On the adaptation of recurrent neural networks for system identification
From MaRDI portal
Publication:6109982
Abstract: This paper presents a transfer learning approach which enables fast and efficient adaptation of Recurrent Neural Network (RNN) models of dynamical systems. A nominal RNN model is first identified using available measurements. The system dynamics are then assumed to change, leading to an unacceptable degradation of the nominal model performance on the perturbed system. To cope with the mismatch, the model is augmented with an additive correction term trained on fresh data from the new dynamic regime. The correction term is learned through a Jacobian Feature Regression (JFR) method defined in terms of the features spanned by the model's Jacobian with respect to its nominal parameters. A non-parametric view of the approach is also proposed, which extends recent work on Gaussian Process (GP) with Neural Tangent Kernel (NTK-GP) to the RNN case (RNTK-GP). This can be more efficient for very large networks or when only few data points are available. Implementation aspects for fast and efficient computation of the correction term, as well as the initial state estimation for the RNN model are described. Numerical examples show the effectiveness of the proposed methodology in presence of significant system variations.
Recommendations
- Continuous-time system identification with neural networks: model structures and fitting criteria
- Synthesis of recurrent neural networks for dynamical system simulation
- Adaptive Capability of Recurrent Neural Networks with Fixed Weights for Series-Parallel System Identification
- Metric entropy limits on recurrent neural network learning of linear dynamical systems
- Mechanical system modelling using recurrent neural networks via quasi- Newton learning methods
Cites work
- Continuous-time system identification with neural networks: model structures and fitting criteria
- Gaussian processes for machine learning.
- scientific article; zbMATH DE number 6982909 (Why is no real title available?)
- scientific article; zbMATH DE number 3997615 (Why is no real title available?)
- scientific article; zbMATH DE number 5060482 (Why is no real title available?)
- The elements of statistical learning. Data mining, inference, and prediction
- The Koopman operator in systems and control. Concepts, methodologies and applications
Cited in
(10)- Dynamic system identification via recurrent multilayer perceptrons
- \(\mathcal L_{2}-\mathcal L_{\infty }\) nonlinear system identification via recurrent neural networks
- Identification of Discrete Event Systems Using the Compound Recurrent Neural Network: Extracting DEVS from Trained Network
- scientific article; zbMATH DE number 3869194 (Why is no real title available?)
- Adaptive Capability of Recurrent Neural Networks with Fixed Weights for Series-Parallel System Identification
- Fast iterative sample transfer identification method for dynamic systems under non-identical distribution
- Heterogeneous mixed kernel correntropy-based robust iterative estimation methods and convergence analysis for the nonlinear system with outliers
- Uncertainty quantification in neural state-space models: applications for experiment design and uncertainty-aware MPC
- Kinematics-informed neural network control on SO(3)
- A neural network-based approach to hybrid systems identification for control
This page was built for publication: On the adaptation of recurrent neural networks for system identification
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6109982)
