Abstract:In recent years, model optimization in the field of computational biology has become a prominent area for development of pharmaceutical drugs. The increased amount of experimental data leads to the increase in complexity of proposed models. With increased complexity comes a necessity for computational algorithms that are able to handle the large datasets that are used to fit model parameters. In this study the ability of simultaneous, hybrid simultaneous, and sequential algorithms are tested on two models representative of computational systems biology. The first case models the cells affected by a virus in a population and serves as a benchmark model for the proposed hybrid algorithm. The second model is the ErbB model and shows the ability of the hybrid sequential and simultaneous method to solve large-scale biological models. Post-processing analysis reveals insights into the model formulation that are important for understanding the specific parameter optimization. A parameter sensitivity analysis reveals shortcomings and difficulties in the ErbB model parameter optimization due to the model formulation rather than the solver capacity. Suggested methods are model reformulation to improve input-to-output model linearity, sensitivity ranking, and choice of solver.
Forecasting models are a central part of many control systems, where high-consequence decisions must be made on long latency control variables. These models are particularly relevant for emerging artificial intelligence (AI)-guided instrumentation, in which prescriptive knowledge is needed to guide autonomous decision-making. Here we describe the implementation of a long short-term memory model (LSTM) for forecasting in situ electron energy loss spectroscopy (EELS) data, one of the richest analytical probes of materials and chemical systems. We describe key considerations for data collection, preprocessing, training, validation, and benchmarking, showing how this approach can yield powerful predictive insight into order-disorder phase transitions. Finally, we comment on how such a model may integrate with emerging AI-guided instrumentation for powerful high-speed experimentation.
The microstructure of commercially made lithium-ion battery electrodes is not necessarily optimal for cell performance. In an attempt to optimize the performance, detailed understanding of the microstructure is necessary. Factors that influence electrode microstructure include high-level variables such as composition and porosity; as well as detailed fabrication conditions that are part of the film-making, drying, and calendering steps. Here we report on continuing efforts to model the particle-level microstructure of porous electrodes and how this relates to fabrication conditions [1]. LAMMPS, a molecular simulation code, was adapted for these mesoscale particle simulations. In particular we simulate an NCM lithium-ion cathode using a superposition of spheres to imitate active particles, carbon, binder, and solvent. Equations of motion coupled to inter-particle forces are solved to simulate particle motion and subsequent immobilization during fabrication steps. Our model is uniquely able to capture the essential physics in both the slurry and dried film. Model results are validated by comparing to experimental results for microstructure, tortuosity, and mechanical properties like viscosity and elasticity. Such simulations are a first step in allowing us to predict electrode microstructure and therefore battery performance from fundamental fabrication conditions. This contributes to the long-term goal of this work, namely to increase the performance of electrodes by improved control of the manufacturing process. This work was supported by the U.S. Department of Energy through the BMR program. References: [1] M. Forouzan et al., J. Power Sources 312, 172-183 (2016). Figure 1: Sequence of images (top to bottom) showing simulated electrode coating process with slurry film (green) moving relative to blade (gray) at shear rate 250 s-1. Figure 1
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.