检索结果-内蒙古大学图书馆

2023 IEEE International Conference on systems, Man, and Cybernetics, SMC 2023

作者： Hilgert, Eric Schwung, Andreas Graduate School for Applied Research in North Rhine-Westphalia Department of Technology and Systems Bochum Germany South Westphalia University of Applied Sciences Department of Automation Technology and Learning Systems Soest Germany

ISBN: (纸本)9798350337020

In this paper, we present a comprehensive approach for designing and analyzing control systems with minmax constraint controllers and machine learning-based virtual sensors. By leveraging the Standard Nonlinear Operator Form (SNOF), we establish the necessary conditions for global asymptotic stability and demonstrate their applicability through an illustrative example based on a modified plant model from the literature. The proposed methodology effectively handles non-linearities and constraints, ensuring stability while providing a systematic procedure for constructing a well-formed SNOF by integrating the plant, virtual sensor, and controller. The successful application of this method in the example highlights its potential for addressing complex control problems involving min-max constraints and virtual sensors in real-world scenarios. This paper contributes to the growing body of knowledge in this area and sets the stage for future advancements, including the exploration of transforming other machine learning architectures into the SNOF and extending the stability analysis to accommodate different types of nonlinearities and constraints. © 2023 IEEE.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Neural Data Fusion Enhanced PD Control for Precision Drone Landing in Synthetic Environments

Neural Data Fusion Enhanced PD Control for Precision Drone L...

引用

2025 IEEE International Conference on Mechatronics, ICM 2025

作者： Altinses, Diyar Torres, David Orlando Salazar Lier, Stefan Schwung, Andreas South Westphalia University of Applied Sciences Department of Automation Technology and learning systems Soest Germany South Westphalia University of Applied Sciences Department of Logistics and Supply Chain Management Meschede Germany

ISBN: (纸本)9798331533892

Unmanned aerial vehicles are increasingly used in applications like surveillance, mapping, and delivery, where precise and safe landings are crucial, especially on small, designated platforms in multimodal systems. However, achieving consistent landing accuracy is challenging due to sensor inaccuracies, environmental disturbances, and dynamic conditions. This paper introduces a synthetic drone simulation environment to test landing scenarios and control strategies. We propose a control framework combining a classical PD controller with a neural network-based data fusion model that corrects sensor noise, particularly in GPS and UWB data, to improve landing precision. Simulation results demonstrate that our approach enhances landing accuracy and robustness compared to traditional methods. © 2025 IEEE.

关键词： Drones

来源：评论

学校读者我要写书评

暂无评论

Synthetic Image Generation for Visual Particle Size Distribution Estimation based on U-NET Convolutional Networks

Synthetic Image Generation for Visual Particle Size Distribu...

引用

2023 IEEE International Conference on Device Intelligence, Computing and Communication Technologies, DICCT 2023

作者： Chavan, Akshay Schmidt, Antonius Klopries, Hendrik Aufderheide, Dominik Faculty of Electrical Engineering Soest Germany South Westphalia University of Applied Sciences Automation Technology and Learning Systems Faculty of Electrical Engineering Soest Germany

ISBN: (纸本)9781665474924

Many applications in the field of bulk material handling require the determination of the Particle Size Distribution (PSD) of the handled materials. Typically, this can only be achieved by manual sampling and a subsequent laboratory analysis. However, the availability of novel image segmentation algorithms based on Convolutional Neural Networks (CNNs) leads to the possibility for an implementation of visual PSD estimation techniques. One prominent example of such approaches is the U-Net image segmentation architecture. However, one fundamental problem for the training of the underlying CNN is the difficulty to gather ground truth data, since the exact correspondence between a set of images and the underlying particle sizes cannot be achieved in practice. This paper introduces a novel synthetic image generation procedure for the visual PSD estimation by using the U-NET structure. The overall approach is validated and verified by using a prototypic implementation of a U-NET segmentation pipeline. It is shown, that the proposed synthetic image dataset generation can be applied for all visual PSD estimation based on machine learning algorithms. © 2023 IEEE.

关键词： Particle size analysis

来源：评论

学校读者我要写书评

暂无评论

Data Imputation Techniques Using the Bag of Functions: Addressing Variable Input Lengths and Missing Data in Time Series Decomposition

Data Imputation Techniques Using the Bag of Functions: Addre...

