Abstract Two-photon lithography (TPL) is a direct laser writing process that enables the fabrication of cm-scale complex three-dimensional polymeric structures with submicrometer resolution. In contrast to the slow and serial writing scheme of conventional TPL, projection TPL (P-TPL) enables rapid printing of entire layers at once. However, process prediction remains a significant challenge in P-TPL due to the lack of computationally efficient models. In this work, we present machine learning-based surrogate models to predict the outcomes of P-TPL to >98% of the accuracy of a physics-based reaction-diffusion finite element simulation. A classification neural network was trained using data generated from the physics-based simulations. This enabled us to achieve computationally efficient and accurate prediction of whether a set of printing conditions will result in precise and controllable polymerization and the desired printing versus no printing or runaway polymerization. We interrogate this surrogate model to investigate the parameter regimes that are promising for successful printing. We predict combinations of photoresist reaction rate constants that are necessary to print for a given set of processing conditions, thereby generating a set of printability maps. The surrogate models reduced the computational time that is required to generate these maps from more than 10 months to less than a second. Thus, these models can enable rapid and informed selection of photoresists and printing parameters during process control and optimization.
more »
« less
Machine learning-enhanced model-based scenario optimization for DIII-D
Scenario development in tokamaks is an open area of investigation that can be approached in a variety of different ways. Experimental trial and error has been the traditional method, but this required a massive amount of experimental time and resources. As high fidelity predictive models have become available, offline development and testing of proposed scenarios has become an option to reduce the required experimental resources. The use of predictive models also offers the possibility of using a numerical optimization process to find the controllable inputs that most closely achieve the desired plasma state. However, this type of optimization can require as many as hundreds or thousands of predictive simulation cases to converge to a solution; many of the commonly used high fidelity models have high computational burdens, so it is only reasonable to run a handful of predictive simulations. In order to make use of numerical optimization approaches, a compromise needs to be found between model fidelity and computational burden. This compromise can be achieved using neural networks surrogates of high fidelity models that retain nearly the same level of accuracy as the models they are trained to replicate while reducing the computation time by orders of magnitude. In this work, a model-based numerical optimization tool for scenario development is described. The predictive model used by the optimizer includes neural network surrogate models integrated into the fast Control-Oriented Transport simulation framework. This optimization scheme is able to converge to the optimal values of the controllable inputs that produce the target plasma scenario by running thousands of predictive simulations in under an hour without sacrificing too much prediction accuracy.
more » « less- NSF-PAR ID:
- 10497292
- Publisher / Repository:
- IOP Publishing
- Date Published:
- Journal Name:
- Nuclear Fusion
- Volume:
- 64
- Issue:
- 5
- ISSN:
- 0029-5515
- Format(s):
- Medium: X Size: Article No. 056018
- Size(s):
- ["Article No. 056018"]
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Simulating the time evolution of physical systems is pivotal in many scientific and engineering problems. An open challenge in simulating such systems is their multi-resolution dynamics: a small fraction of the system is extremely dynamic, and requires very fine-grained resolution, while a majority of the system is changing slowly and can be modeled by coarser spatial scales. Typical learning-based surrogate models use a uniform spatial scale, which needs to resolve to the finest required scale and can waste a huge compute to achieve required accuracy. We introduced Learning controllable Adaptive simulation for Multiresolution Physics (LAMP) as the first full deep learning-based surrogate model that jointly learns the evolution model and optimizes appropriate spatial resolutions that devote more compute to the highly dynamic regions. LAMP consists of a Graph Neural Network (GNN) for learning the forward evolution, and a GNNbased actor-critic for learning the policy of spatial refinement and coarsening. We introduced learning techniques that optimize LAMP with weighted sum of error and computational cost as objective, allowing LAMP to adapt to varying relative importance of error vs. computation tradeoff at inference time. We evaluated our method in a 1D benchmark of nonlinear PDEs and a challenging 2D mesh-based simulation. We demonstrated that our LAMP outperforms state-of-the-art deep learning surrogate models, and can adaptively trade-off computation to improve long-term prediction error: it achieves an average of 33.7% error reduction for 1D nonlinear PDEs, and outperforms MeshGraphNets + classical Adaptive Mesh Refinement (AMR) in 2D mesh-based simulations.more » « less
