Data-Driven Science and Engineering

Brunton, Steven L.; Kutz, J. Nathan

doi:10.1017/9781108380690

Cited by 739 publications

(273 citation statements)

References 403 publications

Supporting

Mentioning

270

Contrasting

Unclassified

Order By: Relevance

“…It identifies the relevant candidates by adopting a sparsity constraint. Two fundamental methods have been proposed: Sparse identification of nonlinear dynamics (SINDy) [20,25] and fast function extraction (FFX) [26]. Both methods were applied in several areas of physical modelling.…”

Section: Model Discovery Methodologymentioning

confidence: 99%

“…Following the idea of parsimonious models we constrain the search to models which optimally balance error and complexity and are not overfitting the data [25]. In principle, given a library a combinatoric study can be carried out, by performing an ordinary least-squares regression for each possible subset of candidates.…”

Section: Model Selection Using Sparsity-promoting Regressionmentioning

confidence: 99%

“…Hence, we follow [25,26] and engage sparsity-promoting regularisation of the underlying least-squares optimisation problem. The model-discovery procedure is divided into two parts: (i) model selection and (ii) model inference, see Figure 1.…”

Section: Model Selection Using Sparsity-promoting Regressionmentioning

confidence: 99%

“…Most commonly the optimal (λ, ρ) combination is found based on a strategy to avoid overfitting of the resulting models, e.g. using crossvalidation [25], for which the data is split into a training and a test set. While the optimisation problem given a grid (λ, ρ) is solved on the former, only the model with the best performance evaluated on the latter survives.…”

Section: Model Selection Using Sparsity-promoting Regressionmentioning

confidence: 99%

“…In [25,28,29] this was done using ordinary least-squares regression for problems in the domains of dynamical systems and biological networks. As mentioned above, the ability of the CFD solver, in which the models will be implemented, to produce a converged solution is sensitive to large coefficients, which has been reported in [11,12,22].…”

Section: Model Inference For Cfdmentioning

confidence: 99%

See 4 more Smart Citations

Discovery of Algebraic Reynolds-Stress Models Using Sparse Symbolic Regression

Schmelzer¹,

Dwight²,

Cinnella³

2019

Flow Turbulence Combust

153

102

View full text Add to dashboard Cite

A novel deterministic symbolic regression method SpaRTA is introduced to infer algebraic stress models for the closure of RANS equations directly from high-fidelity LES or DNS data. The models are written as tensor polynomials and are built from a library of candidate functions. The machine-learning method is based on elastic net regularisation which promotes sparsity of the inferred models. By being data-driven the method relaxes assumptions commonly made in the process of model development. Model-discovery and cross-validation is performed for three cases of separating flows, i.e. periodic hills (Re=10595), converging-diverging channel (Re=12600) and curved backward-facing step (Re=13700). The predictions of the discovered models are significantly improved over the k-ω SST also for a true prediction of the flow over periodic hills at Re=37000. This study shows a systematic assessment of SpaRTA for rapid machine-learning of robust corrections for standard RANS turbulence models.

show abstract

Section: Model Discovery Methodologymentioning

confidence: 99%

Section: Model Selection Using Sparsity-promoting Regressionmentioning

confidence: 99%

Section: Model Selection Using Sparsity-promoting Regressionmentioning

confidence: 99%

Section: Model Selection Using Sparsity-promoting Regressionmentioning

confidence: 99%

Section: Model Inference For Cfdmentioning

confidence: 99%

See 3 more Smart Citations

Discovery of Algebraic Reynolds-Stress Models Using Sparse Symbolic Regression

Schmelzer¹,

Dwight²,

Cinnella³

2019

Flow Turbulence Combust

153

102

View full text Add to dashboard Cite

show abstract

Optimal controller tuning for P&O maximum power point tracking of PV systems using genetic and cuckoo search algorithms

Ahmed

Rahman

Alajmi

2020

Int Trans Electr Energ Syst

View full text Add to dashboard Cite

Background: Tracking the maximum power point (MPPT) of photovoltaic (PV) systems is essential to increase the utilization efficiency. One of the most widely used algorithms is perturb and observe (P&O) technique.Methods: In this paper, tuning algorithms based on genetic algorithm (GA) and cuckoo search (CS) are proposed to provide optimal controller parameters for P&O MPPT. This optimal tuning is crucial in P&O MPPT applications of PV systems where it is difficult to tune the controllers using conventional tuning techniques to attain good performance because of the wide variations in the system parameters and its nonlinearity. The proposed tuning algorithms are used to overcome the drawbacks of the classical P&O MPPT under rapidly changing atmospheric conditions such as oscillation around the maximum power point (MPP) and low convergence speed.Results: The robustness of the algorithms is confirmed against fast varying insolation proving its ability to track the MPP in case of random and fast changing atmospheric conditions. An experimental set up of a 1 kW PV system is carried out and implemented using dSPACE DS1103 controller board and E4360 Keysight modular solar array emulator to validate the performance of the proposed optimal tuning algorithms. The measured results are in a close agreement with the simulation and the analytical system. Conclusion: Experimental and simulation results demonstrate a superior improvement in the transient and steady-state performances and in the maximum power tracking efficiency. The steady-state and the transient efficiencies of the presented work are 99.95% and 94.02% which are almost 17.59% and 10.61%, respectively superior to the classical P&O one and are 14.66% and

show abstract