Refine
Document Type
- Working Paper (13)
- Report (8)
- Article (7)
- Preprint (3)
Has Fulltext
- yes (31)
Keywords
- Optimization (9)
- Optimierung (6)
- Benchmarking (3)
- Modeling (3)
- 3D Printing (2)
- Combined simulation (2)
- Evolutionary Computation (2)
- Evolutionärer Algorithmus (2)
- Globale Optimierung (2)
- Machine Learning (2)
- Maschinelles Lernen (2)
- Soft Computing (2)
- Surrogate (2)
- Surrogate Models (2)
- Surrogate-based (2)
- 3D-Druck (1)
- Abgasreinigung (1)
- Algorithm Tuning (1)
- Algorithmus (1)
- Angewandte Mathematik (1)
- Artificial intelligence (1)
- Automated Learning (1)
- Automation (1)
- BBOB (1)
- Bayesian Learning (1)
- Bayesian Optimization (1)
- Bayesian Regression (1)
- Big Data (1)
- Big data platform (1)
- Business Intelligence (1)
- Ccomputational fluid dynamics (1)
- Cognition (1)
- Computational fluid dynamics (1)
- Conditional inference tree (1)
- Data Analysis (1)
- Data Mining (1)
- Data Modelling (1)
- Data-Warehouse-Konzept (1)
- Decision tree (1)
- Design of Experiments (1)
- Discrete Optimization (1)
- Electrostatic Precipitator (1)
- Ensemble Methods (1)
- Ensemble based modeling (1)
- Entdeckendes Lernen (1)
- Erfahrungsbericht (1)
- Event Detection (1)
- Evolutionsstrategie (1)
- Expensive Optimization (1)
- Experiment (1)
- Experimental Algorithmics (1)
- Faserverbundwerkstoffe (1)
- Feature selection (1)
- Flowcurve (1)
- Forschendes Lernen (1)
- Function Approximation (1)
- Funktionstest (1)
- Gaussian Process (1)
- Genetische Algorithmen (1)
- Genetisches Programmieren (1)
- Health condition monitoring (1)
- Hot rolling (1)
- Imputation (1)
- Industrie 4.0 (1)
- Industry 4.0 (1)
- Knowledge extraction (1)
- Kognition (1)
- Kriging (1)
- Künstliche Intelligenz (1)
- Lineare Regression (1)
- Machine learning (1)
- Massive Online Analysis (1)
- Meta-model (1)
- Metaheuristics (1)
- Metaheuristik (1)
- Metal (1)
- Metamodel (1)
- Metamodell (1)
- Metamodels (1)
- Model Selection (1)
- Modelierung (1)
- Muschelknautz Method of Modelling (1)
- Neural and Evolutionary Computing (1)
- Numerische Strömungssimulation (1)
- On-line Algorithm (1)
- Parallelization (1)
- Performance (1)
- Predictive Analytics (1)
- Promotion (1)
- R (1)
- Referenzmodell (1)
- Regression (1)
- SAP (1)
- SPOT (1)
- Sensor placement (1)
- Sensortechnik (1)
- Sequential Parameter Optimization (1)
- Signalanalyse (1)
- Simulation (1)
- Simulation-based Optimization (1)
- Stacked Generalization (1)
- Stacking (1)
- Standardisierung (1)
- Staubabscheider (1)
- Structural Health Monitoring (1)
- Surrogate Mod (1)
- Surrogate Model (1)
- Surrogate model (1)
- Surrogate model based optimization (1)
- Surrogates (1)
- Tauchrohrtiefe (1)
- Taxonomie (1)
- Taxonomy (1)
- Test Function (1)
- Test function generator (1)
- Time Series (1)
- Trinkwasser (1)
- Univariate Data (1)
- Variable reduction (1)
- Verunreinigung (1)
- Wasserverteilung (1)
- Water distribution systems (1)
Institute
- Fakultät für Informatik und Ingenieurwissenschaften (F10) (31) (remove)
EventDetectR: An efficient Event Detection System (EDS) capable of detecting unexpected water quality conditions. This approach uses multiple algorithms to model the relationship between various multivariate water quality signals. Then the residuals of the models were utilized in constructing the event detection algorithm, which provides a continuous measure of the probability of an event at every time step. The proposed framework was tested for water contamination events with industrial data from automated water quality sensors. The results showed that the framework is reliable with better performance and is highly suitable for event detection.
Data pre-processing is a key research topic in data mining because it plays a
crucial role in improving the accuracy of any data mining algorithm. In most
real world cases, a significant amount of the recorded data is found missing
due to most diverse errors. This loss of data is nearly always unavoidable.
