TY - CHAP A1 - Linner, Thomas A1 - Märkl, Markus A1 - Künzner, Florian A1 - Hagl, Rainer T1 - Analyzing and Optimizing Fixed-Point Operations on a MATLAB Simulink based Motion Control System T2 - 2023 10th International Conference on Electrical Engineering, Computer Science and Informatics (EECSI) N2 - This paper deals with the analysis and optimization of fixed-point operations in the context of motion control systems and the MathWorks MATLAB® Simulink® environment. It presents a procedure for estimating the maximum arithmetic error of fixed-point operations inside a modeled control loop, which can be used to verify compliance with a defined precision. Additionally, it shows the results of a runtime analysis for different numerical data types in combination with two fixed-point operations: addition and multiplication. In order to improve the fixed-point performance for MATLAB® Simulink® models, two optimized implementations are provided, which can improve the fixed-point operations by a factor up to 3, compared to the automatically generated operation implementations. The paper results are based on the MATLAB® R2022b release. KW - Codes KW - fixed-point optimization KW - Hardware KW - Mathematical models KW - model-based design KW - Motion planning KW - Numerical models KW - precision motion control KW - real-time KW - Runtime KW - Software packages Y1 - 2023 U6 - https://doi.org/10.1109/EECSI59885.2023.10295686 SP - 486 EP - 492 ER - TY - CHAP A1 - Märkl, Markus A1 - Hagl, Rainer A1 - Heldwein, M. T1 - Noise Reduction for Sub-Nanometer Position Stability in High Dynamic Linear Drives T2 - 14th International Symposium on Linear Drives for Industry Applications (LDIA) Y1 - 2023 SP - 1 EP - 5 ER - TY - JOUR A1 - Vivekanandan, Deepak A1 - Wirth, Samuel A1 - Karlbauer, Patrick A1 - Klarmann, Noah T1 - A Reinforcement Learning Approach for Scheduling Problems with Improved Generalization through Order Swapping JF - Machine Learning and Knowledge Extraction N2 - The scheduling of production resources (such as associating jobs to machines) plays a vital role for the manufacturing industry not only for saving energy, but also for increasing the overall efficiency. Among the different job scheduling problems, the Job Shop Scheduling Problem (JSSP) is addressed in this work. JSSP falls into the category of NP-hard Combinatorial Optimization Problem (COP), in which solving the problem through exhaustive search becomes unfeasible. Simple heuristics such as First-In, First-Out, Largest Processing Time First and metaheuristics such as taboo search are often adopted to solve the problem by truncating the search space. The viability of the methods becomes inefficient for large problem sizes as it is either far from the optimum or time consuming. In recent years, the research towards using Deep Reinforcement Learning (DRL) to solve COPs has gained interest and has shown promising results in terms of solution quality and computational efficiency. In this work, we provide an novel approach to solve the JSSP examining the objectives generalization and solution effectiveness using DRL. In particular, we employ the Proximal Policy Optimization (PPO) algorithm that adopts the policy-gradient paradigm that is found to perform well in the constrained dispatching of jobs. We incorporated a new method called Order Swapping Mechanism (OSM) in the environment to achieve better generalized learning of the problem. The performance of the presented approach is analyzed in depth by using a set of available benchmark instances and comparing our results with the work of other groups. KW - Job Shop Scheduling KW - Production Scheduling KW - Reinforcement Learning KW - Markov Decision Process KW - Generalization KW - Industry 4.0 Y1 - 2023 U6 - https://doi.org/10.3390/make5020025 VL - 5 IS - 2 SP - 418 EP - 430 ER - TY - CHAP A1 - Parzinger, Michael A1 - Schanda, Ulrich T1 - Analysis of spectra for sound insulation using methods of mathematical statistics and AI - first approaches T2 - Proceedings of Forum Acusticum 2023 N2 - Within the framework of a research focus at the TH Rosenheim on prediction methods for sound and impact sound insulation in timber constructions, methods of mathematical statistics and artificial intelligence are applied to sound insulation. To estimate the potential of those methods, one-third octave band spectra of measured sound insulation of sand-lime brickwork have been