-
Fast inference of numerical model parameters from data is an important prerequisite to generate predictive models for a wide range of applications. Use of sampling-based approaches such as Markov chain Monte Carlo may become intractable when each likelihood evaluation is computationally expensive. New approaches combining variational inference with normalizing flow are characterized by a computational cost that grows only linearly with the dimensionality of the latent variable space, and rely on gradient-based optimization instead of sampling, providing a more efficient approach for Bayesian inference about the model parameters. Moreover, the cost of frequently evaluating an expensive likelihood can be mitigated by replacing the true model with an offline trained surrogate model, such as neural networks. However, this approach might generate significant bias when the surrogate is insufficiently accurate around the posterior modes. To reduce the computational cost without sacrificing inferential accuracy, we propose Normalizing Flow with Adaptive Surrogate (NoFAS), an optimization strategy that alternatively updates the normalizing flow parameters and surrogate model parameters. We also propose an efficient sample weighting scheme for surrogate model training that preserves global accuracy while effectively capturing high posterior density regions. We demonstrate the inferential and computational superiority of NoFAS against various benchmarks, including cases where the underlying model lacks identifiability. The source code and numerical experiments used for this study are available at https://github.com/cedricwangyu/NoFAS.more » « less
-
more » « less
Abstract Radio-frequency quadrupoles (RFQs) are multi-purpose linear particle accelerators that simultaneously bunch and accelerate charged particle beams. They are ubiquitous in accelerator physics, especially as injectors to higher-energy machines, owing to their impressive efficiency. The design and optimization of these devices can be lengthy due to the need to repeatedly perform high-fidelity simulations. Several recent papers have demonstrated that machine learning can be used to build surrogate models (fast-executing replacements of computationally costly beam simulations) for order-of-magnitude computing time speedups. However, while these pilot studies are encouraging, there is room to improve their predictive accuracy. Particularly, beam summary statistics such as emittances (an important figure of merit in particle accelerator physics) have historically been challenging to predict. For the first time, we present a surrogate model trained on 200 000 samples that yields
-
Safe operations of autonomous mobile robots in close proximity to humans, creates a need for enhanced trajectory tracking (with low tracking errors). Linear optimal control techniques such as Linear Quadratic Regulator (LQR) and Model Predictive Control (MPC) have been used successfully for low-speed applications while leveraging their model-based methodology with manageable computational demands. However, model and parameter uncertainties or other unmodeled nonlinearities may cause poor control actions and constraint violations. Nonlinear MPC has emerged as an alternate optimal-control approach but needs to overcome real-time deployment challenges (including fast sampling time, design complexity, and limited computational resources). In recent years, the optimal control-based deployments have benefitted enormously from the ability of Deep Neural Networks (DNNs) to serve as universal function approximators. This has led to deployments in a plethora of previously inaccessible applications – but many aspects of generalizability, benchmarking, and systematic verification and validation coupled with benchmarking have emerged. This paper presents a novel approach to fusing Deep Reinforcement Learning-based (DRL) longitudinal control with a traditional PID lateral controller for autonomous navigation. Our approach follows (i) Generation of an adequate fidelity simulation scenario via a Real2Sim approach; (ii) training a DRL agent within this framework; (iii) Testing the performance and generalizability on alternate scenarios. We use an initial tuned set of the lateral PID controller gains for observing the vehicle response over a range of velocities. Then we use a DRL framework to generate policies for an optimal longitudinal controller that successfully complements the lateral PID to give the best tracking performance for the vehicle.more » « less