Recovery of missing data plays a vital role in avoiding inaccurate data
mining decisions. Most multivariate imputation methods are not compatible
to univariate datasets and the traditional univariate imputation techniques
become highly biased as the missing data gap increases. With the current
technological advancements abundant data is being captured every second.
Hence, we intend to develop a new algorithm that enables maximum
utilization of the available big datasets for imputation. In this paper, we
present a Seasonal and Trend decomposition using Loess (STL) based
Seasonal Moving Window Algorithm, which is capable of handling patterns
with trend as well as cyclic characteristics. We show that the algorithm is
highly suitable for pre-processing of large datasets.
This paper introduces CAAI, a novel cognitive architecture for artificial intelligence in cyber-physical production systems. The goal of the architecture is to reduce the implementation effort for the usage of artificial intelligence algorithms. The core of the CAAI is a cognitive module that processes declarative goals of the user, selects suitable models and algorithms, and creates a configuration for the execution of a processing pipeline on a big data platform. Constant observation and evaluation against performance criteria assess the performance of pipelines for many and varying use cases. Based on these evaluations, the pipelines are automatically adapted if necessary. The modular design with well-defined interfaces enables the reusability and extensibility of pipeline components. A big data platform implements this modular design supported by technologies such as Docker, Kubernetes, and Kafka for virtualization and orchestration of the individual components and their communication. The implementation of the architecture is evaluated using a real-world use case.
When researchers and practitioners in the field of
computational intelligence are confronted with real-world
problems, the question arises which method is the best to
apply. Nowadays, there are several, well established test
suites and well known artificial benchmark functions
available.
However, relevance and applicability of these methods to
real-world problems remains an open question in many
situations. Furthermore, the generalizability of these
methods cannot be taken for granted.
This paper describes a data-driven approach for the
generation of test instances, which is based on
real-world data. The test instance generation uses
data-preprocessing, feature extraction, modeling, and
parameterization. We apply this methodology on a classical
design of experiment real-world project and generate test
instances for benchmarking, e.g. design methods, surrogate
techniques, and optimization algorithms. While most
available results of methods applied on real-world
problems lack availability of the data for comparison,
our future goal is to create a toolbox covering multiple
data sets of real-world projects to provide a test
function generator to the research community.
When using machine learning techniques for learning a function approximation from given data it is often a difficult task to select the right modeling technique.
In many real-world settings is no preliminary knowledge about the objective function available. Then it might be beneficial if the algorithm could learn all models by itself and select the model that suits best to the problem.
This approach is known as automated model selection. In this work we propose a
generalization of this approach.
It combines the predictions of several into one more accurate ensemble surrogate model. This approach is studied in a fundamental way, by first evaluating minimalistic ensembles of only two surrogate models in detail and then proceeding to ensembles with three and more surrogate models.
The results show to what extent combinations of models can perform better than single surrogate models and provides insights into the scalability and robustness of the approach. The study focuses on multi-modal functions topologies, which are important in surrogate-assisted global optimization.
To maximize the throughput of a hot rolling mill,
the number of passes has to be reduced. This can be achieved by maximizing the thickness reduction in each pass. For this purpose, exact predictions of roll force and torque are required. Hence, the predictive models that describe the physical behavior of the product have to be accurate and cover a wide range of different materials.
Due to market requirements a lot of new materials are tested and rolled. If these materials are chosen to be rolled more often, a suitable flow curve has to be established. It is not reasonable to determine those flow curves in laboratory, because of costs and time. A strong demand for quick parameter determination and the optimization of flow curve parameter with minimum costs is the logical consequence. Therefore parameter estimation and the optimization with real data, which were collected during previous runs, is a promising idea. Producers benefit from this data-driven approach and receive a huge gain in flexibility when rolling new
materials, optimizing current production, and increasing quality. This concept would also allow to optimize flow curve parameters, which have already been treated by standard methods. In this article, a new data-driven approach for predicting the physical behavior of the product and setting important parameters is presented.
We demonstrate how the prediction quality of the roll force and roll torque can be optimized sustainably. This offers the opportunity to continuously increase the workload in each pass to the theoretical maximum while product quality and process stability can also be improved.