引用

IEEE International Conference on Industrial technology (ICIT)

作者： David Orlando Salazar Torres Diyar Altinses Andreas Schwung Department of Automation Technology and learning systems South Westphalia University of Applied Sciences Soest Germany

ISBN: (数字)9798331521950

ISBN: (纸本)9798331521967

In time series analysis, the ability to effectively handle data with varying input lengths and missing data is crucial for accurate modeling. This paper presents the Bag-of-Functions-Driven Imputation framework, which leverages sequence-length independent techniques to decompose time series data while accommodating inputs of differing sizes. Unlike traditional methods that require uniform input lengths, the Padding-BoF framework employs a flexible encoding approach, allowing for the integration of variable-length time series and missing elements in the data. Through a series of experiments, we demonstrate that the BoF framework not only ensures precise reconstruction of the original data but also enhances data imputation capabilities by utilizing decomposed components. The results show that this method offers significant advantages in scenarios involving irregular sampling and disparate operational cycles, making it a valuable tool for applications in fields such as finance, healthcare, and industrial monitoring.

关键词： learning systems Adaptation models Accuracy Time series analysis Finance Medical services Imputation Data models Encoding Monitoring

来源：评论

学校读者我要写书评

暂无评论

Bag-of-Future: Time Series Forecasting by Continuous Function Candidates

Bag-of-Future: Time Series Forecasting by Continuous Functio...

引用

IEEE International Symposium on Industrial Electronics (ISIE)

作者： Hendrik Klopries Vignesh Chandramouli Andreas Schwung Automation Technology and Learning Systems South Westphalia University of Applied Sciences Soest Germany

ISBN: (数字)9798350394085

ISBN: (纸本)9798350394092

Time series forecasting is one of the crucial tasks in many industrial fields and bears financial, resource, and control benefits. The leading solutions in this domain are often associated with machine learning approaches. The drawback in many approaches is the missing interpretability of the lookback and the predicted time series. Therefore, we propose a novel method named Bag-of-Future, to introduce interpretability in a backcast as well as forecast of the given series. Our approach aims to select and predict multiple patterns for the input series, based on continuous functions in a preselected bag of functions. Furthermore, stacking multiple encodings results in a continuous parameterization of the given dataset. The predicted bags are used to give precise forecasts and can unveil black box systems. Among synthetic as well as multiple industrial energy datasets we present competitive results in terms of performance, interpretability and pattern detection of the forecast as well as the given time series. Lastly, we present ablations on the main parameters of our approach, evaluated on multiple criterions to provide parameter recommendations.

关键词： Measurement Industrial electronics Time series analysis Stacking Fitting Switches Machine learning

来源：评论

学校读者我要写书评

暂无评论

Foreign Object Separation in Bulk Good systems Using Machine learning and Image Processing

Foreign Object Separation in Bulk Good Systems Using Machine...

引用

IEEE International Conference on Industrial technology (ICIT)

作者： Marlon Löppenberg Hendrik Klopries Jonas Bartsch Andreas Schwung Automation Technology and Learning Systems South Westphalia University of Applied Sciences Soest Germany

ISBN: (数字)9798350340266

ISBN: (纸本)9798350340273

One of the major goals in bulk good systems is to provide high-quality material. Sometimes the bulk good is corrupted by foreign objects, due to insufficient process observations, human mistakes, or unforeseen events. These particles are unwanted in the process. Furthermore, they can lead to unnecessary plant shutdowns or can even lock sub-parts of the plant. Current solutions work with filter techniques to eject foreign objects and ensure high quality. However, many solutions are difficult to implement due to process constraints and give no precise information about the corruption level and their objects. Therefore, a machine learning-based method is developed to detect objects in the bulk material and control an ejection system. In addition, based on the object position, the conveyor speed and air pressure injectors are controlled to separate the detected object during free fall. Consequently, the closed-loop structure and workflow for a custom air-pressure ejection process are presented. The framework is successfully applied in a modular bulk good system and provides adequate results in terms of object detection and object separation.

关键词： YOLO Training Adaptation models Three-dimensional displays Pollution Atmospheric modeling Image processing

来源：评论

学校读者我要写书评

暂无评论

Structured Graph Generation by Evolutionary Algorithm for Program Code Development

Structured Graph Generation by Evolutionary Algorithm for Pr...

引用

Annual Conference of Industrial Electronics Society

作者： Marlon Löppenberg Andreas Schwung Automation Technology and Learning Systems South Westphalia University of Applied Sciences Soest Germany

ISBN: (数字)9781665464543

ISBN: (纸本)9781665464550

Understanding and interpreting complex coupled systems remains one of the biggest challenges in the world. Examples of these applications range from industrial manufacturing to the temporal characteristics of real-world conditions. To address this challenge, this paper presents a novel approach to structured program code development based on graph generation. The problem is considered from the perspective of an inductive link prediction problem structured by an evolutionary algorithm. The self-adaptation of relevant knowledge takes place in a closed loop, where systematic relationships are constantly improved and extended. The required system behaviour is mapped step by step, taking into account constraints, limitations and expert knowledge. Structured graph generation is used to represent logic functions and interpret complex coupled relationships. The presented strategy enables targeted plant control through customised program code concepts, which are used to optimise processes and increase efficiency. The approach is validated through the design of interpretable programmable control logic on an industrial manufacturing process and obtain a comparable solution to the work of a trained professional. The achieved results demonstrate the next level of independent self-optimisation in learning and interpreting logical relationships in automation.