analyzed first. On selected data sets for certain building constructions, physically based calculation approaches according to the \backslashDIN{} series of standards, are compared with purely statistical methods such as GAMLSS (Generalized Additive Models for Location, Scale and Shape Parameters). The parameters derived from these procedures can be used for prediction purposes. The interval estimators resulting from these methods are compared. In addition, methods to classify the separating construction based on measurements are discussed. Thereby, in situ, measurements are used in addition to laboratory measurements. KW - Klassifikation KW - LDA/QDA KW - MLclassifier KW - Regression KW - Schall KW - SVM Y1 - 2023 ER - TY - CHAP A1 - Parzinger, Michael A1 - Schanda, Ulrich T1 - Analyse von Spektren zur Schalldämmung mittels Methoden der KI – erste Ansätze T2 - Fortschritte der Akustik - DAGA Y1 - 2023 ER - TY - CHAP A1 - Kruse, Tobias A1 - Schöpfer, Fabian A1 - Schanda, Ulrich A1 - Mayr, Andreas R. T1 - Input data for the prediction of noise from technical equipment using a heat pump T2 - Proceedings of Forum Acusticum 2023 KW - Berechnungsmethoden KW - Körperschall KW - Körperschallquellen KW - TechnischeAkustik KW - Koerperschall Y1 - 2023 ER - TY - CHAP A1 - Kruse, Tobias A1 - Mayr, Andreas R. A1 - Schöpfer, Fabian A1 - Schanda, Ulrich T1 - Kennwerte für die schalltechnische Prognose von gebäudetechnischen Anlagen am Beispiel einer Wärmepumpe T2 - Fortschritte der Akustik - DAGA KW - Körperschall KW - Installationen KW - TechnischeAkustik KW - GebaeudetechnischeAnlagen KW - Koerperschall Y1 - 2023 ER - TY - CHAP A1 - Hierl, Robin A1 - Schanda, Ulrich A1 - Schöpfer, Fabian A1 - Weise, Yvonne T1 - Experimental development of a timber floor system with improved impact sound insulation in the low-frequency range T2 - Proceedings of Forum Acusticum 2023 Y1 - 2023 ER - TY - CHAP A1 - Schanda, Ulrich A1 - Dullinger, Sandra T1 - Variationen raumakustischer Messungen in einem Unterrichtsraum T2 - Fortschritte der Akustik - DAGA Y1 - 2023 ER - TY - CHAP A1 - Schanda, Ulrich A1 - Hoffbauer, Elias A1 - Lachenmayr, Winfried T1 - Detektion von Flatterechos mittels der Autokorrelationsfunktion der Raumimpulsantwort T2 - Fortschritte der Akustik - DAGA Y1 - 2023 ER - TY - JOUR A1 - Rabold, Andreas A1 - Marxt, Michaela A1 - Schanda, Ulrich T1 - Geeignete Beurteilungsgrößen für die Luftschalldämmung von Haustrennwänden – Bewertung der Schalldämmung bei üblichen Wohngeräuschen JF - Bauphysik KW - Luftschalldämmung KW - Gebäudetrennwand KW - Bewertung Y1 - 2023 U6 - https://doi.org/10.1002/bapi.202200038 VL - 45 IS - 1 SP - 21 EP - 34 ER - TY - CHAP A1 - Schöpfer, Fabian A1 - Mayr, Andreas R. A1 - Schanda, Ulrich A1 - Bayer, Jeremias T1 - Analysis of measured transmission functions in timber buildings T2 - Proceedings of Forum Acusticum 2023 Y1 - 2023 ER - TY - BOOK A1 - Hagl, Rainer A1 - King, Frank A. A1 - Zentgraf, Peter T1 - Ingenieurinformatik BT - Eine Einführung mit MATLAB, Simulink und Stateflow N2 - Umfassende Programmierkenntnisse werden für Ingenieur:innen immer wichtiger, vor allem im Bereich der Produktentwicklung. Die Autoren erklären die Grundlagen der Ingenieurinformatik beispielhaft anhand von MATLAB, Simulink und Stateflow. Neben den Grundkenntnissen lernen Leser:innen auch die Vorteile von MATLAB im Vergleich zu anderen Programmiersprachen kennen und anzuwenden. Die dabei vorgestellten Zusammenhänge und Prinzipien sind allgemeingültig und lassen sich daher auch problemlos auf andere Programmiersprachen übertragen. Konkrete Beispiele und Problemstellungen aus der Ingenieurpraxis werden vorgestellt und Lösungsansätze aufgezeigt. Das Lehrbuch richtet sich an Studierende aller ingenieurwissenschaftlichen Studiengänge. Es eignet sich besonders für die Bachelorausbildung in den Studienschwerpunkten Elektro- und Informationstechnik, Mechatronik, Maschinenbau, Automatisierungstechnik sowie Energie- und Gebäudetechnik. Für die zweite Auflage wurde das Lehrbuch umfassend überarbeitet und an die aktuellen Programmversionen angepasst. Weitere Übungen mit Lösungen sowie Hinweise zur Erstellung eigener Apps und Programmbeispiele stehen auf plus.hanser-fachbuch.de zur Verfügung. Der auf dem Cover abgebildete Programmcode und die animierte Rakete wurden im Projekt „water rocket“ an der Technischen Hochschule Rosenheim im Jahr 2021 unter der Leitung von Prof. Zentgraf