Neural Network Augmented Physics Models for Systems With Partially Unknown Dynamics: Application to Slider–Crank Mechanism

Groote, Wannes De; Kikken, Edward; Hostens, Erik; Hoecke, Sofie Van; Crevecoeur, Guillaume

doi:10.1109/tmech.2021.3058536

Cited by 27 publications

(16 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As has been discussed earlier, by trying to build the interactions between the components of the network informed by physical interactions of the system, the computational complexity of the parameter space reduces drastically. On the other hand, by constraining a data‐driven model appropriately, 89 the structure of the converged network provides insights into the physical interactions of the system. Current research, in physics and engineering, trying to embed physical constraints into the model by altering the architecture of the neural networks has proven to be a better alternative than penalizing using loss functions 13 .…”

Section: Future Directionsmentioning

confidence: 99%

See 1 more Smart Citation

Integration of machine learning and first principles models

et al. 2022

View full text Add to dashboard Cite

Model building and parameter estimation are traditional concepts widely used in chemical, biological, metallurgical, and manufacturing industries. Early modeling methodologies focused on mathematically capturing the process knowledge and domain expertise of the modeler. The models thus developed are termed first principles models (or white‐box models). Over time, computational power became cheaper, and massive amounts of data became available for modeling. This led to the development of cutting edge machine learning models (black‐box models) and artificial intelligence (AI) techniques. Hybrid models (gray‐box models) are a combination of first principles and machine learning models. The development of hybrid models has captured the attention of researchers as this combines the best of both modeling paradigms. Recent attention to this field stems from the interest in explainable AI (XAI), a critical requirement as AI systems become more pervasive. This work aims at identifying and categorizing various hybrid models available in the literature that integrate machine‐learning models with different forms of domain knowledge. Benefits such as enhanced predictive power, extrapolation capabilities, and other advantages of combining the two approaches are summarized. The goal of this article is to consolidate the published corpus in the area of hybrid modeling and develop a comprehensive framework to understand the various techniques presented. This framework can further be used as the foundation to explore rational associations between several models.

show abstract

Section: Future Directionsmentioning

confidence: 99%

“…In an example of modeling a mechatronic system, Groote et al 89 consider modeling a slider‐crank setup. Mechatronic systems generally tend to have partial information about the model as some interactions are unknown.…”

Section: Literature Surveymentioning

confidence: 99%

Integration of machine learning and first principles models

et al. 2022

View full text Add to dashboard Cite

show abstract

“…with: This single-step prediction can be easily coupled with standard machine learning optimizations such as stochastic gradient descent [13], which iterates the optimization on a subset, called mini-batch, of the data in order to speed up the optimization and avoid to compute the gradient with respect to all the samples. In fact, the training algorithm can be fed with mini-batches of (groups of) three consecutive reference points x i−1 n , x i n , x i+1 n ; while common recurrent neural networks may require to be fed with all the n t samples of the time evolution, leading to slow convergence, possible vanishing/exploding gradients [13] or the necessity to reshape the network architecture to perform the time simulation [38].…”

Section: Recurrent Autoencoder For Minimal Coordinates Mappingmentioning

confidence: 99%

Deep learning for model order reduction of multibody systems to minimal coordinates

Angeli

Desmet

Naets

2021

Computer Methods in Applied Mechanics and Engineering

View full text Add to dashboard Cite

show abstract

“…Alternatively, the influence of the neural networks can be attenuated by using them as mappings that compensate for prediction discrepancies of simplified physicsbased models [23], [24]. Furthermore, neural networks have been used to accommodate specific unknown interactions in incomplete yet accurate physics models [25], [26].…”

Section: Introductionmentioning

confidence: 99%

Physics-Based Neural Network Models for Prediction of Cam-Follower Dynamics Beyond Nominal Operations

Groote

Hoecke

Crevecoeur

2022

IEEE/ASME Trans. Mechatron.

Self Cite

View full text Add to dashboard Cite

Cam-follower mechanisms are key in various mechatronic applications to convert rotary to linear reciprocating motions. The dynamic behavior of these systems relies on the design parameters such as the cam shape and follower mass. It appears that for some combinations of system parameters, continuous contact between the cam and follower cannot be assured, leading to harmful periodic impacts. This research presents a data-driven approach to predict the influence of parameter settings on the system dynamics by learning from a limited data set of nominal operating conditions. More specifically, we present a hybrid model architecture encompassing an ordinary differential equation, consisting of a close interconnection of neural and physics-based network layers. Due to an increased generalization established by the physical laws, these physicsbased neural network models exhibit enhanced extrapolation capabilities compared to their black-box counterparts. Consequently, the presented models can accurately simulate the system behavior for parameter settings far beyond the nominal values included in the training data. This way, starting from a limited set of nominal time-series data, we could accurately estimate the set of critical system parameters that lead to hazardous jump phenomena in cam-follower systems.

show abstract

Neural Network Augmented Physics Models for Systems With Partially Unknown Dynamics: Application to Slider–Crank Mechanism

Cited by 27 publications

References 28 publications

Integration of machine learning and first principles models

Integration of machine learning and first principles models

Deep learning for model order reduction of multibody systems to minimal coordinates

Physics-Based Neural Network Models for Prediction of Cam-Follower Dynamics Beyond Nominal Operations

Contact Info

Product

Resources

About