关键词： Programmable control Codes Systematics Manufacturing processes Process control Evolutionary computation Prediction algorithms Logic functions Manufacturing Logic

来源：评论

学校读者我要写书评

暂无评论

Node Reservation Based Incremental learning Network for Object Detection

Node Reservation Based Incremental Learning Network for Obje...

引用

IEEE International Conference on Industrial technology (ICIT)

作者： M. Tahasanul Ibrahim Nikhil Limaye Andreas Schwung Department of Automation Technology and Learning Systems South Westphalia University of Applied Sciences Soest Germany

ISBN: (数字)9798350340266

ISBN: (纸本)9798350340273

Object categorization is a crucial element in AI-driven computer vision systems, with its influence spanning from advanced surveillance technologies to basic projects. This field faces a key challenge in strategically incorporating new objects into existing networks, balancing between efficient training and precise object detection. Our paper introduces an innovative two-step method to fluidly integrate new objects into current networks, maintaining the accuracy of existing object detection accuracy and the integrity of prior knowledge. This method relies on the network's foresight in predicting upcoming class additions during the initial training, allocating specific layers for future expansions. These layers are then selectively activated when new classes are added, simplifying retraining processes and preserving pre-existing knowledge. This approach is particularly beneficial for time-sensitive and resource-limited systems, offering proactive resource allocation and adaptability to new class introductions.

关键词： Training Knowledge engineering Computer vision Adaptation models Surveillance Neural networks Object detection

来源：评论

学校读者我要写书评

暂无评论

Pose Estimation for Robot Grasping Using Dual Quaternions and a Modified PointLoc Architecture

Pose Estimation for Robot Grasping Using Dual Quaternions an...

引用

IEEE Conference on Industrial Electronics and Applications (ICIEA)

作者： Johannes Pöppelbaum William Budiatmadjaja Mochammad Rizky Diprasetya Andreas Schwung Department of Automation Technology and Learning Systems South Westphalia University of Applied Sciences Soest Germany

ISBN: (数字)9798350360868

ISBN: (纸本)9798350360875

In this work, we propose a modified, downscaled version of the PointLoc architecture, intended for pose estimation from an input point cloud in an enclosed, small scale environment. This predicted pose shall later be used in a downstream reinforcement learning for a robot manipulator grasping task. With the proposed smaller PointLoc architecture, we are able to precisely predict the orientation and position of the respective object to grab, while simultaneously reducing the computational complexity and hardware requirements of the original PointLoc architecture. We introduce a dual quaternion valued regressor to incorporate an inductive bias and reduce the amount of trainable parameters within the regressor by almost 16 times in comparison to a conventional real valued counterpart without loss of performance.

关键词： Point cloud compression Solid modeling Three-dimensional displays Quaternions Pose estimation Computer architecture Grasping

来源：评论

学校读者我要写书评

暂无评论

Improving Quaternion Neural Networks with Quaternionic Activation Functions

arXiv

引用

arXiv 2024年

作者： Pöppelbaum, Johannes Schwung, Andreas Department of Automation Technology and Learning Systems South Westphalia University of Applied Sciences Soest Germany

In this paper, we propose novel quaternion activation functions where we modify either the quaternion magnitude or the phase, as an alternative to the commonly used split activation functions. We define criteria that are relevant for quaternion activation functions, and subsequently we propose our novel activation functions based on this analysis. Instead of applying a known activation function like the ReLU or Tanh on the quaternion elements separately, these activation functions consider the quaternion properties and respect the quaternion space H. In particular, all quaternion components are utilized to calculate all output components, carrying out the benefit of the Hamilton product in e.g. the quaternion convolution to the activation functions. The proposed activation functions can be incorporated in arbitrary quaternion valued neural networks trained with gradient descent techniques. We further discuss the derivatives of the proposed activation functions where we observe beneficial properties for the activation functions affecting the phase. Specifically, they prove to be sensitive on basically the whole input range, thus improved gradient flow can be expected. We provide an elaborate experimental evaluation of our proposed quaternion activation functions including comparison with the split ReLU and split Tanh on two image classification tasks using the CIFAR-10 and SVHN dataset. There, especially the quaternion activation functions affecting the phase consistently prove to provide better performance. © 2024, CC BY-NC-ND.

关键词： Gradient methods

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：