| 1. |
|
|
| 2. |
- Persson, Anders, 1982-
(författare)
-
Energy sources and energy harvesting for high-power pneumatic microsystems
-
Annan publikation (övrigt vetenskapligt/konstnärligt)abstract
- Pneumatic microsystems control fluid pressure for various tasks like soft robotics and tactile displays. This study shifts focus to system integration, often overlooked in research when aspects like total size, mass, and power consumption are neglected. The study explores energy source integration efficiency and proposes solutions through both a literature review that identifies key parameters for potential energy sources and microsystem applications, and experimental investigations where mechanical transmission is used to adapt energy sources to application-based requirements. Comparative analysis categorizes energy sources and assesses their performance and their relation to fundamental limits like the average power exerted by the human body. Mechanical transmission's role in optimizing energy source-application compatibility is found promising, at least for moderate optimization. To conclude, bypassing energy transduction enhances energy harvesting efficiency. Mechanical-to-pneumatic systems are efficient, and mechanical transmission optimizes source-application matching. Human body energy remains a critical constraint. For high energy demands, portable batteries become essential.
|
|
| 3. |
|
|
| 4. |
|
|
| 5. |
|
|
| 6. |
|
|
| 7. |
|
|
| 8. |
- Zendejas Medina, Léon, et al.
(författare)
-
Enhancing the mechanical and electrochemical properties of sputtered CoCrFeMnNi thin films by carbon addition
-
Annan publikation (övrigt vetenskapligt/konstnärligt)abstract
- High entropy alloy (HEA) thin films of CrMnFeCoNi are promising materials for polymer electrolyte membrane fuel cells, due to their high fracture resistance and passivation against corrosion. The addition of carbon to HEA films have shown to improve corrosion resistance, hardness and ductility. HEA films with 0%, 6% and 11% amounts of carbon were deposited on Si[100] and polyimide substrates by magnetron co-sputtering using a graphite target and a sintered compound target for this study. Characterization using XRD and TEM showed that the samples were crystalline with a mixture of ccp and α-Mn at 0 at-% C, and underwent amorphization from 6 at-% C. No evidence of either free carbon or a crystalline carbide phase was found, and the films displayed homogenous composition. Nanoindentation measurements on coated Si [100] substrate showed that the hardness of the thin films changed with increasing carbon content. The tensile fragmentation test method was shown to be useful to characterize the strength distribution and fracture energy at crack onset for the present materials. This method revealed fewer cracks at all levels of strain for the carbon containing films compared to the carbon free ones. Consequentially, there was a trend of higher fracture energies by addition of carbon because of a shifting towards of crack onset strain. Electrochemical studies showed that the carbon had a beneficial effect on the passive layer, lowering the corrosion current and the passive current during with increasing carbon.
|
|
| 9. |
- Anubhab, Ghosh, et al.
(författare)
-
DeepBayes -- an estimator for parameter estimation in stochastic nonlinear dynamical models
-
Annan publikation (övrigt vetenskapligt/konstnärligt)abstract
- Stochastic nonlinear dynamical systems are ubiquitous in modern, real-world applications. Yet, estimating the unknown parameters of stochastic, nonlinear dynamical models remains a challenging problem. The majority of existing methods employ maximum likelihood or Bayesian estimation. However, these methods suffer from some limitations, most notably the substantial computational time for inference coupled with limited flexibility in application. In this work, we propose DeepBayes estimators that leverage the power of deep recurrent neural networks in learning an estimator. The method consists of first training a recurrent neural network to minimize the mean-squared estimation error over a set of synthetically generated data using models drawn from the model set of interest. The a priori trained estimator can then be used directly for inference by evaluating the network with the estimation data. The deep recurrent neural network architectures can be trained offline and ensure significant time savings during inference. We experiment with two popular recurrent neural networks -- long short term memory network (LSTM) and gated recurrent unit (GRU). We demonstrate the applicability of our proposed method on different example models and perform detailed comparisons with state-of-the-art approaches. We also provide a study on a real-world nonlinear benchmark problem. The experimental evaluations show that the proposed approach is asymptotically as good as the Bayes estimator.
|
|
| 10. |
|
|
