Technology in Bratislava Trnava Slovakia

ISBN: (数字)9798350394900

ISBN: (纸本)9798350394917

Recent developments in the production sector are slowly beginning to test Small and Medium Enterprises (SME) and their ability to compete with the remainder of the market, especially the enterprises specializing in large-scale mass production. The ever-increasing demand gradually creates a large gap between the various competitors and often forces the SME to take risks when making important decisions regarding production control. One of the solutions to this problem can be the use of a simulation based Digital Twin (DT) as a decisionmaking tool for production control. Naturally, the established on-site hardware and software differs for each job-shop, thus making the creation of generally applicable solutions almost impossible. The biggest issue that needs to be solved is the discrepancy in data between the on-site software and the newly created DT of the production line. Manual data transformation and transfer is one of the options for the integration process, but it requires an experienced operator with a specific skillset to perform the actions necessary for seamless execution of this process. Automating the process of data integration between two systems helps alleviate the entry requirements for the DT technology implementation into a job-shop environment. The presented case study explores the possibility of using automated algorithms for production order data evaluation and processing which can be applied to other data types with minor adjustments to the methods proposed.

关键词： Mass production Mechatronics Data integration Production control Reliability engineering Software Hardware Digital twins Testing Sorting

来源：评论

学校读者我要写书评

暂无评论

AI-based automated active learning for discovery of hidden dynamic processes: A use case in light microscopy

arXiv

引用

arXiv 2023年

作者： Friederich, Nils Sitcheu, Angelo Yamachui Neumann, Oliver Eroğlu-Kayıkçı, Süheyla Prizak, Roshan Hilbert, Lennart Mikut, Ralf Institute for Automation and Applied Informatics Karlsruhe Institute of Technology Hermann-von-Helmholtz-Platz 1 Eggenstein-Leopoldshafen76344 Germany Institute of Biological and Chemical Systems Karlsruhe Institute of Technology Hermann-von-Helmholtz-Platz 1 Eggenstein-Leopoldshafen76344 Germany Zoological Institute Karlsruhe Institute of Technology Fritz-Haber-Weg 4 Karlsruhe76131 Germany

In the biomedical environment, experiments assessing dynamic processes are primarily performed by a human acquisition supervisor. Contemporary implementations of such experiments frequently aim to acquire a maximum number of relevant events from sometimes several hundred parallel, non-synchronous processes. Since in some high-throughput experiments, only one or a few instances of a given process can be observed simultaneously, a strategy for planning and executing an efficient acquisition paradigm is essential. To address this problem, we present two new methods in this paper. The first method, Encoded Dynamic Process (EDP), is Artificial Intelligence (AI)-based and represents dynamic processes so as to allow prediction of pseudo-time values from single still images. Second, with Experiment automation Pipeline for Dynamic Processes (EAPDP), we present a Machine Learning Operations (MLOps)-based pipeline that uses the extracted knowledge from EDP to efficiently schedule acquisition in biomedical experiments for dynamic processes in practice. In a first experiment, we show that the pre-trained State-Of-The-Art (SOTA) object segmentation method Contour Proposal Networks (CPN) works reliably as a module of EAPDP to extract the relevant object for EDP from the acquired three-dimensional image stack. © 2023, CC BY.

关键词： Pipelines

来源：评论

学校读者我要写书评

暂无评论

Generation of synthetic Point Clouds for MEMS LiDAR Sensor

TechRxiv

引用

TechRxiv 2022年

作者： Berens, Felix Elser, Stefan Reischl, Markus Institute for Automation and Applied Informatics Karlsruhe Institute of Technology Karlsruhe76344 Germany Institute for Artificial Intelligence Ravensburg-Weingarten University of Applied Sciences Weingarten88250 Germany

Sensory data is essential for the training of methods in autonomous driving like object detection, odometry, or SLAM. MEMS LiDAR sensors can be very valuable for autonomous vehicles because they are less prone to shock and wear compared to motorized optomechanical LiDAR sensors. Recording real-world data is complicated and expensive. An alternative is simulated data, but for MEMS LiDAR sensors there is no publicly available software to simulate this type of sensor. With this paper, we introduce a method to simulate data recorded by a MEMS LiDAR sensor like the Blickfeld Cube 1 (and other MEMS LiDAR sensors as well). For this, we use the open-source autonomous driving simulation environment CARLA (our method is available online1). We compare our synthetic point cloud with a real-world point cloud and evaluate the similarity. Moreover, we demonstrate the application of our method on the problem of the optimal sensor configuration. © 2022, CC BY.

关键词： Autonomous vehicles

来源：评论

学校读者我要写书评

暂无评论

Forecasting energy time series with profile neural networks 20

Forecasting energy time series with profile neural networks

引用

11th ACM International Conference on Future Energy Systems, e-Energy 2020

作者： Heidrich, Benedikt Turowski, Marian Ludwig, Nicole Mikut, Ralf Hagenmeyer, Veit Karlsruhe Institute of Technology Institute for Automation and Applied Informatics Germany

ISBN: (纸本)9781450380096

Forecasting the energy demand is essential for network operators to balance the grid, in particular with the increasing share of renewable energy sources. Neural networks, especially deep neural networks, have shown promising results in recent forecasting tasks. However, they often struggle learning periodicities in time series efficiently. In line with the finding that deep learning can be improved with statistical information, we introduce profile neural networks based on the fast and promising convolutional neural networks. The underlying idea of profile neural networks is that decomposing periodic energy time series into a standard load profile, a trend, and a colorful noise module improves the forecasting accuracy. The proposed deep neural network architecture is applied to real-world electricity data from buildings on a university campus, more specifically of one building with strong seasonal variation and one building with weak seasonal variation. The new architecture outperforms current state-of-the-art deep learning benchmark models regarding the forecasting accuracy on forecast horizons of one day and one week-ahead, improving the mean absolute scaled error by up to 25%, as well as regarding the trade-off between training time and accuracy. © 2020 Owner/Author.

关键词： Time series

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：