Physics guided neural networks for modelling of non-linear dynamics

Robinson, Haakon; Pawar, Suraj; Rasheed, Adil; San, Omer

doi:10.1016/j.neunet.2022.07.023

Cited by 30 publications

(14 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This creates a need to enhance the generalizability and data efficiency of data‐driven approaches for biological systems. As a result, researchers are actively exploring ways to combine physics‐based and neural network models to form a new field known as physics‐guided AI or physics‐guided ML (PGML) 118–122 . In this section, we orient researchers around key concepts of how quantum computing can be integrated into the study of the hierarchical complexity of living organisms and address the above‐mentioned problems.…”

Section: Scaling Behavior In Living Systems With Quantum Computingmentioning

confidence: 99%

Complexity of life sciences in quantum and AI era

Pyrkov,

Aliper,

Bezrukov

et al. 2024

WIREs Comput Mol Sci

View full text Add to dashboard Cite

Having made significant advancements in understanding living organisms at various levels such as genes, cells, molecules, tissues, and pathways, the field of life sciences is now shifting towards integrating these components into the bigger picture to understand their collective behavior. Such a shift of perspective requires a general conceptual framework for understanding complexity in life sciences which is currently elusive, a transition being facilitated by large‐scale data collection, unprecedented computational power, and new analytical tools. In recent years, life sciences have been revolutionized with AI methods, and quantum computing is touted to be the next most significant leap in technology. Here, we provide a theoretical framework to orient researchers around key concepts of how quantum computing can be integrated into the study of the hierarchical complexity of living organisms and discuss recent advances in quantum computing for life sciences.This article is categorized under: Data Science > Artificial Intelligence/Machine Learning Quantum Computing > Algorithms Structure and Mechanism > Computational Biochemistry and Biophysics

show abstract

Section: Scaling Behavior In Living Systems With Quantum Computingmentioning

confidence: 99%

Complexity of life sciences in quantum and AI era

Pyrkov,

Aliper,

Bezrukov

et al. 2024

WIREs Comput Mol Sci

View full text Add to dashboard Cite

show abstract

“…Here, the LIBS spectra are normalized with the following formula: ( ) is the minimum (maximum) original intensity among the pixels in the th spectrum. Following earlier work [22] in which LIBS was combined with machine learning to estimate the physical parameters of materials, the feature is selected according to a linear correlation defined by the following steps as described in [33]:…”

Section: Unified Multivariate Model and Its Performancementioning

confidence: 99%

“…In addition, hyperparameter optimization was conducted with cross-validation and grid-search parameter tuning. Details can be found in [33]. Specifically, the procedure involves: (1) defining a range and step size for the values of the model hyperparameters (including the number of features); (2) identifying a possible hyperparameter combination and evaluating the associated model performance via cross-validation; (3) systematically exploring all possible combinations of hyperparameters and assessing the corresponding model performance through cross-validation; (4) the hyperparameter combination yielding the optimal model performance is used as the model's optimal hyperparameters.…”

Section: Unified Multivariate Model and Its Performancementioning

confidence: 99%

Estimating the grain size of microgranular material using laser-induced breakdown spectroscopy combined with machine learning algorithms

ZHANG 张,

LI 李,

YANG 杨

et al. 2024

Plasma Sci. Technol.

View full text Add to dashboard Cite

Recent work validated a new method for estimating grain size of microgranular materials in the range of tens-to-hundreds micrometers using laser-induced breakdown spectroscopy (LIBS). In that situation, univariate analysis was performed and a piecewise model has to be constructed for achieving the estimation of the grain size within such a wide range. This is due to the fact that a complex dependence of plasma formation environment (i.e., the status of luminous plasma and therefore LIBS signal to be measured) on grain size occurs in the size range studied there. In the present work, we tentatively construct a unified calibration model suitable for LIBS-based estimation of those grain sizes. Specifically, two unified multivariate calibration models are constructed based on back-propagation neural network (BPNN) algorithms using the feature selection strategies with and without considering physical prior knowledge, respectively. By detailed analysis of the performances of the two multivariate models, it was found that, a unified calibration model can be constructed successfully based on BPNN algorithms for estimating the grain size in the range of tens-to-hundreds micrometers. It was also found that this model constructed with a physics-guided feature selection strategy has better prediction performances. This study has practical significance in developing the technology for material analysis using LIBS, especially in the case that LIBS signal exhibits a complex dependence on the material parameter to be estimated.

show abstract

“…Since their breakthrough, PINNs or similar approaches used in PINNs have been applied across almost all branches of engineering and science: applied mathematics [37][38][39][40][41][42] ; seismic response modeling 43 ; flow problems [44][45][46][47] ; thermal-related problems 48,49 ; thermochemical curing processes. 50 A new method of integrating domain knowledge or physics laws was tried by Robinson et al 51 To address benchmark problems, they utilized an approach that involved injecting physical information into an intermediate layer of a PINN during training. Despite the introduction of this approach, it is still unclear how the injection of physical laws into the different types of PINNs relates to the nature of their modeling.…”

Section: Introductionmentioning

confidence: 99%

Event‐triggered model reference adaptive control system design for SISO plants using meta‐learning‐based physics‐informed neural networks without labeled data and transfer learning

Duanyai,

Song,

Konghuayrob

et al. 2024

Adaptive Control & Signal

View full text Add to dashboard Cite

SummaryThis paper examines the controllability of a novel Lyapunov‐based model reference adaptive control (MRAC) system designed with a meta‐learning‐based physics‐informed neural network (MLPINN) for linear and nonlinear single‐input and single‐output (SISO) plants without labeled data (MLPINN‐MRAC system). It is devised with the benefits of several techniques: the integration of the identification process in a training mode into an online control mode (straightforward design); no labeled data generation for online identification by a physics‐informed neural network; the prevention of degradation in tracking performance by meta‐learning, as the system triggers a meta‐learning process only when an error threshold detects the deterioration (high efficiency and the reduction of computation cost); quick adaptation to new inputs and an updated control input for each sub‐time span by transfer learning. It is worth noting that the frequency of the meta‐learning event detection significantly affects the step response stability of the nonlinear plant. To achieve a better quality of the stabilization, more frequent event detection is necessary for both beginning and end intervals in control. Sixteen triggering events are enough to shape the acceptable step response of the nonlinear plant, and 44 triggering events achieve the plant's desired step response with its minor modeling error. While, as for the linear plant, a single triggering event is sufficient to attain its tolerable step response and modeling error, implying that intensive event detection is not critical in the identification. It is obvious that the MLPINN‐MRAC system functions well and is more beneficial and efficient for the nonlinear plant.

show abstract

Physics guided neural networks for modelling of non-linear dynamics

Cited by 30 publications

References 52 publications

Complexity of life sciences in quantum and AI era

Complexity of life sciences in quantum and AI era

Estimating the grain size of microgranular material using laser-induced breakdown spectroscopy combined with machine learning algorithms

Event‐triggered model reference adaptive control system design for SISO plants using meta‐learning‐based physics‐informed neural networks without labeled data and transfer learning

Contact Info

Product

Resources

About