Cnn-Dp: Composite Neural Network with Differential Propagation For Impulsive Nonlinear Dynamics

Lee, Hyeonbeen; Han, Seongji; Choi, Hee-Sun; Kim, Jin-Gyun

doi:10.2139/ssrn.4353533

Cited by 2 publications

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The development of efficient training algorithms and improved performance of hardware have made it is possible to predict more complex dynamical systems. The ability of neural networks to describe nonlinearity is exploited in a variety of ways, including techniques based on composite neural networks that can effectively improve model accuracy by utilizing low and high-fidelity data [16] and differential propagation composite neural networks that can predict impulsive responses with high accuracy [17]. A data-driven DNN metamodeling technique for general-purpose MBD systems has been presented [18], and DNN techniques that effectively describe high nonlinearity and achieve real-time predictions over a large number of degrees of freedom have been studied for MBD systems with flexible bodies [19].…”

mentioning

confidence: 99%

SID-Net: Machine learning based system identificationframework for rigid and flexible multibody dynamics

Jang,

Han,

Kim

et al. 2023

Preprint

View full text Add to dashboard Cite

Modeling and simulation of dynamic systems is widely used in mechanical system design and control. System identification (SI), a process of correlation using experimental or target data, is essential for the reliability of implemented numerical models. To actualize the process, it is crucial to understand the relationship between numerous modeling parameters that affect the system responses. Modeling and simulating nonlinear systems such as multibody dynamics, is particularly difficult owing to their characteristic enormous assumptions and approximations; furthermore, the computational demands are also crucial during the iterative SI process. Therefore, we propose an effective SI framework for rigid and flexible multibody dynamics using artificial neural networks. The framework consists of two phases: a system metamodel and a SI network (SIN). Through supervised learning, a metamodel that can provide nonlinear time transient responses based on the input system design parameters in real time was generated. The loss minimization process for initializing a new neural network that, utilizes a SIN to find the input design parameters that generate the desired target output was also introduced. A new loss function was defined for minimizing the difference between the output of the trained system metamodel, following which the target output was defined. The performance of the proposed framework was evaluated in well-designed numerical and experimental examples.

show abstract

mentioning

confidence: 99%