More Like this

1. Neural Networks with Physics-Informed Architectures and Constraints for Dynamical Systems Modeling

   Djeumou, Franck.; Neary, Cyrus.; Goubault, Eric.; Putot, Sylvie.; Topcu, Ufuk. (January 2022, 4th Annual Conference on Learning for Dynamics and Control)

   Firoozi, R.; Mehr, N.; Yel, E.; Antonova, R.; Bohg, J.; Schwager, M.; Kochenderfer, M. (Ed.)

   Effective inclusion of physics-based knowledge into deep neural network models of dynamical sys- tems can greatly improve data efficiency and generalization. Such a priori knowledge might arise from physical principles (e.g., conservation laws) or from the system’s design (e.g., the Jacobian matrix of a robot), even if large portions of the system dynamics remain unknown. We develop a framework to learn dynamics models from trajectory data while incorporating a priori system knowledge as inductive bias. More specifically, the proposed framework uses physics-based side information to inform the structure of the neural network itself, and to place constraints on the values of the outputs and the internal states of the model. It represents the system’s vector field as a composition of known and unknown functions, the latter of which are parametrized by neural networks. The physics-informed constraints are enforced via the augmented Lagrangian method during the model’s training. We experimentally demonstrate the benefits of the proposed approach on a variety of dynamical systems – including a benchmark suite of robotics environments featur- ing large state spaces, non-linear dynamics, external forces, contact forces, and control inputs. By exploiting a priori system knowledge during training, the proposed approach learns to predict the system dynamics two orders of magnitude more accurately than a baseline approach that does not include prior knowledge, given the same training dataset. 

   more » « less
2. Physics-Constrained Taylor Neural Networks for Learning and Control of Dynamical Systems

   Nguyen, Nam T; Tique, Juan C; Nghiem, Truong X (July 2025, IEEE)

   Data-driven approaches are increasingly popular for identifying dynamical systems due to improved accuracy and availability of sensor data. However, relying solely on data for identification does not guarantee that the identified systems will maintain their physical properties or that the predicted models will generalize well. In this paper, we propose a novel method for data-driven system identification by integrating a neural network as the first-order derivative of the learned dynamics in a Taylor series instead of learning the dynamical function directly. In addition, for dynamical systems with known monotonic properties, our approach can ensure monotonicity by constraining the neural network derivative to be non-positive or non-negative to the respective inputs, resulting in Monotonic Taylor Neural Networks (MTNN). Such constraints are enforced by either a specialized neural network architecture or regularization in the loss function for training. The proposed method demonstrates better performance compared to methods without the physics-based monotonicity constraints when tested on experimental data from an HVAC system and a temperature control testbed. Furthermore, MTNN shows good performance in a control application of a model predictive controller for a nonlinear MIMO system, illustrating the practical application of our method. 

   more » « less
3. Probability flow solution of the Fokker–Planck equation

   https://doi.org/10.1088/2632-2153/ace2aa  

   Boffi, Nicholas M.; Vanden-Eijnden, Eric (July 2023, Machine Learning: Science and Technology)

   Abstract The method of choice for integrating the time-dependent Fokker–Planck equation (FPE) in high-dimension is to generate samples from the solution via integration of the associated stochastic differential equation (SDE). Here, we study an alternative scheme based on integrating an ordinary differential equation that describes the flow of probability. Acting as a transport map, this equation deterministically pushes samples from the initial density onto samples from the solution at any later time. Unlike integration of the stochastic dynamics, the method has the advantage of giving direct access to quantities that are challenging to estimate from trajectories alone, such as the probability current, the density itself, and its entropy. The probability flow equation depends on the gradient of the logarithm of the solution (its ‘score’), and so isa-prioriunknown. To resolve this dependence, we model the score with a deep neural network that is learned on-the-fly by propagating a set of samples according to the instantaneous probability current. We show theoretically that the proposed approach controls the Kullback–Leibler (KL) divergence from the learned solution to the target, while learning on external samples from the SDE does not control either direction of the KL divergence. Empirically, we consider several high-dimensional FPEs from the physics of interacting particle systems. We find that the method accurately matches analytical solutions when they are available as well as moments computed via Monte-Carlo when they are not. Moreover, the method offers compelling predictions for the global entropy production rate that out-perform those obtained from learning on stochastic trajectories, and can effectively capture non-equilibrium steady-state probability currents over long time intervals. 

   more » « less
4. Deconstructing the Inductive Biases of Hamiltonian Neural Networks

   Gruver, Nate; Finzi, Marc A; Stanton, Samuel; Wilson, Andrew (January 2022, International Conference on Learning Representations)

   Physics-inspired neural networks (NNs), such as Hamiltonian or Lagrangian NNs, dramatically outperform other learned dynamics models by leveraging strong inductive biases. These models, however, are challenging to apply to many real world systems, such as those that don’t conserve energy or contain contacts, a common setting for robotics and reinforcement learning. In this paper, we examine the inductive biases that make physics-inspired models successful in practice. We show that, contrary to conventional wisdom, the improved generalization of HNNs is the result of modeling acceleration directly and avoiding artificial complexity from the coordinate system, rather than symplectic structure or energy conservation. We show that by relaxing the inductive biases of these models, we can match or exceed performance on energy-conserving systems while dramatically improving performance on practical, non-conservative systems. We extend this approach to constructing transition models for common Mujoco environments, showing that our model can appropriately balance inductive biases with the flexibility required for model-based control. 

   more » « less
5. Knowledge Guided Machine Learning for Extracting, Preserving, and Adapting Physics-aware Features

   https://doi.org/10.1137/1.9781611978032.82  

   He, Erhu; Xie, Yiqun; Liu, Licheng; Jin, Zhenong; Zhang, Dajun; Jia, Xiaowei (April 2024, SIAM International Conference on Data Mining (SDM) 2024)

   Training machine learning (ML) models for scientific problems is often challenging due to limited observation data. To overcome this challenge, prior works commonly pre-train ML models using simulated data before having them fine-tuned with small real data. Despite the promise shown in initial research across different domains, these methods cannot ensure improved performance after fine-tuning because (i) they are not designed for extracting generalizable physics-aware features during pre-training, (ii) the features learned from pre-training can be distorted by the fine-tuning process. In this paper, we propose a new learning method for extracting, preserving, and adapting physics-aware features. We build a knowledge-guided neural network (KGNN) model based on known dependencies amongst physical variables, which facilitate extracting physics-aware feature representation from simulated data. Then we fine-tune this model by alternately updating the encoder and decoder of the KGNN model to enhance the prediction while preserving the physics-aware features learned through pre-training. We further propose to adapt the model to new testing scenarios via a teacher-student learning framework based on the model uncertainty. The results demonstrate that the proposed method outperforms many baselines by a good margin, even using sparse training data or under out-of-sample testing scenarios. 

   more » « less