Die Reinhaltung der Luft spielt heute mehr denn je eine wichtige Rolle. In Gesellschaft und Politik wird über Dieselfahrverbote in Innenstädten diskutiert, um die Feinstaubbelastung in den Städten zu senken. Besonders die Industrie steht vor der Aufgabe, den Partikelausstoß zu senken und Wege zu finden, um eine gesunde Luft zu wahren. Zur Abgasreinigung werden oft Filter eingesetzt. Diese weisen aber hohe Energieverluste auf. Die ständige Reinigung oder der Wechsel der Filter kostet Zeit und Geld. Daher ist neben Filtern eine der gängigsten Methoden die Abgasreinigung durch Staubabscheider. Staubabscheider funktionieren filterlos. Dadurch entfällt eine wiederkehrende Filterreinigung, beziehungsweise der regelmäßige Filtertausch. Die Technik der Staubabscheider hat ihren Ursprung in der Natur. Aus der Betrachtung von Zyklonen (in den Tropen vorkommende Wirbelstürme) wurde ein Verfahren entwickelt, um staubhaltige Fluide von den Verunreinigungen zu trennen. Die Abgasreinigung mittels Zyklon-Staubabscheider wird in vielen verschiedenen
Industrien eingesetzt, heutzutage meist als Vorabscheider. Beispiele hierfür sind die
braunkohleverarbeitende Industrie, die Gesteinsindustrie und die papier- oder holzverarbeitende Industrie, insbesondere dort, wo viel Staub oder auch größere Späne in die Luft gelangen. Auch im Alltag sind Zyklon-Staubabscheider zu finden. Hier kommen sie in beutellosen Staubsaugern oder als Vorabscheider von Staubsaugern bei der Holzverarbeitung zum Einsatz.
Die Vorgänge im Staubabscheider-Zyklon sind bereits durch mathematische Modelle beschrieben worden. Hierbei handelt es sich um Näherungen, jedoch nicht um
die exakte Abbildung der Realität, weswegen bis heute die Modelle immer wieder weiterentwickelt und verbessert werden. Eine CFD (Computional Fluid Dynamics)Simulation bringt meist die besten Ergebnisse, ist jedoch sehr aufwendig und muss für jeden Staubabscheider neu entwickelt werden. Daher wird noch immer an der Weiterentwicklung der mathematischen Modelle gearbeitet, um eine Berechnung zu optimieren, die für alle Staubabscheider gilt. Muschelknautz hat in diesem Bereich über Jahre hinweg geforscht und so eine der
wichtigsten Methoden zur Berechnung von Zyklonabscheidern entwickelt. Diese stimmt oft sehr gut mit der Realität überein. Betrachtet man jedoch die Tiefe des Tauchrohres im Zyklon, fällt auf, dass der Abscheidegrad maximal wird, wenn das Tauchrohr nicht in den Abscheideraum ragt, sondern mit dem Deckel des Zyklons abschließt. Dieses Phänomen tritt weder bei den durchgeführten CFD-Simulationen noch bei den durchgeführten Messungen am Bauteil auf. Ziel der Arbeit ist es, diese Unstimmigkeit zwischen Berechnung und Messung zu untersuchen und Gründe hierfür herauszufinden. Darum wird zunächst der Stand der Technik und das Muschelknautz’sche Modell
vorgestellt, um im Anschluss die Berechnungsmethode genauer zu untersuchen. So soll festgestellt werden, ob die Ursache der Abweichungen zur Realität bei einer Analyse der Berechnungsmethode ersichtlich wird. Beispielsweise soll überprüft werden, ob die Schlussfolgerung einer maximalen Abscheideleistung bei minimaler Tauchrohrtiefe von speziellen Faktoren abhängt. Es wird eine Reihe von Beispielrechnungen durchgeführt, mit deren Hilfe der Zusammenhang
von Abscheidegrad und Tauchrohrtiefe ersichtlich wird. Hierbei werden die Geometrieparameter des Abscheiders variiert, um deren Einfluss auf die Tauchrohrtiefe
zu untersuchen.
Architecural aproaches are considered to simplify the generation of re-usable building blocks in the field of data warehousing. While SAP’s Layer Scalable Architecure (LSA) offers a reference model for creating data warehousing infrastructure based on SAP software, extented reference models are needed to guide the integration of SAP and non-SAP tools. Therefore, SAP’s LSA is compared to the Data Warehouse Architectural Reference Model (DWARM), which aims to cover the classical data warehouse topologies.
Verunreinigungen im Wassernetz können weite Teile der Bevölkerung unmittelbar gefährden. Gefahrenpotenziale bestehen dabei nicht nur durch mögliche kriminelle Handlungen und terroristische Anschläge. Auch Betriebsstörungen, Systemfehler und Naturkatastrophen können zu Verunreinigungen führen.
In this paper we present a comparison of different data driven modeling methods. The first instance of a data driven linear Bayesian model is compared with several linear regression models, a Kriging model and a genetic programming model.
The models are build on industrial data for the development of a robust gas sensor.
The data contain limited amount of samples and a high variance.
The mean square error of the models implemented in a test dataset is used as the comparison strategy.
The results indicate that standard linear regression approaches as well as Kriging and GP show good results,
whereas the Bayesian approach, despite the fact that it requires additional resources, does not lead to improved results.