Optimization and control of a thin film growth process: A hybrid first principles/artificial neural network based multiscale modelling approach

Iterative multiscale and multi-physics computations for operando catalyst nanostructure elucidation and kinetic modeling

Modern heterogeneous catalysis has benefitted immensely from computational predictions of catalyst structure and its evolution under reaction conditions, first-principles mechanistic investigations, and detailed kinetic modeling, which are rungs on a multiscale workflow. Establishing connections across these rungs and integration with experiments have been challenging. Here, operando catalyst structure prediction techniques using density functional theory simulations and ab initio thermodynamics calculations, molecular dynamics, and machine learning techniques are presented. Surface structure characterization by computational spectroscopic and machine learning techniques is then discussed. Hierarchical approaches in kinetic parameter estimation involving semi-empirical, data-driven, and first-principles calculations and detailed kinetic modeling via mean-field microkinetic modeling and kinetic Monte Carlo simulations are discussed along with methods and the need for uncertainty quantification. With these as the background, this article proposes a bottom-up hierarchical and closed loop modeling framework incorporating consistency checks and iterative refinements at each level and across levels.

Industrial data science – a review of machine learning applications for chemical and process industries

Understand and optimize industrial processes via machine learning and chemical engineering principles.

Artificial Neural Networks Based Optimization Techniques: A Review

In the last few years, intensive research has been done to enhance artificial intelligence (AI) using optimization techniques. In this paper, we present an extensive review of artificial neural networks (ANNs) based optimization algorithm techniques with some of the famous optimization techniques, e.g., genetic algorithm (GA), particle swarm optimization (PSO), artificial bee colony (ABC), and backtracking search algorithm (BSA) and some modern developed techniques, e.g., the lightning search algorithm (LSA) and whale optimization algorithm (WOA), and many more. The entire set of such techniques is classified as algorithms based on a population where the initial population is randomly created. Input parameters are initialized within the specified range, and they can provide optimal solutions. This paper emphasizes enhancing the neural network via optimization algorithms by manipulating its tuned parameters or training parameters to obtain the best structure network pattern to dissolve the problems in the best way. This paper includes some results for improving the ANN performance by PSO, GA, ABC, and BSA optimization techniques, respectively, to search for optimal parameters, e.g., the number of neurons in the hidden layers and learning rate. The obtained neural net is used for solving energy management problems in the virtual power plant system.

Performance enhancement of commercial ethylene oxide reactor by artificial intelligence approach

The present work emphasizes the development of a generic methodology that addresses the core issue of any running chemical plant, i.e., how to maintain a delicate balance between profit and environmental impact. Here, an ethylene oxide (EO) production plant has been taken as a case study. The production of EO takes place in a multiphase catalytic reactor, the reliable first principle-based model of which is still not available in the literature. Artificial neural network (ANN) was therefore applied to develop a data-driven model of the complex reactor with the help of actual industrial data. The model successfully built up a correlation between the catalyst selectivity and temperature with other operational parameters. A hybrid multi-objective metaheuristic optimization technique, namely ANN-multi-objective genetic algorithm (MOGA) algorithm was used to develop a Pareto diagram of selectivity versus reactor temperature. The Pareto diagram will help the plant engineers to make a strategy on what operating conditions to be maintained to make a delicate balance between profit and environmental impact. It was also found that by applying this hybrid ANN-MOGA modeling and optimization technique, for a 720 KTA ethylene glycol plant, approximately 32,345 ton/year of carbon-di-oxide emission into the atmosphere can be reduced. Along with the reduction of environmental impact, this hybrid approach enables the plant to reduce raw material cost of nine million USD per annum simultaneously.

Development of a hybrid model for a partially known intracellular signaling pathway through correction term estimation and neural network modeling

Developing an accurate first-principle model is an important step in employing systems biology approaches to analyze an intracellular signaling pathway. However, an accurate first-principle model is difficult to be developed since it requires in-depth mechanistic understandings of the signaling pathway. Since underlying mechanisms such as the reaction network structure are not fully understood, significant discrepancy exists between predicted and actual signaling dynamics. Motivated by these considerations, this work proposes a hybrid modeling approach that combines a first-principle model and an artificial neural network (ANN) model so that predictions of the hybrid model surpass those of the original model. First, the proposed approach determines an optimal subset of model states whose dynamics should be corrected by the ANN by examining the correlation between each state and outputs through relative order. Second, an L2-regularized least-squares problem is solved to infer values of the correction terms that are necessary to minimize the discrepancy between the model predictions and available measurements. Third, an ANN is developed to generalize relationships between the values of the correction terms and the system dynamics. Lastly, the original first-principle model is coupled with the developed ANN to finalize the hybrid model development so that the model will possess generalized prediction capabilities while retaining the model interpretability. We have successfully validated the proposed methodology with two case studies, simplified apoptosis and lipopolysaccharide-induced NFκB signaling pathways, to develop hybrid models with in silico and in vitro measurements, respectively.

