Refine
Year of publication
Document type
- Article (peer-reviewed) (1698) (remove)
Language
- English (1445)
- German (250)
- Multiple languages (3)
Keywords
- Electrical impedance tomography (74)
- Mechanical ventilation (31)
- Exercise (24)
- Grinding (17)
- Surface roughness (17)
- Bone mineral density (16)
- Parameter identification (16)
- Acute respiratory distress syndrome (15)
- Machine learning (13)
- Körperliches Training (12)
Pulmonary response prediction through personalized basis functions in a virtual patient model
(2024)
Introduction: The present study investigated the role of training intensity in the dose–response relationship between endurance training and cardiorespiratory fitness (CRF). The hypothesis was that beginners would benefit from an increase in training intensity after an initial training phase, even if the energy expenditure was not altered. For this purpose, 26 weeks of continuous moderate training (control group, CON) was compared to training with gradually increasing intensity (intervention group, INC) but constant energy expenditure.
Methods: Thirty-one healthy, untrained subjects (13 men, 18 women; 46±8 years; body mass index 25.4 ± 3.3 kg m−2; maximum oxygen uptake, VO2max −1 −1 34 ± 4 ml min kg ) trained for 10 weeks with moderate intensity [3 days/week for 50 min/session at 55% heart rate reserve (HRreserve)] before allocation to one of two groups. A minimization technique was used to ensure homogeneous groups. While group CON continued with moderate intensity for 16 weeks, the INC group trained at 70% HRreserve for 8weeks and thereafter participated in a 4 × 4 training program (high-intensity interval training, HIIT) for 8 weeks. Constant energy expenditure was ensured by indirect calorimetry and corresponding adjustment of the training volume. Treadmill tests were performed at baseline and after 10, 18, and 26 weeks.
Results: The INC group showed improved VO2max (3.4 ± 2.7 ml kg−1 min−1) to a significantly greater degree than the CON group (0.4 ± 2.9 ml kg−1 min−1) (P = 0.020). In addition, the INC group exhibited improved Vmax (1.7 ± 0.7 km h−1) to a significantly greater degree than the CON group (1.0 ± 0.5 km h−1) (P = 0.001). The reduction of resting HR was significantly larger in the INC group (7±4bpm) than in the CON group (2±6bpm) (P=0.001). The mean heart rate in the submaximal exercise test was reduced significantly in the CON group (5±6bpm; P=0.007) and in the INC group (8±7bpm; P=0.001), without a significant interaction between group and time point.
Data processed in context is more meaningful, easier to understand and has higher information content, hence it derives its semantic meaning from the surrounding context. Even in the field of acoustic signal processing. In this work, a Deep Learning based approach using Ensemble Neural Networks to integrate context into a learning system is presented. For this purpose, different use cases are considered and the method is demonstrated using acoustic signal processing of machine sound data for valves, pumps and slide rails. Mel-spectrograms are used to train convolutional neural networks in order to analyse acoustic data using image processing techniques.
On Consistency Viability and Admissibility in Constrained Ensemble and Hierarchical Control Systems
(2023)
Several control architectures, such as decentralized, distributed, and hierarchical control, have been elaborated over the past decades for controlling systems composed of a set of subsystems. However, computational complexity and constraint satisfaction are still challenging tasks. We present an approach to control an ensemble of similar heterogeneous systems with input and state constraints via an identical control input. This control input is globally admissible and computed based on an aggregated system that reflects the overall behavior of the ensemble. To limit the computational complexity of the control task, the aggregated system is designed such that its dimension is independent of the number of subsystems. To guarantee viability, i.e., state constraint satisfaction for all times, appropriate consistency conditions are derived based on invariant set theory. The presented approach is illustrated with a numerical example.
In this paper, we derive set constraints for a reduced order model and augment them into a model predictive control (MPC) scheme to ensure safe operation of the large-scale ensemble system. For the control feedback, only the aggregated information of the whole system is required. For the constraint satisfaction, we consider an adaptive tube formulation to characterize the deviation between the reduced order model and the ensemble system. Employing the robust control invariant set, we ensure recursive feasibility and initial feasibility under an easily verifiable condition.
XAutoML : A Visual Analytics Tool for Understanding and Validating Automated Machine Learning
(2023)
Increasing Resilience of Production Systems by Dynamic Context Modelling and Process Adaption
(2023)