Machine learning for accelerating macroscopic parameters prediction for poroelasticity problem in stochastic media

Vasilyeva, Maria; Tyrylgin, Aleksei

doi:10.1016/j.camwa.2020.09.024

Cited by 23 publications

(8 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is partly the result of a suitable choice of features, the novel rank-adaptive training strategy and the continuity induced by using a mesh transformation instead of automatic re-meshing. Compared to other works such as [61,62], where standard generic methods like artificial neural networks and polynomial chaos are used in a similar context, we gain several orders of accuracy.…”

Section: Discussionmentioning

confidence: 87%

Numerical upscaling of parametric microstructures in a possibilistic uncertainty framework with tensor trains

et al. 2022

View full text Add to dashboard Cite

A fuzzy arithmetic framework for the efficient possibilistic propagation of shape uncertainties based on a novel fuzzy edge detection method is introduced. The shape uncertainties stem from a blurred image that encodes the distribution of two phases in a composite material. The proposed framework employs computational homogenisation to upscale the shape uncertainty to a effective material with fuzzy material properties. For this, many samples of a linear elasticity problem have to be computed, which is significantly sped up by a highly accurate low-rank tensor surrogate. To ensure the continuity of the underlying mapping from shape parametrisation to the upscaled material behaviour, a diffeomorphism is constructed by generating an appropriate family of meshes via transformation of a reference mesh. The shape uncertainty is then propagated to measure the distance of the upscaled material to the isotropic and orthotropic material class. Finally, the fuzzy effective material is used to compute bounds for the average displacement of a non-homogenized material with uncertain star-shaped inclusion shapes.

show abstract

Section: Discussionmentioning

confidence: 87%

Numerical upscaling of parametric microstructures in a possibilistic uncertainty framework with tensor trains

et al. 2022

View full text Add to dashboard Cite

show abstract

“…The field of Machine Learning (ML) is a wide area in different topics and research works, e.g., prediction problems. Vasileva et al [5], proposed the fast calculation for the parameters of macroscope based on deep neural network construction. The presented system is to find the relationship between stochastic and macroscopic parameters.…”

Section: Related Workmentioning

confidence: 99%

A Framework of Vehicular Security and Demand Service Prediction Based on Data Analysis Integrated with Blockchain Approach

Shahbazi

Byun

2021

Sensors

View full text Add to dashboard Cite

The prediction of taxi demand service has become a recently attractive area of research along with large-scale and potential applications in the intelligent transportation system. The demand process is divided into two main parts: Picking-up and dropping-off demand based on passenger habit. Taxi demand prediction is a great concept for drivers and passengers, and is designed platforms for ride-hailing and municipal managers. The majority of research has focused on forecasting the pick-up part of demand service and specifying the interconnection of spatial and temporal correlations. In this study, the main focus is to overcome the access point of non-registered users for having fake transactions using taxi services and predicting taxi demand pick-up and drop-off information. The integration of machine learning techniques and blockchain framework is considered a possible solution for this problem. The blockchain technique was selected as an effective technique for protecting and controlling the real-time system. Historical data analysis was processed by extracting the three higher related sections for the intervening time, namely closeness and trend. Next, the pick-up and drop-off taxi prediction task was processed based on constructing the components of multi-task learning and spatiotemporal feature extraction. The combination of feature embedding performance and Long Short-Term Memory (LSTM) obtain the pick-up and drop-off correlation by fusing the historical data spatiotemporal features. Finally, the taxi demand pick-up and drop-off prediction were processed based on the combination of the external factors. The experimental result is based on a real dataset in Jeju Island, South Korea, to show the proposed system’s efficacy and performance compared with other state-of-art models.

show abstract

“…ML methods are used in battery research to understand issues such as degradation and safety. They were used to predict the state of charge and state of health of a battery using electro-impedance spectroscopy, , internal short circuit detection, optimal safety under abuse testing, , capacity fade, and remaining useful life predictions − and effective porous media properties. − …”

Section: Introductionmentioning

confidence: 99%

Mesoscale Machine Learning Analytics for Electrode Property Estimation

et al. 2022

View full text Add to dashboard Cite

The development of next-generation batteries with high areal and volumetric energy density requires the use of high active material mass loading electrodes. This typically reduces the power density, but the push for rapid charging has propelled innovation in microstructure design for improved transport and electrochemical conversion efficiency. This requires accurate effective electrode property estimation, such as tortuosity, electronic conductivity, and interfacial area. Obtaining this information solely from experiments and 3D mesoscale simulations is time-consuming while empirical relations are limited to simplified microstructure geometry. In this work, we propose an alternate route for rapid characterization of electrode microstructural effective properties using machine learning (ML). Using the Li-ion battery graphite anode electrode as an exemplar system, we generate a comprehensive data set of ∼17 000 electrode microstructures. These consist of various shapes, sizes, orientations, and chemical compositions, and characterize their effective properties using 3D mesoscale simulations. A low dimensional representation of each microstructure is achieved by calculating a set of comprehensive physical descriptors and eliminating redundant features. The mesoscale ML analytics based on porous electrode microstructural characteristics achieves prediction accuracy of more than 90% for effective property estimation.

show abstract

Machine learning for accelerating macroscopic parameters prediction for poroelasticity problem in stochastic media

Cited by 23 publications

References 25 publications

Numerical upscaling of parametric microstructures in a possibilistic uncertainty framework with tensor trains

Numerical upscaling of parametric microstructures in a possibilistic uncertainty framework with tensor trains

A Framework of Vehicular Security and Demand Service Prediction Based on Data Analysis Integrated with Blockchain Approach

Mesoscale Machine Learning Analytics for Electrode Property Estimation

Contact Info

Product

Resources

About