PUMA publications for /user/ifsw/bayesianoptimizationMon Apr 20 16:13:32 CEST 2026Advanced laser processing and its optimization with machine learning2023myown Welding Laser BayesianOptimization GaussianProcesses MachineLearning Ablation Optimization The flexibility of new laser sources and process-monitoring enables new possibilities in laser-based production technology, for instance the combination of different laser processes with many adjustable parameters. The fusion of domain knowledge and probabilistic models in the form of hybrid models allows an efficient optimization of these processes with machine learning. This can be a key technology to realize self-learning laser-based universal machines in the future. The article discusses some examples where algorithm-based optimization, partly supported by hybrid models, can already greatly reduce the effort required to find suitable process parameters.Mon Apr 20 16:09:36 CEST 2026Advanced laser processing and its optimization with machine learningLaser Applications in Microelectronic and Optoelectronic Manufacturing (LAMOM) XXVIII2023myown Welding Laser BayesianOptimization GaussianProcesses MachineLearning Ablation Optimization The flexibility of new laser sources and process-monitoring enables new possibilities in laser-based production technology, for instance the combination of different laser processes with many adjustable parameters. The fusion of domain knowledge and probabilistic models in the form of hybrid models allows an efficient optimization of these processes with machine learning. This can be a key technology to realize self-learning laser-based universal machines in the future. The article discusses some examples where algorithm-based optimization, partly supported by hybrid models, can already greatly reduce the effort required to find suitable process parameters.Mon Apr 20 10:38:19 CEST 2026Light: Advanced Manufacturing332Laser material processing optimization using bayesian optimization: a generic tool52024myown Welding peer Laser DesignOfExperiments BayesianOptimization Ablation Cutting Fri Apr 10 15:59:18 CEST 2026Journal of Intelligent ManufacturingfebEfficient parameter selection in laser welding via Bayesian optimization2026myown Welding peer Aluminum BayesianOptimization X-ray Optimization 27Identifying suitable process parameters is essential for developing effective laser welding processes. Traditionally, this involves extensive experimentation and relies heavily on expert knowledge. Bayesian optimization can be used to minimize both the experimental effort and the need for expert information. This study demonstrates the merit of Bayesian optimization for laser welding and explains the methodology for implementing the optimization technique. A strategy for selecting evaluation methods and the design of a suitable cost function to meet specific quality criteria is proposed. For the experimental demonstration, butt joint laser welding of AA1050 aluminum alloy was performed with options to adapt the laser power, welding speed, focus position, and the intensity distribution of the laser beam by changing the power distribution in a multi-core fiber. The success of the optimization was validated by finding several parameter sets producing welds that met the defined quality levels. Furthermore, the properties of the underlying surrogate model of the Bayesian optimizer generated further information that helped to improve the welding process.Fri Dec 05 10:30:29 CET 2025Enhancing the laser weld seam quality with dynamic beam shaping and Bayesian optimization2025myown BayesianOptimization laserwelding Mon Jun 30 12:57:18 CEST 2025Forum Bildverarbeitung 2024 = Image Pocessing Forum 2024sep1–12Automated image-based parameter optimization for single-pulse laser drilling2024Drilling myown peer BayesianOptimization Mon Sep 23 09:27:54 CEST 2024Improving the weld seam quality in laser welding processes by means of Bayesian optimization2024myown Welding Laser BayesianOptimization X-ray The determination of appropriate process parameters is crucial for the development of laser welding processes. This usually requires extensive and time-consuming experimentation combined with expert knowledge. To reduce the number of experiments required to determine appropriate process parameters, Bayesian optimization was used in this work. Bead on plate laser welding of AA5754 samples was performed while optimizing the laser power, the welding speed, the focus position and the power distribution in the core-ring fiber laser system with the objective of achieving welds with a specific weld depth and low number of defects at high welding speeds. The welds were evaluated using X-ray imaging and height measurements. A cost function was developed to quantify the overall weld quality based on the weld properties. It is demonstrated that the Bayesian optimizer can determine appropriate process parameters for the given objective, based on a cost function, within a comparatively small number of 29 experiments.Mon Sep 23 09:21:21 CEST 2024Procedia CIRP772–775Improving the weld seam quality in laser welding processes by means of Bayesian optimization1242024myown Welding peer Laser BayesianOptimization X-ray The determination of appropriate process parameters is crucial for the development of laser welding processes. This usually requires extensive and time-consuming experimentation combined with expert knowledge. To reduce the number of experiments required to determine appropriate process parameters, Bayesian optimization was used in this work. Bead on plate laser welding of AA5754 samples was performed while optimizing the laser power, the welding speed, the focus position and the power distribution in the core-ring fiber laser system with the objective of achieving welds with a specific weld depth and low number of defects at high welding speeds. The welds were evaluated using X-ray imaging and height measurements. A cost function was developed to quantify the overall weld quality based on the weld properties. It is demonstrated that the Bayesian optimizer can determine appropriate process parameters for the given objective, based on a cost function, within a comparatively small number of 29 experiments.