entwickelt. Die Projekt-Unterlagen und der zugehörige Filmbeitrag können unter www.th-rosenheim.de/rt-WaterRocket eingesehen werden. KW - Grundlagen der Programmierung KW - Grafische Bedienoberflächen KW - Zahlenformate KW - Numerische Integration KW - Zeitgesteuerte Systeme (Simulink) KW - Ereignisdiskrete Systeme (Stateflow) KW - Paralleles Rechnen KW - Symbolisches Rechnen Y1 - 2023 U6 - https://doi.org/10.3139/9783446477308 PB - Carl Hanser Verlag GmbH & Co. KG CY - München ET - 2 ER - TY - CHAP A1 - Wohlschläger, Maximilian A1 - Leiter, Nina A1 - Dietlmeier, Maximilian A1 - Löder, Martin G.J. A1 - Versen, Martin A1 - Laforsch, Christian T1 - Comparison of Two Classification Methods Trained with FD-FLIM Data to Identify and Distinguish Plastics from Environmental Materials T2 - 2023 International Joint Conference on Neural Networks (IJCNN) N2 - Previous research on identifying plastic types and differentiating plastics from environmental material is promising by utilizing the specific fluorescence lifetime, but the evaluation still has to be automated. Therefore, an automated Gaussian analysis is developed for evaluating frequency-domain fluorescence lifetime images of plastics and environmental materials. Furthermore, we applied a “Multilayer Perceptron” and “Random Forest Classifier” to the data resulting from the Gaussian analysis of the frequency domain fluorescence lifetime imaging microscopy data. The classification results show high F1-scores, whereby the best “Multilayer Perceptron” and “Random Forest Classifier” achieved an F1-score of 90%. Thus, identifying and differentiating plastics and environmental materials is possible by applying a “Multilayer Perceptron” or “Random Forest Classifier” to the Gaussian-analyzed imaged fluorescence lifetime data. KW - Fluorescence KW - Microscopy KW - Plastics KW - FD-FLIM KW - Neural networks KW - Environment KW - Frequency-domain analysis KW - Gaussian analysis KW - MLP KW - Multilayer perceptrons KW - Random forests KW - RFC Y1 - 2023 U6 - https://doi.org/10.1109/IJCNN54540.2023.10191054 SP - 1 EP - 9 ER - TY - CHAP A1 - Wohlschläger, Maximilian A1 - Khan, Yamna A1 - Leiter, Nina A1 - Versen, Martin A1 - Löder, Martin A1 - Laforsch, Christian T1 - Development of a BLOB-detection algorithm based on DoG to detect Plastic in an environmental matrix using FD-FLIM T2 - Optica Sensing Congress 2023 (AIS, FTS, HISE, Sensors, ES) N2 - The direct identification of plastics in an environmental matrix is heavily researched. We successfully developed a BLOB-detection algorithm based on differences of Gaussians to identify HDPE particles in an artificial environmental matrix using FD-FLIM. KW - Diode lasers KW - Fluorescence lifetime imaging KW - Phase shift KW - Neural networks KW - Optical filters KW - Spatial resolution Y1 - 2023 U6 - https://doi.org/10.1364/ES.2023.EW4E.4 ER - TY - CHAP A1 - Wohlschläger, Maximilian A1 - Khan, Yamna A1 - Leiter, Nina A1 - Versen, Martin A1 - Löder, Martin A1 - Laforsch, Christian T1 - Combining BLOB-Detection and MLP to Detect and Identify Plastics in an Environmental Matrix T2 - 2023 IEEE Sensors Applications Symposium (SAS) N2 - Environmental pollution by plastics is an increasing problem. However, state-of-the-art methods have significant disadvantages in detecting and identifying plastics directly in an environmental matrix. In this study, we propose a blob detection algorithm in combination with a neural network for fast and automated identification of plastics and non-plastics in a single fluorescence lifetime image. Therefore an artificial environmental matrix is prepared that contains soil, grass, spruce and HDPE (high density polyethylene) particles. Several FD-FLIM (frequency domain fluorescence lifetime imaging microscopy) images are taken, and the detection algorithm and the neural network are applied. We successfully demonstrated the suitability of the thresholding algorithm and the binary classification of the HDPE particles directly in the environmental matrix. KW - Fluorescence KW - Plastics KW - FD-FLIM KW - Neural networks KW - MLP KW - blob detection KW - Classification algorithms KW - Dogs KW - fluorescence lifetime KW - plastics identification KW - Soil KW - Thresholding (Imaging) Y1 - 2023 U6 - https://doi.org/10.1109/SAS58821.2023.10254171 SP - 1 EP - 5 ER - TY - CHAP A1 - Schwarz, Jonas A1 - Wohlschläger, Maximilian A1 - Leiter, Nina A1 - Auer, Veronika A1 - Risse, Michael A1 - Versen, Martin T1 - Frequency Domain Fluorescence Lifetime Imaging Microscopy (FD-FLIM) analysis of Quercus robur samples for origin differentiation purposes T2 - Optica Sensing Congress 2023 (AIS, FTS, HISE, Sensors, ES) N2 - Increasing demand for wood products requires methods to determine its harvest origin and ensure sustainable and legal sourcing. In 15 out of 21 cases, the origin of Quercus robur was differentiable in FD-FLIM studies. KW - Fluorescence lifetime imaging KW - Phase shift KW - Phase modulation KW - Laser sources KW - Bandpass filters KW - Frequency modulation Y1 - 2023 U6 - https://doi.org/10.1364/AIS.2023.JTu4A.10 ER - TY - JOUR A1 - Maniyattu, Georgekutty Jose A1 - Geegy, Eldho A1 - Wohlschläger, Maximilian A1 - Leiter, Nina A1 - Versen, Martin A1 - Laforsch, Christian T1 - Multilayer Perceptron Development to Identify Plastics Using Fluorescence Lifetime Imaging Microscopy JF - EDFA Technical Articles N2 - Existing plastic analysis techniques such as Fourier transform infrared spectroscopy and Raman spectroscopy are problematic because samples must be anhydrous and identification can be hindered by additives. This article describes a new approach that has been successfully demonstrated in which plastics can be classified by neural networks that are trained, validated, and tested by frequency domain fluorescence lifetime imaging microscopy measurements. Y1 - 2023 U6 - https://doi.org/10.31399/asm.edfa.2023-3.p031 VL - 25 IS - 3 SP - 31 EP - 37 ER - TY - CHAP A1 - Leiter, Nina A1 - Wohlschläger, Maximilian A1 - Dietlmeier, Maximilian A1 - Versen, Martin A1 - Löder, Martin A1 - Laforsch, Christian T1 - Comparative Analysis of Fluorescence Properties of Post-Consumer Wood Using FD-FLIM T2 - 2023 IEEE Sensors Applications Symposium (SAS) N2 - The FD-FLIM technique has a high potential for automated post-consumer wood sorting. A problem of analysing post-consumer wood fluorescence properties is the uncertainty of the post-consumer wood category as the origin of the samples are unknown. In this study, the fluorescence properties of actual post-consumer wood is compared with prepared wood samples. The post-consumer wood samples display slightly different fluorescence intensities and lifetimes due to environmental influences and a higher sample diversity. For improved training of evaluation algorithms for post-consumer wood sorting, the prepared sample set should be extended or the post-consumer wood should be additionally analysed in the laboratory. KW - fluorescence KW - Fluorescence KW - Neural networks KW - fluorescence properties KW - Moisture KW - post-consumer wood KW - Sensors KW - Training KW - Uncertainty KW - Visualization KW - waste wood Y1 - 2023 U6 - https://doi.org/10.1109/SAS58821.2023.10254052 SP - 1 EP - 6 ER - TY - CHAP A1 - Leiter, Nina A1 - Dietlmeier, Maximilian A1 - Wohlschläger, Maximilian A1 - Löder, Martin G.J. A1 - Versen, Martin A1 - Laforsch, Christian T1 - Development of a Neural Network for Automatic Classification of Post-Consumer Wood Using Rapid-FLIM T2 - 2023 IEEE Sensors Applications Symposium (SAS) N2 - The economic use of wood is a growing sector, not only because of the significant advantage of wood to retain CO 2 . It is crucial to increase the material recycling of wood in several lifecycles, but currently, there is no reliable post-consumer wood sorting technique in line. This research measures 365 post-consumer wood samples of classes A1-4 four times with the frequency domain fluorescence lifetime imaging microscopy (FD-FLIM) subset method rapid fluorescence lifetime imaging microscopy (Rapid-FLIM). The data is analyzed on their statistical features. Four neural networks based on Multilayer perceptron are then trained and tested with twelve statistical features extracted from the Rapid-FLIM images. The best model for this application contains the optimizer RMSprop, the activation function SELU and the loss function binary crossentropy. The best model of this structure could achieve a false positive ratio of 4.79 % over the ten folds. KW - fluorescence KW - Fluorescence KW - Microscopy KW - Neural networks KW - MLP KW - Current measurement KW - Feature extraction KW - Loss measurement KW - Post-consumer wood classification KW - Rapid-FLIM KW - Rapid-Fluorescence Lifetime Imaging Microscopy KW - Time measurement Y1 - 2023 U6 - https://doi.org/10.1109/SAS58821.2023.10254174 SP - 1 EP - 6 ER -