TY - JOUR A1 - Weber, Felix A1 - Huber, Peter A1 - Borchsenius, Fredrik A1 - Braun, Christian T1 - Performance of TMDI for Tall Building Damping JF - Actuators N2 - This study investigates the vibration reduction of tall wind-excited buildings using a tuned mass damper (TMD) with an inerter (TMDI). The performance of the TMDI is computed as a function of the floor to which the inerter is grounded as this parameter strongly influences the vibration reduction of the building and for the case when the inerter is grounded to the earth whereby the absolute acceleration of the corresponding inerter terminal is zero. Simulations are made for broadband and harmonic excitations of the first three bending modes, and the conventional TMD is used as a benchmark. It is found that the inerter performs best when grounded to the earth because, then, the inerter force is in proportion to the absolute acceleration of only the pendulum mass, but not to the relative acceleration of the two inerter terminals, which is demonstrated by the mass matrix. However, if the inerter is grounded to a floor below the pendulum mass, the TMDI only outperforms the TMD if the inerter is grounded to a floor within approximately the first third of the building’s height. For the most realistic case, where the inerter is grounded to a floor in the vicinity of the pendulum mass, the TMDI performs far worse than the classical TMD. KW - damping KW - inerter KW - simulation KW - tall building KW - TMD KW - TMDI KW - wind excitation Y1 - 2020 U6 - https://doi.org/10.3390/act9040139 VL - 9 IS - 4 SP - 1 EP - 13 PB - MPDI ER - TY - JOUR A1 - Weber, Felix A1 - Borchsenius, Fredrik A1 - Distl, Johann A1 - Braun, Christian T1 - Performance of Numerically Optimized Tuned Mass Damper with Inerter (TMDI) JF - Applied Sciences N2 - In recent years, the Tuned Mass Damper with inerter (TMDI) has received significant attention. The inerter is defined to exert a force that is in proportion to the relative acceleration of the two inerter terminals. Here, two TMDI topologies are investigated. The conventional topology is given by the inerter being in parallel to the spring and viscous damper of the TMDI. The other topology is the serial arrangement of spring, inerter and viscous damper being in parallel to the stiffness of the mass spring oscillator of the TMDI. While the first topology intends to increase the inertial force of the TMDI, the second topology aims at producing an additional degree of freedom. The considered TMDI concepts are simulated for harmonic and random excitations, with parameters set according to those described in the literature and with numerically optimized parameters which minimize the primary structure displacement response. The classical TMD is used as a benchmark. The findings are twofold. The conventional TMDI with typical inertance ratio of 1% and the very small value of 0.02% performs significantly worse than the classical TMD with the same mass ratio. In contrast, the TMDI with an additional degree of freedom can improve the mitigation of the primary structure if the inertance ratio is set very small and if the TMDI parameters are numerically optimized. KW - damping KW - inerter KW - optimization KW - TMD KW - TMDI KW - vibration Y1 - 2022 U6 - https://doi.org/10.3390/app12126204 VL - 12 IS - 12 SP - 1 EP - 15 PB - MDPI CY - Basel ER - TY - JOUR A1 - Graßmann, Felix A1 - Mengelkamp, Judith A1 - Brandl, Caroline A1 - Harsch, Sebastian A1 - Zimmermann, Martina E. A1 - Linkohr, Birgit A1 - Peters, Annette A1 - Heid, Iris M. A1 - Palm, Christoph A1 - Weber, Bernhard H. F. T1 - A Deep Learning Algorithm for Prediction of Age-Related Eye Disease Study Severity Scale for Age-Related Macular Degeneration from Color Fundus Photography JF - Ophtalmology N2 - Purpose Age-related macular degeneration (AMD) is a common threat to vision. While classification of disease stages is critical to understanding disease risk and progression, several systems based on color fundus photographs are known. Most of these require in-depth and time-consuming analysis of fundus images. Herein, we present an automated computer-based classification algorithm. Design Algorithm development for AMD classification based on a large collection of color fundus images. Validation is performed on a cross-sectional, population-based study. Participants. We included 120 656 manually graded color fundus images from 3654 Age-Related Eye Disease Study (AREDS) participants. AREDS participants were >55 years of age, and non-AMD sight-threatening diseases were excluded at recruitment. In addition, performance of our algorithm was evaluated in 5555 fundus images from the population-based Kooperative Gesundheitsforschung in der Region Augsburg (KORA; Cooperative Health Research in the Region of Augsburg) study. Methods. We defined 13 classes (9 AREDS steps, 3 late AMD stages, and 1 for ungradable images) and trained several convolution deep learning architectures. An ensemble of network architectures improved prediction accuracy. An independent dataset was used to evaluate the performance of our algorithm in a population-based study. Main Outcome Measures. κ Statistics and accuracy to evaluate the concordance between predicted and expert human grader classification. Results. A network ensemble of 6 different neural net architectures predicted the 13 classes in the AREDS test set with a quadratic weighted κ of 92% (95% confidence interval, 89%–92%) and an overall accuracy of 63.3%. In the independent KORA dataset, images wrongly classified as AMD were mainly the result of a macular reflex observed in young individuals. By restricting the KORA analysis to individuals >55 years of age and prior exclusion of other retinopathies, the weighted and unweighted κ increased to 50% and 63%, respectively. Importantly, the algorithm detected 84.2% of all fundus images with definite signs of early or late AMD. Overall, 94.3% of healthy fundus images were classified correctly. Conclusions Our deep learning algoritm revealed a weighted κ outperforming human graders in the AREDS study and is suitable to classify AMD fundus images in other datasets using individuals >55 years of age. KW - Senile Makuladegeneration KW - Krankheitsverlauf KW - Mustererkennung KW - Maschinelles Lernen Y1 - 2018 U6 - https://doi.org/10.1016/j.ophtha.2018.02.037 N1 - Corresponding authors: Bernhard H. F. Weber, University of Regensburg, and Christoph Palm VL - 125 IS - 9 SP - 1410 EP - 1420 PB - Elsevier ER -