An intracellular signaling pathway is often represented by a set of nonlinear ordinary differential equations, which translate our current knowledge about the signaling pathway into a testable mathematical model. However, predictions from such models are often subject to high uncertainty since many signaling pathways are only partially known beforehand. In this study, we propose a systematic approach to develop a hybrid model to improve model accuracy by combining machine learning and the first-principle modeling. Specifically, model correction terms are learned from discrepancy between model predictions and measurements, and these terms are added to the first-principle model to enhance the prediction accuracy. Once these correction terms are learned from the data, an artificial neural network (ANN) model is developed to find an empirical relation between the model and the correction terms so that the developed ANN can be used to posses improved predictive capabilities even in new operating conditions (i.e., generalizability). The final hybrid model is then constructed by coupling the first-principle model with the developed ANN.

Assessing conformal thin film growth under nonstochastic deposition conditions: application of a phenomenological model of roughness replication to synthetic topographic images

In this work, a simple method to follow the evolution of the surface of thin films during growth on substrates characterised by high roughness is detailed. To account for real cases as much as possible, the approach presented is based on the hypothesis that deposition takes place under nonstochastic conditions, such as those typical of many thin film processes in industry and technology. In this context, previous models for roughness replication, which are mainly based on idealised deposition conditions, cannot be applied and thus ad hoc approaches are required for achieving quantitative predictions. Here it is suggested that under nonstochastic conditions a phenomenological relation can be proposed, mainly based on local roughening of surface, to monitor the statistical similarity between the film and the substrate during growth or, in other words, to detect changes of the bare substrate morphological profile occurring during the film growth on top. Such approximation is based on surface representation in terms of power spectral density of surface heights, derived from topographic images; in this work, such method will be tested on two separate batches of synthetic images which simulate thin films growth onto a real rough substrate. In particular, two growth models will be implemented: the first reproduces the surface profile obtained during an atomic force microscopy measurement by using a simple geometrical envelope of surface, regardless the thin film growth mechanism; the second reproduces the columnar growth expected under nonstochastic deposition conditions. It will be shown that the approach introduced is capable to highlight differences between the two batches and, in the second case, to quantitatively account for the replication of the substrate roughness during growth. The results obtained here are potentially interesting in that they account essentially for the geometrical features of the surfaces, and as such they can be applied to synthetic depositions that reproduce different thin film depositions and experimental contexts.

Integrating Feedback Control and Run-to-Run Control in Multi-Wafer Thermal Atomic Layer Deposition of Thin Films

There is currently a lack of understanding of the deposition profile in a batch atomic layer deposition (ALD) process. Also, no on-line control scheme has been proposed to resolve the prevalent disturbances. Motivated by this, we develop a computational fluid dynamics (CFD) model and an integrated online run-to-run and feedback control scheme. Specifically, we analyze a furnace reactor for a SiO2 thin-film ALD with BTBAS and ozone as precursors. Initially, a high-fidelity 2D axisymmetric multiscale CFD model is developed using ANSYS Fluent for the gas-phase characterization and the surface thin-film deposition, based on a kinetic Monte-Carlo (kMC) model database. To deal with the disturbance during reactor operation, a proportional integral (PI) control scheme is adopted, which manipulates the inlet precursor concentration to drive the precursor partial pressure to the set-point, ensuring the complete substrate coverage. Additionally, the CFD model is utilized to investigate a wide range of operating conditions, and a regression model is developed to describe the relationship between the half-cycle time and the feed flow rate. A run-to-run (R2R) control scheme using an exponentially weighted moving average (EWMA) strategy is developed to regulate the half-cycle time for the furnace ALD process between batches.

Real-Time Optimization and Control of Nonlinear Processes Using Machine Learning

Machine learning has attracted extensive interest in the process engineering field, due to the capability of modeling complex nonlinear process behavior. This work presents a method for combining neural network models with first-principles models in real-time optimization (RTO) and model predictive control (MPC) and demonstrates the application to two chemical process examples. First, the proposed methodology that integrates a neural network model and a first-principles model in the optimization problems of RTO and MPC is discussed. Then, two chemical process examples are presented. In the first example, a continuous stirred tank reactor (CSTR) with a reversible exothermic reaction is studied. A feed-forward neural network model is used to approximate the nonlinear reaction rate and is combined with a first-principles model in RTO and MPC. An RTO is designed to find the optimal reactor operating condition balancing energy cost and reactant conversion, and an MPC is designed to drive the process to the optimal operating condition. A variation in energy price is introduced to demonstrate that the developed RTO scheme is able to minimize operation cost and yields a closed-loop performance that is very close to the one attained by RTO/MPC using the first-principles model. In the second example, a distillation column is used to demonstrate an industrial application of the use of machine learning to model nonlinearities in RTO. A feed-forward neural network is first built to obtain the phase equilibrium properties and then combined with a first-principles model in RTO, which is designed to maximize the operation profit and calculate optimal set-points for the controllers. A variation in feed concentration is introduced to demonstrate that the developed RTO scheme can increase operation profit for all considered conditions.